diff --git "a/checkpoint-4218/trainer_state.json" "b/checkpoint-4218/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-4218/trainer_state.json" @@ -0,0 +1,29559 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.9994666666666667, + "eval_steps": 500, + "global_step": 4218, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0007111111111111111, + "grad_norm": 1480.1832275390625, + "learning_rate": 4.739336492890996e-08, + "loss": 7.0544, + "step": 1 + }, + { + "epoch": 0.0014222222222222223, + "grad_norm": 1544.5272216796875, + "learning_rate": 9.478672985781992e-08, + "loss": 7.3129, + "step": 2 + }, + { + "epoch": 0.0021333333333333334, + "grad_norm": 1485.66259765625, + "learning_rate": 1.4218009478672986e-07, + "loss": 7.0652, + "step": 3 + }, + { + "epoch": 0.0028444444444444446, + "grad_norm": 1452.4617919921875, + "learning_rate": 1.8957345971563984e-07, + "loss": 7.1338, + "step": 4 + }, + { + "epoch": 0.0035555555555555557, + "grad_norm": 1482.7576904296875, + "learning_rate": 2.3696682464454978e-07, + "loss": 7.1056, + "step": 5 + }, + { + "epoch": 0.004266666666666667, + "grad_norm": 1447.8955078125, + "learning_rate": 2.843601895734597e-07, + "loss": 7.0752, + "step": 6 + }, + { + "epoch": 0.004977777777777778, + "grad_norm": 1340.7593994140625, + "learning_rate": 3.317535545023697e-07, + "loss": 6.9306, + "step": 7 + }, + { + "epoch": 0.005688888888888889, + "grad_norm": 770.9786987304688, + "learning_rate": 3.791469194312797e-07, + "loss": 6.0171, + "step": 8 + }, + { + "epoch": 0.0064, + "grad_norm": 759.1358032226562, + "learning_rate": 4.265402843601896e-07, + "loss": 5.9473, + "step": 9 + }, + { + "epoch": 0.0071111111111111115, + "grad_norm": 711.0708618164062, + "learning_rate": 4.7393364928909956e-07, + "loss": 5.8716, + "step": 10 + }, + { + "epoch": 0.007822222222222222, + "grad_norm": 785.1265258789062, + "learning_rate": 5.213270142180095e-07, + "loss": 5.2725, + "step": 11 + }, + { + "epoch": 0.008533333333333334, + "grad_norm": 734.9464111328125, + "learning_rate": 5.687203791469194e-07, + "loss": 5.1252, + "step": 12 + }, + { + "epoch": 0.009244444444444444, + "grad_norm": 694.8941650390625, + "learning_rate": 6.161137440758294e-07, + "loss": 4.9818, + "step": 13 + }, + { + "epoch": 0.009955555555555556, + "grad_norm": 670.718505859375, + "learning_rate": 6.635071090047394e-07, + "loss": 4.8992, + "step": 14 + }, + { + "epoch": 0.010666666666666666, + "grad_norm": 382.98980712890625, + "learning_rate": 7.109004739336493e-07, + "loss": 3.6864, + "step": 15 + }, + { + "epoch": 0.011377777777777778, + "grad_norm": 525.0732421875, + "learning_rate": 7.582938388625594e-07, + "loss": 3.4779, + "step": 16 + }, + { + "epoch": 0.012088888888888889, + "grad_norm": 534.5843505859375, + "learning_rate": 8.056872037914692e-07, + "loss": 3.0006, + "step": 17 + }, + { + "epoch": 0.0128, + "grad_norm": 275.4285583496094, + "learning_rate": 8.530805687203792e-07, + "loss": 2.6444, + "step": 18 + }, + { + "epoch": 0.013511111111111111, + "grad_norm": 161.94793701171875, + "learning_rate": 9.004739336492892e-07, + "loss": 2.4706, + "step": 19 + }, + { + "epoch": 0.014222222222222223, + "grad_norm": 919.6431274414062, + "learning_rate": 9.478672985781991e-07, + "loss": 2.8082, + "step": 20 + }, + { + "epoch": 0.014933333333333333, + "grad_norm": 1683.3765869140625, + "learning_rate": 9.95260663507109e-07, + "loss": 2.5242, + "step": 21 + }, + { + "epoch": 0.015644444444444443, + "grad_norm": 138.06727600097656, + "learning_rate": 1.042654028436019e-06, + "loss": 2.0763, + "step": 22 + }, + { + "epoch": 0.016355555555555557, + "grad_norm": 238.34657287597656, + "learning_rate": 1.090047393364929e-06, + "loss": 1.9594, + "step": 23 + }, + { + "epoch": 0.017066666666666667, + "grad_norm": 134.41326904296875, + "learning_rate": 1.1374407582938388e-06, + "loss": 1.8378, + "step": 24 + }, + { + "epoch": 0.017777777777777778, + "grad_norm": 176.9913330078125, + "learning_rate": 1.184834123222749e-06, + "loss": 1.8537, + "step": 25 + }, + { + "epoch": 0.018488888888888888, + "grad_norm": 386.27081298828125, + "learning_rate": 1.2322274881516587e-06, + "loss": 3.1778, + "step": 26 + }, + { + "epoch": 0.0192, + "grad_norm": 423.3983459472656, + "learning_rate": 1.2796208530805687e-06, + "loss": 3.1127, + "step": 27 + }, + { + "epoch": 0.019911111111111112, + "grad_norm": 259.02099609375, + "learning_rate": 1.3270142180094788e-06, + "loss": 2.6918, + "step": 28 + }, + { + "epoch": 0.020622222222222222, + "grad_norm": 231.3286590576172, + "learning_rate": 1.3744075829383887e-06, + "loss": 2.2661, + "step": 29 + }, + { + "epoch": 0.021333333333333333, + "grad_norm": 126.32612609863281, + "learning_rate": 1.4218009478672987e-06, + "loss": 2.0996, + "step": 30 + }, + { + "epoch": 0.022044444444444443, + "grad_norm": 499.4660949707031, + "learning_rate": 1.4691943127962086e-06, + "loss": 2.0082, + "step": 31 + }, + { + "epoch": 0.022755555555555557, + "grad_norm": 78.92438507080078, + "learning_rate": 1.5165876777251187e-06, + "loss": 1.801, + "step": 32 + }, + { + "epoch": 0.023466666666666667, + "grad_norm": 78.19351196289062, + "learning_rate": 1.5639810426540287e-06, + "loss": 1.7307, + "step": 33 + }, + { + "epoch": 0.024177777777777777, + "grad_norm": 51.95500946044922, + "learning_rate": 1.6113744075829384e-06, + "loss": 1.6417, + "step": 34 + }, + { + "epoch": 0.024888888888888887, + "grad_norm": 47.872398376464844, + "learning_rate": 1.6587677725118483e-06, + "loss": 1.5986, + "step": 35 + }, + { + "epoch": 0.0256, + "grad_norm": 26.960590362548828, + "learning_rate": 1.7061611374407585e-06, + "loss": 1.4681, + "step": 36 + }, + { + "epoch": 0.02631111111111111, + "grad_norm": 17.61344337463379, + "learning_rate": 1.7535545023696684e-06, + "loss": 1.4302, + "step": 37 + }, + { + "epoch": 0.027022222222222222, + "grad_norm": 19.154991149902344, + "learning_rate": 1.8009478672985784e-06, + "loss": 1.3414, + "step": 38 + }, + { + "epoch": 0.027733333333333332, + "grad_norm": 15.859538078308105, + "learning_rate": 1.8483412322274883e-06, + "loss": 1.2776, + "step": 39 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 14.852712631225586, + "learning_rate": 1.8957345971563982e-06, + "loss": 1.224, + "step": 40 + }, + { + "epoch": 0.029155555555555556, + "grad_norm": 11.642457008361816, + "learning_rate": 1.943127962085308e-06, + "loss": 1.148, + "step": 41 + }, + { + "epoch": 0.029866666666666666, + "grad_norm": 11.555150985717773, + "learning_rate": 1.990521327014218e-06, + "loss": 1.1438, + "step": 42 + }, + { + "epoch": 0.030577777777777777, + "grad_norm": 10.368565559387207, + "learning_rate": 2.037914691943128e-06, + "loss": 1.1025, + "step": 43 + }, + { + "epoch": 0.03128888888888889, + "grad_norm": 8.658889770507812, + "learning_rate": 2.085308056872038e-06, + "loss": 1.0243, + "step": 44 + }, + { + "epoch": 0.032, + "grad_norm": 8.126867294311523, + "learning_rate": 2.1327014218009483e-06, + "loss": 1.0131, + "step": 45 + }, + { + "epoch": 0.032711111111111114, + "grad_norm": 8.33288860321045, + "learning_rate": 2.180094786729858e-06, + "loss": 0.9836, + "step": 46 + }, + { + "epoch": 0.03342222222222222, + "grad_norm": 7.693348407745361, + "learning_rate": 2.2274881516587678e-06, + "loss": 0.9934, + "step": 47 + }, + { + "epoch": 0.034133333333333335, + "grad_norm": 8.185802459716797, + "learning_rate": 2.2748815165876777e-06, + "loss": 0.9503, + "step": 48 + }, + { + "epoch": 0.03484444444444444, + "grad_norm": 12.168658256530762, + "learning_rate": 2.322274881516588e-06, + "loss": 0.9128, + "step": 49 + }, + { + "epoch": 0.035555555555555556, + "grad_norm": 8.569781303405762, + "learning_rate": 2.369668246445498e-06, + "loss": 0.9199, + "step": 50 + }, + { + "epoch": 0.03626666666666667, + "grad_norm": 6.235218524932861, + "learning_rate": 2.417061611374408e-06, + "loss": 0.9084, + "step": 51 + }, + { + "epoch": 0.036977777777777776, + "grad_norm": 15.239593505859375, + "learning_rate": 2.4644549763033174e-06, + "loss": 0.8608, + "step": 52 + }, + { + "epoch": 0.03768888888888889, + "grad_norm": 13.00948715209961, + "learning_rate": 2.5118483412322274e-06, + "loss": 0.8314, + "step": 53 + }, + { + "epoch": 0.0384, + "grad_norm": 7.194519519805908, + "learning_rate": 2.5592417061611373e-06, + "loss": 0.8527, + "step": 54 + }, + { + "epoch": 0.03911111111111111, + "grad_norm": 10.14979362487793, + "learning_rate": 2.606635071090048e-06, + "loss": 0.7809, + "step": 55 + }, + { + "epoch": 0.039822222222222224, + "grad_norm": 25.219594955444336, + "learning_rate": 2.6540284360189576e-06, + "loss": 0.8643, + "step": 56 + }, + { + "epoch": 0.04053333333333333, + "grad_norm": 11.078099250793457, + "learning_rate": 2.7014218009478675e-06, + "loss": 0.7833, + "step": 57 + }, + { + "epoch": 0.041244444444444445, + "grad_norm": 12.409817695617676, + "learning_rate": 2.7488151658767775e-06, + "loss": 0.7844, + "step": 58 + }, + { + "epoch": 0.04195555555555556, + "grad_norm": 5.695272922515869, + "learning_rate": 2.7962085308056874e-06, + "loss": 0.8127, + "step": 59 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 9.806700706481934, + "learning_rate": 2.8436018957345973e-06, + "loss": 0.805, + "step": 60 + }, + { + "epoch": 0.04337777777777778, + "grad_norm": 10.424652099609375, + "learning_rate": 2.8909952606635073e-06, + "loss": 0.7934, + "step": 61 + }, + { + "epoch": 0.044088888888888886, + "grad_norm": 6.070518493652344, + "learning_rate": 2.938388625592417e-06, + "loss": 0.7402, + "step": 62 + }, + { + "epoch": 0.0448, + "grad_norm": 11.389714241027832, + "learning_rate": 2.985781990521327e-06, + "loss": 0.8066, + "step": 63 + }, + { + "epoch": 0.04551111111111111, + "grad_norm": 16.57745933532715, + "learning_rate": 3.0331753554502375e-06, + "loss": 0.749, + "step": 64 + }, + { + "epoch": 0.04622222222222222, + "grad_norm": 6.85017204284668, + "learning_rate": 3.0805687203791474e-06, + "loss": 0.8037, + "step": 65 + }, + { + "epoch": 0.046933333333333334, + "grad_norm": 33.74360275268555, + "learning_rate": 3.1279620853080574e-06, + "loss": 0.8024, + "step": 66 + }, + { + "epoch": 0.04764444444444445, + "grad_norm": 18.29521369934082, + "learning_rate": 3.1753554502369673e-06, + "loss": 0.7485, + "step": 67 + }, + { + "epoch": 0.048355555555555554, + "grad_norm": 23.733505249023438, + "learning_rate": 3.222748815165877e-06, + "loss": 0.7689, + "step": 68 + }, + { + "epoch": 0.04906666666666667, + "grad_norm": 17.502437591552734, + "learning_rate": 3.2701421800947867e-06, + "loss": 0.7761, + "step": 69 + }, + { + "epoch": 0.049777777777777775, + "grad_norm": 11.277215003967285, + "learning_rate": 3.3175355450236967e-06, + "loss": 0.7664, + "step": 70 + }, + { + "epoch": 0.05048888888888889, + "grad_norm": 10.196959495544434, + "learning_rate": 3.3649289099526066e-06, + "loss": 0.7308, + "step": 71 + }, + { + "epoch": 0.0512, + "grad_norm": 8.489730834960938, + "learning_rate": 3.412322274881517e-06, + "loss": 0.7208, + "step": 72 + }, + { + "epoch": 0.05191111111111111, + "grad_norm": 8.371442794799805, + "learning_rate": 3.459715639810427e-06, + "loss": 0.7574, + "step": 73 + }, + { + "epoch": 0.05262222222222222, + "grad_norm": 15.589056015014648, + "learning_rate": 3.507109004739337e-06, + "loss": 0.6975, + "step": 74 + }, + { + "epoch": 0.05333333333333334, + "grad_norm": 13.982439041137695, + "learning_rate": 3.5545023696682468e-06, + "loss": 0.723, + "step": 75 + }, + { + "epoch": 0.054044444444444444, + "grad_norm": 9.172749519348145, + "learning_rate": 3.6018957345971567e-06, + "loss": 0.7323, + "step": 76 + }, + { + "epoch": 0.05475555555555556, + "grad_norm": 8.153841972351074, + "learning_rate": 3.6492890995260666e-06, + "loss": 0.682, + "step": 77 + }, + { + "epoch": 0.055466666666666664, + "grad_norm": 11.31055736541748, + "learning_rate": 3.6966824644549766e-06, + "loss": 0.7168, + "step": 78 + }, + { + "epoch": 0.05617777777777778, + "grad_norm": 5.345701217651367, + "learning_rate": 3.7440758293838865e-06, + "loss": 0.6928, + "step": 79 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 9.976143836975098, + "learning_rate": 3.7914691943127964e-06, + "loss": 0.6706, + "step": 80 + }, + { + "epoch": 0.0576, + "grad_norm": 4.655991554260254, + "learning_rate": 3.838862559241707e-06, + "loss": 0.7054, + "step": 81 + }, + { + "epoch": 0.05831111111111111, + "grad_norm": 10.711212158203125, + "learning_rate": 3.886255924170616e-06, + "loss": 0.6971, + "step": 82 + }, + { + "epoch": 0.05902222222222222, + "grad_norm": 5.5640153884887695, + "learning_rate": 3.933649289099527e-06, + "loss": 0.6633, + "step": 83 + }, + { + "epoch": 0.05973333333333333, + "grad_norm": 5.140533447265625, + "learning_rate": 3.981042654028436e-06, + "loss": 0.6556, + "step": 84 + }, + { + "epoch": 0.060444444444444446, + "grad_norm": 12.882429122924805, + "learning_rate": 4.0284360189573465e-06, + "loss": 0.6253, + "step": 85 + }, + { + "epoch": 0.06115555555555555, + "grad_norm": 4.929982662200928, + "learning_rate": 4.075829383886256e-06, + "loss": 0.667, + "step": 86 + }, + { + "epoch": 0.06186666666666667, + "grad_norm": 9.255619049072266, + "learning_rate": 4.123222748815166e-06, + "loss": 0.6881, + "step": 87 + }, + { + "epoch": 0.06257777777777777, + "grad_norm": 7.011388301849365, + "learning_rate": 4.170616113744076e-06, + "loss": 0.677, + "step": 88 + }, + { + "epoch": 0.0632888888888889, + "grad_norm": 13.98145580291748, + "learning_rate": 4.218009478672986e-06, + "loss": 0.6596, + "step": 89 + }, + { + "epoch": 0.064, + "grad_norm": 11.323680877685547, + "learning_rate": 4.265402843601897e-06, + "loss": 0.649, + "step": 90 + }, + { + "epoch": 0.06471111111111111, + "grad_norm": 9.969672203063965, + "learning_rate": 4.312796208530806e-06, + "loss": 0.6417, + "step": 91 + }, + { + "epoch": 0.06542222222222223, + "grad_norm": 7.269873142242432, + "learning_rate": 4.360189573459716e-06, + "loss": 0.6305, + "step": 92 + }, + { + "epoch": 0.06613333333333334, + "grad_norm": 9.048823356628418, + "learning_rate": 4.407582938388626e-06, + "loss": 0.6253, + "step": 93 + }, + { + "epoch": 0.06684444444444444, + "grad_norm": 8.705512046813965, + "learning_rate": 4.4549763033175355e-06, + "loss": 0.6383, + "step": 94 + }, + { + "epoch": 0.06755555555555555, + "grad_norm": 9.974411010742188, + "learning_rate": 4.502369668246446e-06, + "loss": 0.6086, + "step": 95 + }, + { + "epoch": 0.06826666666666667, + "grad_norm": 7.942433834075928, + "learning_rate": 4.549763033175355e-06, + "loss": 0.5871, + "step": 96 + }, + { + "epoch": 0.06897777777777778, + "grad_norm": 11.440200805664062, + "learning_rate": 4.597156398104266e-06, + "loss": 0.6236, + "step": 97 + }, + { + "epoch": 0.06968888888888888, + "grad_norm": 8.418744087219238, + "learning_rate": 4.644549763033176e-06, + "loss": 0.5911, + "step": 98 + }, + { + "epoch": 0.0704, + "grad_norm": 10.286561965942383, + "learning_rate": 4.691943127962086e-06, + "loss": 0.6297, + "step": 99 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 10.335427284240723, + "learning_rate": 4.739336492890996e-06, + "loss": 0.6222, + "step": 100 + }, + { + "epoch": 0.07182222222222222, + "grad_norm": 6.2219014167785645, + "learning_rate": 4.7867298578199055e-06, + "loss": 0.5676, + "step": 101 + }, + { + "epoch": 0.07253333333333334, + "grad_norm": 5.207602500915527, + "learning_rate": 4.834123222748816e-06, + "loss": 0.6253, + "step": 102 + }, + { + "epoch": 0.07324444444444445, + "grad_norm": 9.122442245483398, + "learning_rate": 4.881516587677725e-06, + "loss": 0.6062, + "step": 103 + }, + { + "epoch": 0.07395555555555555, + "grad_norm": 7.860543251037598, + "learning_rate": 4.928909952606635e-06, + "loss": 0.6314, + "step": 104 + }, + { + "epoch": 0.07466666666666667, + "grad_norm": 7.804473876953125, + "learning_rate": 4.976303317535545e-06, + "loss": 0.6139, + "step": 105 + }, + { + "epoch": 0.07537777777777778, + "grad_norm": 7.323083877563477, + "learning_rate": 5.023696682464455e-06, + "loss": 0.5739, + "step": 106 + }, + { + "epoch": 0.07608888888888889, + "grad_norm": 6.951669216156006, + "learning_rate": 5.071090047393366e-06, + "loss": 0.5727, + "step": 107 + }, + { + "epoch": 0.0768, + "grad_norm": 5.1571736335754395, + "learning_rate": 5.118483412322275e-06, + "loss": 0.5838, + "step": 108 + }, + { + "epoch": 0.07751111111111111, + "grad_norm": 9.363823890686035, + "learning_rate": 5.165876777251185e-06, + "loss": 0.5725, + "step": 109 + }, + { + "epoch": 0.07822222222222222, + "grad_norm": 5.150230407714844, + "learning_rate": 5.213270142180096e-06, + "loss": 0.5534, + "step": 110 + }, + { + "epoch": 0.07893333333333333, + "grad_norm": 6.8626298904418945, + "learning_rate": 5.260663507109005e-06, + "loss": 0.5502, + "step": 111 + }, + { + "epoch": 0.07964444444444445, + "grad_norm": 4.936388969421387, + "learning_rate": 5.308056872037915e-06, + "loss": 0.5954, + "step": 112 + }, + { + "epoch": 0.08035555555555556, + "grad_norm": 7.439772129058838, + "learning_rate": 5.355450236966825e-06, + "loss": 0.5671, + "step": 113 + }, + { + "epoch": 0.08106666666666666, + "grad_norm": 6.3812336921691895, + "learning_rate": 5.402843601895735e-06, + "loss": 0.5995, + "step": 114 + }, + { + "epoch": 0.08177777777777778, + "grad_norm": 3.6189188957214355, + "learning_rate": 5.4502369668246446e-06, + "loss": 0.5321, + "step": 115 + }, + { + "epoch": 0.08248888888888889, + "grad_norm": 13.35216999053955, + "learning_rate": 5.497630331753555e-06, + "loss": 0.6124, + "step": 116 + }, + { + "epoch": 0.0832, + "grad_norm": 6.6187238693237305, + "learning_rate": 5.5450236966824644e-06, + "loss": 0.5544, + "step": 117 + }, + { + "epoch": 0.08391111111111112, + "grad_norm": 9.263525009155273, + "learning_rate": 5.592417061611375e-06, + "loss": 0.5643, + "step": 118 + }, + { + "epoch": 0.08462222222222222, + "grad_norm": 7.294285297393799, + "learning_rate": 5.639810426540285e-06, + "loss": 0.5918, + "step": 119 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 13.273073196411133, + "learning_rate": 5.687203791469195e-06, + "loss": 0.5337, + "step": 120 + }, + { + "epoch": 0.08604444444444445, + "grad_norm": 13.174317359924316, + "learning_rate": 5.734597156398105e-06, + "loss": 0.5808, + "step": 121 + }, + { + "epoch": 0.08675555555555556, + "grad_norm": 6.576754093170166, + "learning_rate": 5.7819905213270145e-06, + "loss": 0.5664, + "step": 122 + }, + { + "epoch": 0.08746666666666666, + "grad_norm": 7.516620635986328, + "learning_rate": 5.829383886255925e-06, + "loss": 0.5582, + "step": 123 + }, + { + "epoch": 0.08817777777777777, + "grad_norm": 7.960477828979492, + "learning_rate": 5.876777251184834e-06, + "loss": 0.553, + "step": 124 + }, + { + "epoch": 0.08888888888888889, + "grad_norm": 5.772348403930664, + "learning_rate": 5.924170616113745e-06, + "loss": 0.531, + "step": 125 + }, + { + "epoch": 0.0896, + "grad_norm": 4.3808979988098145, + "learning_rate": 5.971563981042654e-06, + "loss": 0.5708, + "step": 126 + }, + { + "epoch": 0.0903111111111111, + "grad_norm": 5.318278789520264, + "learning_rate": 6.018957345971565e-06, + "loss": 0.5355, + "step": 127 + }, + { + "epoch": 0.09102222222222223, + "grad_norm": 4.8401312828063965, + "learning_rate": 6.066350710900475e-06, + "loss": 0.4985, + "step": 128 + }, + { + "epoch": 0.09173333333333333, + "grad_norm": 4.3814873695373535, + "learning_rate": 6.1137440758293845e-06, + "loss": 0.5498, + "step": 129 + }, + { + "epoch": 0.09244444444444444, + "grad_norm": 4.231261730194092, + "learning_rate": 6.161137440758295e-06, + "loss": 0.5557, + "step": 130 + }, + { + "epoch": 0.09315555555555556, + "grad_norm": 4.202973365783691, + "learning_rate": 6.208530805687204e-06, + "loss": 0.536, + "step": 131 + }, + { + "epoch": 0.09386666666666667, + "grad_norm": 3.509845495223999, + "learning_rate": 6.255924170616115e-06, + "loss": 0.5408, + "step": 132 + }, + { + "epoch": 0.09457777777777777, + "grad_norm": 4.4882073402404785, + "learning_rate": 6.303317535545023e-06, + "loss": 0.5098, + "step": 133 + }, + { + "epoch": 0.0952888888888889, + "grad_norm": 4.336206912994385, + "learning_rate": 6.350710900473935e-06, + "loss": 0.6021, + "step": 134 + }, + { + "epoch": 0.096, + "grad_norm": 3.6297757625579834, + "learning_rate": 6.398104265402843e-06, + "loss": 0.5099, + "step": 135 + }, + { + "epoch": 0.09671111111111111, + "grad_norm": 5.969592571258545, + "learning_rate": 6.445497630331754e-06, + "loss": 0.5151, + "step": 136 + }, + { + "epoch": 0.09742222222222222, + "grad_norm": 2.9240684509277344, + "learning_rate": 6.492890995260665e-06, + "loss": 0.5377, + "step": 137 + }, + { + "epoch": 0.09813333333333334, + "grad_norm": 7.893035411834717, + "learning_rate": 6.5402843601895735e-06, + "loss": 0.5468, + "step": 138 + }, + { + "epoch": 0.09884444444444444, + "grad_norm": 5.830024242401123, + "learning_rate": 6.587677725118484e-06, + "loss": 0.5751, + "step": 139 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 6.471724510192871, + "learning_rate": 6.635071090047393e-06, + "loss": 0.5898, + "step": 140 + }, + { + "epoch": 0.10026666666666667, + "grad_norm": 4.967070579528809, + "learning_rate": 6.682464454976304e-06, + "loss": 0.5602, + "step": 141 + }, + { + "epoch": 0.10097777777777778, + "grad_norm": 8.279498100280762, + "learning_rate": 6.729857819905213e-06, + "loss": 0.5284, + "step": 142 + }, + { + "epoch": 0.10168888888888888, + "grad_norm": 4.938235759735107, + "learning_rate": 6.777251184834124e-06, + "loss": 0.5978, + "step": 143 + }, + { + "epoch": 0.1024, + "grad_norm": 8.82091236114502, + "learning_rate": 6.824644549763034e-06, + "loss": 0.5212, + "step": 144 + }, + { + "epoch": 0.10311111111111111, + "grad_norm": 8.72518539428711, + "learning_rate": 6.8720379146919435e-06, + "loss": 0.5064, + "step": 145 + }, + { + "epoch": 0.10382222222222222, + "grad_norm": 5.962033748626709, + "learning_rate": 6.919431279620854e-06, + "loss": 0.5165, + "step": 146 + }, + { + "epoch": 0.10453333333333334, + "grad_norm": 6.5562334060668945, + "learning_rate": 6.966824644549763e-06, + "loss": 0.496, + "step": 147 + }, + { + "epoch": 0.10524444444444445, + "grad_norm": 6.653872966766357, + "learning_rate": 7.014218009478674e-06, + "loss": 0.5318, + "step": 148 + }, + { + "epoch": 0.10595555555555555, + "grad_norm": 4.76288366317749, + "learning_rate": 7.061611374407583e-06, + "loss": 0.5482, + "step": 149 + }, + { + "epoch": 0.10666666666666667, + "grad_norm": 7.857875823974609, + "learning_rate": 7.1090047393364935e-06, + "loss": 0.5289, + "step": 150 + }, + { + "epoch": 0.10737777777777778, + "grad_norm": 4.920173168182373, + "learning_rate": 7.156398104265403e-06, + "loss": 0.4855, + "step": 151 + }, + { + "epoch": 0.10808888888888889, + "grad_norm": 7.848324775695801, + "learning_rate": 7.203791469194313e-06, + "loss": 0.5437, + "step": 152 + }, + { + "epoch": 0.1088, + "grad_norm": 6.917397499084473, + "learning_rate": 7.251184834123224e-06, + "loss": 0.5611, + "step": 153 + }, + { + "epoch": 0.10951111111111111, + "grad_norm": 4.4537034034729, + "learning_rate": 7.298578199052133e-06, + "loss": 0.5334, + "step": 154 + }, + { + "epoch": 0.11022222222222222, + "grad_norm": 6.10968542098999, + "learning_rate": 7.345971563981044e-06, + "loss": 0.5255, + "step": 155 + }, + { + "epoch": 0.11093333333333333, + "grad_norm": 4.133383274078369, + "learning_rate": 7.393364928909953e-06, + "loss": 0.5269, + "step": 156 + }, + { + "epoch": 0.11164444444444445, + "grad_norm": 3.7927308082580566, + "learning_rate": 7.4407582938388635e-06, + "loss": 0.5328, + "step": 157 + }, + { + "epoch": 0.11235555555555556, + "grad_norm": 8.12065601348877, + "learning_rate": 7.488151658767773e-06, + "loss": 0.5151, + "step": 158 + }, + { + "epoch": 0.11306666666666666, + "grad_norm": 4.926440238952637, + "learning_rate": 7.535545023696683e-06, + "loss": 0.5301, + "step": 159 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 6.377720355987549, + "learning_rate": 7.582938388625593e-06, + "loss": 0.5198, + "step": 160 + }, + { + "epoch": 0.11448888888888889, + "grad_norm": 4.956860542297363, + "learning_rate": 7.630331753554503e-06, + "loss": 0.5217, + "step": 161 + }, + { + "epoch": 0.1152, + "grad_norm": 3.0651938915252686, + "learning_rate": 7.677725118483414e-06, + "loss": 0.5088, + "step": 162 + }, + { + "epoch": 0.11591111111111112, + "grad_norm": 4.404562473297119, + "learning_rate": 7.725118483412322e-06, + "loss": 0.5246, + "step": 163 + }, + { + "epoch": 0.11662222222222222, + "grad_norm": 5.535040855407715, + "learning_rate": 7.772511848341233e-06, + "loss": 0.5868, + "step": 164 + }, + { + "epoch": 0.11733333333333333, + "grad_norm": 2.9453532695770264, + "learning_rate": 7.819905213270143e-06, + "loss": 0.5308, + "step": 165 + }, + { + "epoch": 0.11804444444444444, + "grad_norm": 7.674310207366943, + "learning_rate": 7.867298578199053e-06, + "loss": 0.56, + "step": 166 + }, + { + "epoch": 0.11875555555555556, + "grad_norm": 6.85412073135376, + "learning_rate": 7.914691943127962e-06, + "loss": 0.4974, + "step": 167 + }, + { + "epoch": 0.11946666666666667, + "grad_norm": 5.095519065856934, + "learning_rate": 7.962085308056872e-06, + "loss": 0.5036, + "step": 168 + }, + { + "epoch": 0.12017777777777777, + "grad_norm": 5.24214506149292, + "learning_rate": 8.009478672985783e-06, + "loss": 0.4878, + "step": 169 + }, + { + "epoch": 0.12088888888888889, + "grad_norm": 4.314838886260986, + "learning_rate": 8.056872037914693e-06, + "loss": 0.4972, + "step": 170 + }, + { + "epoch": 0.1216, + "grad_norm": 5.7503790855407715, + "learning_rate": 8.104265402843603e-06, + "loss": 0.5253, + "step": 171 + }, + { + "epoch": 0.1223111111111111, + "grad_norm": 2.826552152633667, + "learning_rate": 8.151658767772512e-06, + "loss": 0.5053, + "step": 172 + }, + { + "epoch": 0.12302222222222223, + "grad_norm": 6.72640323638916, + "learning_rate": 8.199052132701422e-06, + "loss": 0.4968, + "step": 173 + }, + { + "epoch": 0.12373333333333333, + "grad_norm": 4.787735462188721, + "learning_rate": 8.246445497630333e-06, + "loss": 0.5424, + "step": 174 + }, + { + "epoch": 0.12444444444444444, + "grad_norm": 4.478790760040283, + "learning_rate": 8.293838862559243e-06, + "loss": 0.5014, + "step": 175 + }, + { + "epoch": 0.12515555555555555, + "grad_norm": 3.7871813774108887, + "learning_rate": 8.341232227488152e-06, + "loss": 0.5386, + "step": 176 + }, + { + "epoch": 0.12586666666666665, + "grad_norm": 4.360183238983154, + "learning_rate": 8.388625592417062e-06, + "loss": 0.4934, + "step": 177 + }, + { + "epoch": 0.1265777777777778, + "grad_norm": 4.067202091217041, + "learning_rate": 8.436018957345973e-06, + "loss": 0.4819, + "step": 178 + }, + { + "epoch": 0.1272888888888889, + "grad_norm": 7.427547454833984, + "learning_rate": 8.483412322274883e-06, + "loss": 0.4642, + "step": 179 + }, + { + "epoch": 0.128, + "grad_norm": 5.075253009796143, + "learning_rate": 8.530805687203793e-06, + "loss": 0.5334, + "step": 180 + }, + { + "epoch": 0.1287111111111111, + "grad_norm": 3.8813982009887695, + "learning_rate": 8.578199052132702e-06, + "loss": 0.4935, + "step": 181 + }, + { + "epoch": 0.12942222222222222, + "grad_norm": 6.568417549133301, + "learning_rate": 8.625592417061612e-06, + "loss": 0.536, + "step": 182 + }, + { + "epoch": 0.13013333333333332, + "grad_norm": 2.845282793045044, + "learning_rate": 8.672985781990521e-06, + "loss": 0.4971, + "step": 183 + }, + { + "epoch": 0.13084444444444446, + "grad_norm": 7.286205291748047, + "learning_rate": 8.720379146919431e-06, + "loss": 0.5461, + "step": 184 + }, + { + "epoch": 0.13155555555555556, + "grad_norm": 6.1420087814331055, + "learning_rate": 8.767772511848342e-06, + "loss": 0.4969, + "step": 185 + }, + { + "epoch": 0.13226666666666667, + "grad_norm": 7.751103401184082, + "learning_rate": 8.815165876777252e-06, + "loss": 0.5049, + "step": 186 + }, + { + "epoch": 0.13297777777777778, + "grad_norm": 5.039268970489502, + "learning_rate": 8.862559241706162e-06, + "loss": 0.5297, + "step": 187 + }, + { + "epoch": 0.13368888888888888, + "grad_norm": 6.319336891174316, + "learning_rate": 8.909952606635071e-06, + "loss": 0.5133, + "step": 188 + }, + { + "epoch": 0.1344, + "grad_norm": 4.74379825592041, + "learning_rate": 8.957345971563981e-06, + "loss": 0.4926, + "step": 189 + }, + { + "epoch": 0.1351111111111111, + "grad_norm": 5.595324993133545, + "learning_rate": 9.004739336492892e-06, + "loss": 0.48, + "step": 190 + }, + { + "epoch": 0.13582222222222223, + "grad_norm": 3.853137969970703, + "learning_rate": 9.052132701421802e-06, + "loss": 0.5309, + "step": 191 + }, + { + "epoch": 0.13653333333333334, + "grad_norm": 5.981559753417969, + "learning_rate": 9.09952606635071e-06, + "loss": 0.5023, + "step": 192 + }, + { + "epoch": 0.13724444444444445, + "grad_norm": 5.102012634277344, + "learning_rate": 9.146919431279621e-06, + "loss": 0.5368, + "step": 193 + }, + { + "epoch": 0.13795555555555555, + "grad_norm": 7.003080368041992, + "learning_rate": 9.194312796208532e-06, + "loss": 0.5112, + "step": 194 + }, + { + "epoch": 0.13866666666666666, + "grad_norm": 5.92706823348999, + "learning_rate": 9.241706161137442e-06, + "loss": 0.4578, + "step": 195 + }, + { + "epoch": 0.13937777777777777, + "grad_norm": 6.2215681076049805, + "learning_rate": 9.289099526066352e-06, + "loss": 0.5407, + "step": 196 + }, + { + "epoch": 0.1400888888888889, + "grad_norm": 4.936273574829102, + "learning_rate": 9.336492890995261e-06, + "loss": 0.5564, + "step": 197 + }, + { + "epoch": 0.1408, + "grad_norm": 3.8595407009124756, + "learning_rate": 9.383886255924171e-06, + "loss": 0.4938, + "step": 198 + }, + { + "epoch": 0.14151111111111112, + "grad_norm": 4.272466659545898, + "learning_rate": 9.431279620853082e-06, + "loss": 0.5123, + "step": 199 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 2.8903748989105225, + "learning_rate": 9.478672985781992e-06, + "loss": 0.4639, + "step": 200 + }, + { + "epoch": 0.14293333333333333, + "grad_norm": 4.8672332763671875, + "learning_rate": 9.5260663507109e-06, + "loss": 0.5554, + "step": 201 + }, + { + "epoch": 0.14364444444444444, + "grad_norm": 3.552577257156372, + "learning_rate": 9.573459715639811e-06, + "loss": 0.4967, + "step": 202 + }, + { + "epoch": 0.14435555555555554, + "grad_norm": 4.522824764251709, + "learning_rate": 9.620853080568721e-06, + "loss": 0.4958, + "step": 203 + }, + { + "epoch": 0.14506666666666668, + "grad_norm": 3.9506609439849854, + "learning_rate": 9.668246445497632e-06, + "loss": 0.4993, + "step": 204 + }, + { + "epoch": 0.14577777777777778, + "grad_norm": 4.214148044586182, + "learning_rate": 9.715639810426542e-06, + "loss": 0.5546, + "step": 205 + }, + { + "epoch": 0.1464888888888889, + "grad_norm": 4.218770980834961, + "learning_rate": 9.76303317535545e-06, + "loss": 0.5632, + "step": 206 + }, + { + "epoch": 0.1472, + "grad_norm": 3.12326717376709, + "learning_rate": 9.810426540284361e-06, + "loss": 0.4867, + "step": 207 + }, + { + "epoch": 0.1479111111111111, + "grad_norm": 4.085453033447266, + "learning_rate": 9.85781990521327e-06, + "loss": 0.5307, + "step": 208 + }, + { + "epoch": 0.1486222222222222, + "grad_norm": 2.9664077758789062, + "learning_rate": 9.905213270142182e-06, + "loss": 0.4732, + "step": 209 + }, + { + "epoch": 0.14933333333333335, + "grad_norm": 2.9920878410339355, + "learning_rate": 9.95260663507109e-06, + "loss": 0.5213, + "step": 210 + }, + { + "epoch": 0.15004444444444445, + "grad_norm": 2.7713584899902344, + "learning_rate": 1e-05, + "loss": 0.4891, + "step": 211 + }, + { + "epoch": 0.15075555555555556, + "grad_norm": 2.898153305053711, + "learning_rate": 1.004739336492891e-05, + "loss": 0.4878, + "step": 212 + }, + { + "epoch": 0.15146666666666667, + "grad_norm": 4.835503101348877, + "learning_rate": 1.0094786729857822e-05, + "loss": 0.5536, + "step": 213 + }, + { + "epoch": 0.15217777777777777, + "grad_norm": 2.7085673809051514, + "learning_rate": 1.0142180094786732e-05, + "loss": 0.5312, + "step": 214 + }, + { + "epoch": 0.15288888888888888, + "grad_norm": 3.421114206314087, + "learning_rate": 1.018957345971564e-05, + "loss": 0.5575, + "step": 215 + }, + { + "epoch": 0.1536, + "grad_norm": 3.9022581577301025, + "learning_rate": 1.023696682464455e-05, + "loss": 0.5092, + "step": 216 + }, + { + "epoch": 0.15431111111111112, + "grad_norm": 3.2885355949401855, + "learning_rate": 1.0284360189573461e-05, + "loss": 0.486, + "step": 217 + }, + { + "epoch": 0.15502222222222223, + "grad_norm": 4.1236114501953125, + "learning_rate": 1.033175355450237e-05, + "loss": 0.4703, + "step": 218 + }, + { + "epoch": 0.15573333333333333, + "grad_norm": 3.8225886821746826, + "learning_rate": 1.037914691943128e-05, + "loss": 0.5147, + "step": 219 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 2.5520408153533936, + "learning_rate": 1.0426540284360192e-05, + "loss": 0.4871, + "step": 220 + }, + { + "epoch": 0.15715555555555555, + "grad_norm": 6.766590595245361, + "learning_rate": 1.0473933649289101e-05, + "loss": 0.5114, + "step": 221 + }, + { + "epoch": 0.15786666666666666, + "grad_norm": 4.376341342926025, + "learning_rate": 1.052132701421801e-05, + "loss": 0.5259, + "step": 222 + }, + { + "epoch": 0.1585777777777778, + "grad_norm": 5.930453777313232, + "learning_rate": 1.056872037914692e-05, + "loss": 0.4864, + "step": 223 + }, + { + "epoch": 0.1592888888888889, + "grad_norm": 3.8706393241882324, + "learning_rate": 1.061611374407583e-05, + "loss": 0.5319, + "step": 224 + }, + { + "epoch": 0.16, + "grad_norm": 5.06313943862915, + "learning_rate": 1.066350710900474e-05, + "loss": 0.509, + "step": 225 + }, + { + "epoch": 0.1607111111111111, + "grad_norm": 6.23012638092041, + "learning_rate": 1.071090047393365e-05, + "loss": 0.5068, + "step": 226 + }, + { + "epoch": 0.16142222222222222, + "grad_norm": 4.813814163208008, + "learning_rate": 1.075829383886256e-05, + "loss": 0.5417, + "step": 227 + }, + { + "epoch": 0.16213333333333332, + "grad_norm": 2.9707915782928467, + "learning_rate": 1.080568720379147e-05, + "loss": 0.5393, + "step": 228 + }, + { + "epoch": 0.16284444444444446, + "grad_norm": 4.354058265686035, + "learning_rate": 1.085308056872038e-05, + "loss": 0.4999, + "step": 229 + }, + { + "epoch": 0.16355555555555557, + "grad_norm": 3.935194969177246, + "learning_rate": 1.0900473933649289e-05, + "loss": 0.5337, + "step": 230 + }, + { + "epoch": 0.16426666666666667, + "grad_norm": 4.134055137634277, + "learning_rate": 1.0947867298578201e-05, + "loss": 0.5043, + "step": 231 + }, + { + "epoch": 0.16497777777777778, + "grad_norm": 4.548140048980713, + "learning_rate": 1.099526066350711e-05, + "loss": 0.5374, + "step": 232 + }, + { + "epoch": 0.16568888888888889, + "grad_norm": 3.072282314300537, + "learning_rate": 1.104265402843602e-05, + "loss": 0.5632, + "step": 233 + }, + { + "epoch": 0.1664, + "grad_norm": 6.970465660095215, + "learning_rate": 1.1090047393364929e-05, + "loss": 0.5454, + "step": 234 + }, + { + "epoch": 0.1671111111111111, + "grad_norm": 7.205247402191162, + "learning_rate": 1.1137440758293841e-05, + "loss": 0.5304, + "step": 235 + }, + { + "epoch": 0.16782222222222223, + "grad_norm": 4.45557165145874, + "learning_rate": 1.118483412322275e-05, + "loss": 0.5205, + "step": 236 + }, + { + "epoch": 0.16853333333333334, + "grad_norm": 3.764113426208496, + "learning_rate": 1.1232227488151658e-05, + "loss": 0.5115, + "step": 237 + }, + { + "epoch": 0.16924444444444445, + "grad_norm": 5.614161014556885, + "learning_rate": 1.127962085308057e-05, + "loss": 0.4851, + "step": 238 + }, + { + "epoch": 0.16995555555555555, + "grad_norm": 3.8806092739105225, + "learning_rate": 1.132701421800948e-05, + "loss": 0.4576, + "step": 239 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 7.451761245727539, + "learning_rate": 1.137440758293839e-05, + "loss": 0.5021, + "step": 240 + }, + { + "epoch": 0.17137777777777777, + "grad_norm": 6.800906658172607, + "learning_rate": 1.1421800947867298e-05, + "loss": 0.4962, + "step": 241 + }, + { + "epoch": 0.1720888888888889, + "grad_norm": 2.5431292057037354, + "learning_rate": 1.146919431279621e-05, + "loss": 0.4419, + "step": 242 + }, + { + "epoch": 0.1728, + "grad_norm": 7.403611660003662, + "learning_rate": 1.1516587677725119e-05, + "loss": 0.5122, + "step": 243 + }, + { + "epoch": 0.17351111111111112, + "grad_norm": 6.277567386627197, + "learning_rate": 1.1563981042654029e-05, + "loss": 0.547, + "step": 244 + }, + { + "epoch": 0.17422222222222222, + "grad_norm": 4.7829670906066895, + "learning_rate": 1.1611374407582941e-05, + "loss": 0.495, + "step": 245 + }, + { + "epoch": 0.17493333333333333, + "grad_norm": 4.416293144226074, + "learning_rate": 1.165876777251185e-05, + "loss": 0.5184, + "step": 246 + }, + { + "epoch": 0.17564444444444444, + "grad_norm": 4.348495006561279, + "learning_rate": 1.1706161137440758e-05, + "loss": 0.4863, + "step": 247 + }, + { + "epoch": 0.17635555555555554, + "grad_norm": 3.891252040863037, + "learning_rate": 1.1753554502369669e-05, + "loss": 0.5238, + "step": 248 + }, + { + "epoch": 0.17706666666666668, + "grad_norm": 4.528507709503174, + "learning_rate": 1.180094786729858e-05, + "loss": 0.4867, + "step": 249 + }, + { + "epoch": 0.17777777777777778, + "grad_norm": 4.057559013366699, + "learning_rate": 1.184834123222749e-05, + "loss": 0.5191, + "step": 250 + }, + { + "epoch": 0.1784888888888889, + "grad_norm": 5.301110744476318, + "learning_rate": 1.1895734597156398e-05, + "loss": 0.4933, + "step": 251 + }, + { + "epoch": 0.1792, + "grad_norm": 3.9183449745178223, + "learning_rate": 1.1943127962085309e-05, + "loss": 0.488, + "step": 252 + }, + { + "epoch": 0.1799111111111111, + "grad_norm": 5.223655700683594, + "learning_rate": 1.1990521327014219e-05, + "loss": 0.5429, + "step": 253 + }, + { + "epoch": 0.1806222222222222, + "grad_norm": 4.26966667175293, + "learning_rate": 1.203791469194313e-05, + "loss": 0.5272, + "step": 254 + }, + { + "epoch": 0.18133333333333335, + "grad_norm": 5.591842174530029, + "learning_rate": 1.2085308056872038e-05, + "loss": 0.4939, + "step": 255 + }, + { + "epoch": 0.18204444444444445, + "grad_norm": 5.419987201690674, + "learning_rate": 1.213270142180095e-05, + "loss": 0.4952, + "step": 256 + }, + { + "epoch": 0.18275555555555556, + "grad_norm": 4.122593879699707, + "learning_rate": 1.2180094786729859e-05, + "loss": 0.5105, + "step": 257 + }, + { + "epoch": 0.18346666666666667, + "grad_norm": 3.3791682720184326, + "learning_rate": 1.2227488151658769e-05, + "loss": 0.462, + "step": 258 + }, + { + "epoch": 0.18417777777777777, + "grad_norm": 5.889245986938477, + "learning_rate": 1.2274881516587678e-05, + "loss": 0.4769, + "step": 259 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 4.383092403411865, + "learning_rate": 1.232227488151659e-05, + "loss": 0.5093, + "step": 260 + }, + { + "epoch": 0.1856, + "grad_norm": 4.57335901260376, + "learning_rate": 1.2369668246445498e-05, + "loss": 0.4915, + "step": 261 + }, + { + "epoch": 0.18631111111111112, + "grad_norm": 4.621091365814209, + "learning_rate": 1.2417061611374409e-05, + "loss": 0.4719, + "step": 262 + }, + { + "epoch": 0.18702222222222223, + "grad_norm": 3.4973011016845703, + "learning_rate": 1.2464454976303319e-05, + "loss": 0.456, + "step": 263 + }, + { + "epoch": 0.18773333333333334, + "grad_norm": 4.299407482147217, + "learning_rate": 1.251184834123223e-05, + "loss": 0.4957, + "step": 264 + }, + { + "epoch": 0.18844444444444444, + "grad_norm": 4.250063419342041, + "learning_rate": 1.2559241706161138e-05, + "loss": 0.5123, + "step": 265 + }, + { + "epoch": 0.18915555555555555, + "grad_norm": 3.087158203125, + "learning_rate": 1.2606635071090047e-05, + "loss": 0.5205, + "step": 266 + }, + { + "epoch": 0.18986666666666666, + "grad_norm": 6.040894985198975, + "learning_rate": 1.2654028436018959e-05, + "loss": 0.4493, + "step": 267 + }, + { + "epoch": 0.1905777777777778, + "grad_norm": 5.1678242683410645, + "learning_rate": 1.270142180094787e-05, + "loss": 0.55, + "step": 268 + }, + { + "epoch": 0.1912888888888889, + "grad_norm": 4.0269293785095215, + "learning_rate": 1.2748815165876778e-05, + "loss": 0.486, + "step": 269 + }, + { + "epoch": 0.192, + "grad_norm": 4.772781848907471, + "learning_rate": 1.2796208530805687e-05, + "loss": 0.514, + "step": 270 + }, + { + "epoch": 0.1927111111111111, + "grad_norm": 3.316976547241211, + "learning_rate": 1.2843601895734599e-05, + "loss": 0.4632, + "step": 271 + }, + { + "epoch": 0.19342222222222222, + "grad_norm": 3.822645902633667, + "learning_rate": 1.2890995260663507e-05, + "loss": 0.4826, + "step": 272 + }, + { + "epoch": 0.19413333333333332, + "grad_norm": 3.5536272525787354, + "learning_rate": 1.2938388625592418e-05, + "loss": 0.4998, + "step": 273 + }, + { + "epoch": 0.19484444444444443, + "grad_norm": 3.0050196647644043, + "learning_rate": 1.298578199052133e-05, + "loss": 0.4761, + "step": 274 + }, + { + "epoch": 0.19555555555555557, + "grad_norm": 5.281590938568115, + "learning_rate": 1.3033175355450238e-05, + "loss": 0.4655, + "step": 275 + }, + { + "epoch": 0.19626666666666667, + "grad_norm": 3.89361834526062, + "learning_rate": 1.3080568720379147e-05, + "loss": 0.5219, + "step": 276 + }, + { + "epoch": 0.19697777777777778, + "grad_norm": 4.476886749267578, + "learning_rate": 1.3127962085308057e-05, + "loss": 0.527, + "step": 277 + }, + { + "epoch": 0.1976888888888889, + "grad_norm": 3.896406888961792, + "learning_rate": 1.3175355450236968e-05, + "loss": 0.4425, + "step": 278 + }, + { + "epoch": 0.1984, + "grad_norm": 3.3174450397491455, + "learning_rate": 1.3222748815165878e-05, + "loss": 0.5191, + "step": 279 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 3.872722625732422, + "learning_rate": 1.3270142180094787e-05, + "loss": 0.5188, + "step": 280 + }, + { + "epoch": 0.19982222222222223, + "grad_norm": 4.094091415405273, + "learning_rate": 1.3317535545023699e-05, + "loss": 0.5166, + "step": 281 + }, + { + "epoch": 0.20053333333333334, + "grad_norm": 3.5000228881835938, + "learning_rate": 1.3364928909952607e-05, + "loss": 0.4823, + "step": 282 + }, + { + "epoch": 0.20124444444444445, + "grad_norm": 3.134237766265869, + "learning_rate": 1.3412322274881518e-05, + "loss": 0.4977, + "step": 283 + }, + { + "epoch": 0.20195555555555555, + "grad_norm": 2.0882067680358887, + "learning_rate": 1.3459715639810426e-05, + "loss": 0.4767, + "step": 284 + }, + { + "epoch": 0.20266666666666666, + "grad_norm": 4.1349358558654785, + "learning_rate": 1.3507109004739339e-05, + "loss": 0.4695, + "step": 285 + }, + { + "epoch": 0.20337777777777777, + "grad_norm": 2.7574007511138916, + "learning_rate": 1.3554502369668247e-05, + "loss": 0.5008, + "step": 286 + }, + { + "epoch": 0.20408888888888888, + "grad_norm": 4.918444633483887, + "learning_rate": 1.3601895734597158e-05, + "loss": 0.4797, + "step": 287 + }, + { + "epoch": 0.2048, + "grad_norm": 4.640432834625244, + "learning_rate": 1.3649289099526068e-05, + "loss": 0.5149, + "step": 288 + }, + { + "epoch": 0.20551111111111112, + "grad_norm": 3.4757654666900635, + "learning_rate": 1.3696682464454978e-05, + "loss": 0.545, + "step": 289 + }, + { + "epoch": 0.20622222222222222, + "grad_norm": 3.6345794200897217, + "learning_rate": 1.3744075829383887e-05, + "loss": 0.4681, + "step": 290 + }, + { + "epoch": 0.20693333333333333, + "grad_norm": 3.446122407913208, + "learning_rate": 1.3791469194312797e-05, + "loss": 0.4661, + "step": 291 + }, + { + "epoch": 0.20764444444444444, + "grad_norm": 3.1782402992248535, + "learning_rate": 1.3838862559241708e-05, + "loss": 0.5071, + "step": 292 + }, + { + "epoch": 0.20835555555555554, + "grad_norm": 2.8736047744750977, + "learning_rate": 1.3886255924170618e-05, + "loss": 0.5215, + "step": 293 + }, + { + "epoch": 0.20906666666666668, + "grad_norm": 2.925722122192383, + "learning_rate": 1.3933649289099527e-05, + "loss": 0.488, + "step": 294 + }, + { + "epoch": 0.20977777777777779, + "grad_norm": 3.132317543029785, + "learning_rate": 1.3981042654028435e-05, + "loss": 0.4938, + "step": 295 + }, + { + "epoch": 0.2104888888888889, + "grad_norm": 3.099607467651367, + "learning_rate": 1.4028436018957347e-05, + "loss": 0.4483, + "step": 296 + }, + { + "epoch": 0.2112, + "grad_norm": 3.9765737056732178, + "learning_rate": 1.4075829383886258e-05, + "loss": 0.4907, + "step": 297 + }, + { + "epoch": 0.2119111111111111, + "grad_norm": 3.711031675338745, + "learning_rate": 1.4123222748815166e-05, + "loss": 0.5083, + "step": 298 + }, + { + "epoch": 0.2126222222222222, + "grad_norm": 5.24534797668457, + "learning_rate": 1.4170616113744078e-05, + "loss": 0.4602, + "step": 299 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 5.1365485191345215, + "learning_rate": 1.4218009478672987e-05, + "loss": 0.4968, + "step": 300 + }, + { + "epoch": 0.21404444444444445, + "grad_norm": 4.154321670532227, + "learning_rate": 1.4265402843601896e-05, + "loss": 0.491, + "step": 301 + }, + { + "epoch": 0.21475555555555556, + "grad_norm": 3.7960476875305176, + "learning_rate": 1.4312796208530806e-05, + "loss": 0.5134, + "step": 302 + }, + { + "epoch": 0.21546666666666667, + "grad_norm": 4.072483539581299, + "learning_rate": 1.4360189573459718e-05, + "loss": 0.4909, + "step": 303 + }, + { + "epoch": 0.21617777777777777, + "grad_norm": 2.84600567817688, + "learning_rate": 1.4407582938388627e-05, + "loss": 0.5059, + "step": 304 + }, + { + "epoch": 0.21688888888888888, + "grad_norm": 3.9269962310791016, + "learning_rate": 1.4454976303317535e-05, + "loss": 0.4587, + "step": 305 + }, + { + "epoch": 0.2176, + "grad_norm": 3.4242045879364014, + "learning_rate": 1.4502369668246448e-05, + "loss": 0.4984, + "step": 306 + }, + { + "epoch": 0.21831111111111112, + "grad_norm": 4.920367240905762, + "learning_rate": 1.4549763033175356e-05, + "loss": 0.5181, + "step": 307 + }, + { + "epoch": 0.21902222222222223, + "grad_norm": 5.413025856018066, + "learning_rate": 1.4597156398104267e-05, + "loss": 0.5046, + "step": 308 + }, + { + "epoch": 0.21973333333333334, + "grad_norm": 3.142439365386963, + "learning_rate": 1.4644549763033175e-05, + "loss": 0.5409, + "step": 309 + }, + { + "epoch": 0.22044444444444444, + "grad_norm": 3.2751197814941406, + "learning_rate": 1.4691943127962087e-05, + "loss": 0.5042, + "step": 310 + }, + { + "epoch": 0.22115555555555555, + "grad_norm": 3.547229051589966, + "learning_rate": 1.4739336492890996e-05, + "loss": 0.4818, + "step": 311 + }, + { + "epoch": 0.22186666666666666, + "grad_norm": 2.253777503967285, + "learning_rate": 1.4786729857819906e-05, + "loss": 0.5111, + "step": 312 + }, + { + "epoch": 0.2225777777777778, + "grad_norm": 3.496878147125244, + "learning_rate": 1.4834123222748817e-05, + "loss": 0.4711, + "step": 313 + }, + { + "epoch": 0.2232888888888889, + "grad_norm": 2.3660967350006104, + "learning_rate": 1.4881516587677727e-05, + "loss": 0.4759, + "step": 314 + }, + { + "epoch": 0.224, + "grad_norm": 4.4830546379089355, + "learning_rate": 1.4928909952606636e-05, + "loss": 0.5417, + "step": 315 + }, + { + "epoch": 0.2247111111111111, + "grad_norm": 3.642277956008911, + "learning_rate": 1.4976303317535546e-05, + "loss": 0.4919, + "step": 316 + }, + { + "epoch": 0.22542222222222222, + "grad_norm": 3.0057177543640137, + "learning_rate": 1.5023696682464456e-05, + "loss": 0.5031, + "step": 317 + }, + { + "epoch": 0.22613333333333333, + "grad_norm": 3.1279497146606445, + "learning_rate": 1.5071090047393367e-05, + "loss": 0.4864, + "step": 318 + }, + { + "epoch": 0.22684444444444443, + "grad_norm": 3.451179027557373, + "learning_rate": 1.5118483412322275e-05, + "loss": 0.4609, + "step": 319 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 2.4904258251190186, + "learning_rate": 1.5165876777251186e-05, + "loss": 0.4846, + "step": 320 + }, + { + "epoch": 0.22826666666666667, + "grad_norm": 3.2437591552734375, + "learning_rate": 1.5213270142180096e-05, + "loss": 0.5086, + "step": 321 + }, + { + "epoch": 0.22897777777777778, + "grad_norm": 2.530247926712036, + "learning_rate": 1.5260663507109007e-05, + "loss": 0.4512, + "step": 322 + }, + { + "epoch": 0.2296888888888889, + "grad_norm": 2.9988605976104736, + "learning_rate": 1.5308056872037915e-05, + "loss": 0.4714, + "step": 323 + }, + { + "epoch": 0.2304, + "grad_norm": 2.5761773586273193, + "learning_rate": 1.5355450236966827e-05, + "loss": 0.532, + "step": 324 + }, + { + "epoch": 0.2311111111111111, + "grad_norm": 3.54913592338562, + "learning_rate": 1.5402843601895736e-05, + "loss": 0.5416, + "step": 325 + }, + { + "epoch": 0.23182222222222224, + "grad_norm": 2.688767194747925, + "learning_rate": 1.5450236966824645e-05, + "loss": 0.5213, + "step": 326 + }, + { + "epoch": 0.23253333333333334, + "grad_norm": 4.006014347076416, + "learning_rate": 1.5497630331753553e-05, + "loss": 0.4729, + "step": 327 + }, + { + "epoch": 0.23324444444444445, + "grad_norm": 3.4370555877685547, + "learning_rate": 1.5545023696682465e-05, + "loss": 0.4722, + "step": 328 + }, + { + "epoch": 0.23395555555555556, + "grad_norm": 3.203117847442627, + "learning_rate": 1.5592417061611377e-05, + "loss": 0.478, + "step": 329 + }, + { + "epoch": 0.23466666666666666, + "grad_norm": 3.6933863162994385, + "learning_rate": 1.5639810426540286e-05, + "loss": 0.5272, + "step": 330 + }, + { + "epoch": 0.23537777777777777, + "grad_norm": 2.7507715225219727, + "learning_rate": 1.5687203791469198e-05, + "loss": 0.5084, + "step": 331 + }, + { + "epoch": 0.23608888888888888, + "grad_norm": 2.888516664505005, + "learning_rate": 1.5734597156398107e-05, + "loss": 0.4913, + "step": 332 + }, + { + "epoch": 0.2368, + "grad_norm": 2.248821496963501, + "learning_rate": 1.5781990521327015e-05, + "loss": 0.5013, + "step": 333 + }, + { + "epoch": 0.23751111111111112, + "grad_norm": 4.56018590927124, + "learning_rate": 1.5829383886255924e-05, + "loss": 0.4976, + "step": 334 + }, + { + "epoch": 0.23822222222222222, + "grad_norm": 3.273653745651245, + "learning_rate": 1.5876777251184836e-05, + "loss": 0.5077, + "step": 335 + }, + { + "epoch": 0.23893333333333333, + "grad_norm": 5.10419225692749, + "learning_rate": 1.5924170616113745e-05, + "loss": 0.5007, + "step": 336 + }, + { + "epoch": 0.23964444444444444, + "grad_norm": 3.2050716876983643, + "learning_rate": 1.5971563981042653e-05, + "loss": 0.5049, + "step": 337 + }, + { + "epoch": 0.24035555555555554, + "grad_norm": 3.639308214187622, + "learning_rate": 1.6018957345971565e-05, + "loss": 0.4926, + "step": 338 + }, + { + "epoch": 0.24106666666666668, + "grad_norm": 2.6481032371520996, + "learning_rate": 1.6066350710900474e-05, + "loss": 0.499, + "step": 339 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 3.020235300064087, + "learning_rate": 1.6113744075829386e-05, + "loss": 0.4376, + "step": 340 + }, + { + "epoch": 0.2424888888888889, + "grad_norm": 2.5266494750976562, + "learning_rate": 1.6161137440758295e-05, + "loss": 0.4629, + "step": 341 + }, + { + "epoch": 0.2432, + "grad_norm": 2.8628244400024414, + "learning_rate": 1.6208530805687207e-05, + "loss": 0.4802, + "step": 342 + }, + { + "epoch": 0.2439111111111111, + "grad_norm": 2.955493688583374, + "learning_rate": 1.6255924170616116e-05, + "loss": 0.4644, + "step": 343 + }, + { + "epoch": 0.2446222222222222, + "grad_norm": 3.423654079437256, + "learning_rate": 1.6303317535545024e-05, + "loss": 0.4798, + "step": 344 + }, + { + "epoch": 0.24533333333333332, + "grad_norm": 2.424535036087036, + "learning_rate": 1.6350710900473933e-05, + "loss": 0.5061, + "step": 345 + }, + { + "epoch": 0.24604444444444445, + "grad_norm": 3.7033872604370117, + "learning_rate": 1.6398104265402845e-05, + "loss": 0.5026, + "step": 346 + }, + { + "epoch": 0.24675555555555556, + "grad_norm": 2.1360037326812744, + "learning_rate": 1.6445497630331754e-05, + "loss": 0.4891, + "step": 347 + }, + { + "epoch": 0.24746666666666667, + "grad_norm": 2.835291862487793, + "learning_rate": 1.6492890995260666e-05, + "loss": 0.5253, + "step": 348 + }, + { + "epoch": 0.24817777777777777, + "grad_norm": 2.873575210571289, + "learning_rate": 1.6540284360189574e-05, + "loss": 0.4906, + "step": 349 + }, + { + "epoch": 0.24888888888888888, + "grad_norm": 2.3434009552001953, + "learning_rate": 1.6587677725118486e-05, + "loss": 0.4996, + "step": 350 + }, + { + "epoch": 0.2496, + "grad_norm": 2.464038610458374, + "learning_rate": 1.6635071090047395e-05, + "loss": 0.5281, + "step": 351 + }, + { + "epoch": 0.2503111111111111, + "grad_norm": 2.9241421222686768, + "learning_rate": 1.6682464454976304e-05, + "loss": 0.5086, + "step": 352 + }, + { + "epoch": 0.2510222222222222, + "grad_norm": 3.2049553394317627, + "learning_rate": 1.6729857819905216e-05, + "loss": 0.4413, + "step": 353 + }, + { + "epoch": 0.2517333333333333, + "grad_norm": 2.213411808013916, + "learning_rate": 1.6777251184834124e-05, + "loss": 0.4911, + "step": 354 + }, + { + "epoch": 0.25244444444444447, + "grad_norm": 3.3933403491973877, + "learning_rate": 1.6824644549763033e-05, + "loss": 0.5, + "step": 355 + }, + { + "epoch": 0.2531555555555556, + "grad_norm": 2.119575023651123, + "learning_rate": 1.6872037914691945e-05, + "loss": 0.5175, + "step": 356 + }, + { + "epoch": 0.2538666666666667, + "grad_norm": 2.9153764247894287, + "learning_rate": 1.6919431279620854e-05, + "loss": 0.5086, + "step": 357 + }, + { + "epoch": 0.2545777777777778, + "grad_norm": 1.5884883403778076, + "learning_rate": 1.6966824644549766e-05, + "loss": 0.4993, + "step": 358 + }, + { + "epoch": 0.2552888888888889, + "grad_norm": 3.5453898906707764, + "learning_rate": 1.7014218009478674e-05, + "loss": 0.4999, + "step": 359 + }, + { + "epoch": 0.256, + "grad_norm": 2.350055456161499, + "learning_rate": 1.7061611374407587e-05, + "loss": 0.4847, + "step": 360 + }, + { + "epoch": 0.2567111111111111, + "grad_norm": 5.420952320098877, + "learning_rate": 1.7109004739336495e-05, + "loss": 0.5169, + "step": 361 + }, + { + "epoch": 0.2574222222222222, + "grad_norm": 4.110228538513184, + "learning_rate": 1.7156398104265404e-05, + "loss": 0.5168, + "step": 362 + }, + { + "epoch": 0.2581333333333333, + "grad_norm": 2.933438301086426, + "learning_rate": 1.7203791469194316e-05, + "loss": 0.4796, + "step": 363 + }, + { + "epoch": 0.25884444444444443, + "grad_norm": 3.365131378173828, + "learning_rate": 1.7251184834123225e-05, + "loss": 0.5612, + "step": 364 + }, + { + "epoch": 0.25955555555555554, + "grad_norm": 2.980142116546631, + "learning_rate": 1.7298578199052133e-05, + "loss": 0.5585, + "step": 365 + }, + { + "epoch": 0.26026666666666665, + "grad_norm": 3.9425296783447266, + "learning_rate": 1.7345971563981042e-05, + "loss": 0.4856, + "step": 366 + }, + { + "epoch": 0.26097777777777775, + "grad_norm": 3.4514007568359375, + "learning_rate": 1.7393364928909954e-05, + "loss": 0.4774, + "step": 367 + }, + { + "epoch": 0.2616888888888889, + "grad_norm": 2.103703260421753, + "learning_rate": 1.7440758293838863e-05, + "loss": 0.4751, + "step": 368 + }, + { + "epoch": 0.2624, + "grad_norm": 3.1878414154052734, + "learning_rate": 1.7488151658767775e-05, + "loss": 0.532, + "step": 369 + }, + { + "epoch": 0.26311111111111113, + "grad_norm": 3.8282370567321777, + "learning_rate": 1.7535545023696683e-05, + "loss": 0.5511, + "step": 370 + }, + { + "epoch": 0.26382222222222224, + "grad_norm": 2.920881986618042, + "learning_rate": 1.7582938388625595e-05, + "loss": 0.488, + "step": 371 + }, + { + "epoch": 0.26453333333333334, + "grad_norm": 2.713918685913086, + "learning_rate": 1.7630331753554504e-05, + "loss": 0.5888, + "step": 372 + }, + { + "epoch": 0.26524444444444445, + "grad_norm": 2.43331241607666, + "learning_rate": 1.7677725118483413e-05, + "loss": 0.4798, + "step": 373 + }, + { + "epoch": 0.26595555555555556, + "grad_norm": 2.925508499145508, + "learning_rate": 1.7725118483412325e-05, + "loss": 0.4866, + "step": 374 + }, + { + "epoch": 0.26666666666666666, + "grad_norm": 2.2425506114959717, + "learning_rate": 1.7772511848341233e-05, + "loss": 0.5016, + "step": 375 + }, + { + "epoch": 0.26737777777777777, + "grad_norm": 2.517987012863159, + "learning_rate": 1.7819905213270142e-05, + "loss": 0.5051, + "step": 376 + }, + { + "epoch": 0.2680888888888889, + "grad_norm": 2.347808361053467, + "learning_rate": 1.7867298578199054e-05, + "loss": 0.4542, + "step": 377 + }, + { + "epoch": 0.2688, + "grad_norm": 2.208350658416748, + "learning_rate": 1.7914691943127963e-05, + "loss": 0.4637, + "step": 378 + }, + { + "epoch": 0.2695111111111111, + "grad_norm": 1.702507495880127, + "learning_rate": 1.7962085308056875e-05, + "loss": 0.4637, + "step": 379 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 2.2131340503692627, + "learning_rate": 1.8009478672985784e-05, + "loss": 0.485, + "step": 380 + }, + { + "epoch": 0.27093333333333336, + "grad_norm": 5.122796535491943, + "learning_rate": 1.8056872037914696e-05, + "loss": 0.5032, + "step": 381 + }, + { + "epoch": 0.27164444444444447, + "grad_norm": 2.579911708831787, + "learning_rate": 1.8104265402843604e-05, + "loss": 0.487, + "step": 382 + }, + { + "epoch": 0.2723555555555556, + "grad_norm": 4.667227745056152, + "learning_rate": 1.8151658767772513e-05, + "loss": 0.4654, + "step": 383 + }, + { + "epoch": 0.2730666666666667, + "grad_norm": 3.9245495796203613, + "learning_rate": 1.819905213270142e-05, + "loss": 0.4931, + "step": 384 + }, + { + "epoch": 0.2737777777777778, + "grad_norm": 3.613002061843872, + "learning_rate": 1.8246445497630334e-05, + "loss": 0.5042, + "step": 385 + }, + { + "epoch": 0.2744888888888889, + "grad_norm": 3.349091053009033, + "learning_rate": 1.8293838862559242e-05, + "loss": 0.5031, + "step": 386 + }, + { + "epoch": 0.2752, + "grad_norm": 2.8045475482940674, + "learning_rate": 1.8341232227488154e-05, + "loss": 0.497, + "step": 387 + }, + { + "epoch": 0.2759111111111111, + "grad_norm": 2.3788719177246094, + "learning_rate": 1.8388625592417063e-05, + "loss": 0.4717, + "step": 388 + }, + { + "epoch": 0.2766222222222222, + "grad_norm": 3.9690022468566895, + "learning_rate": 1.8436018957345975e-05, + "loss": 0.5356, + "step": 389 + }, + { + "epoch": 0.2773333333333333, + "grad_norm": 2.6112773418426514, + "learning_rate": 1.8483412322274884e-05, + "loss": 0.4859, + "step": 390 + }, + { + "epoch": 0.2780444444444444, + "grad_norm": 4.7624945640563965, + "learning_rate": 1.8530805687203792e-05, + "loss": 0.5055, + "step": 391 + }, + { + "epoch": 0.27875555555555553, + "grad_norm": 4.399750232696533, + "learning_rate": 1.8578199052132704e-05, + "loss": 0.5128, + "step": 392 + }, + { + "epoch": 0.27946666666666664, + "grad_norm": 2.4322690963745117, + "learning_rate": 1.8625592417061613e-05, + "loss": 0.5018, + "step": 393 + }, + { + "epoch": 0.2801777777777778, + "grad_norm": 4.3571038246154785, + "learning_rate": 1.8672985781990522e-05, + "loss": 0.5154, + "step": 394 + }, + { + "epoch": 0.2808888888888889, + "grad_norm": 2.6631150245666504, + "learning_rate": 1.872037914691943e-05, + "loss": 0.4854, + "step": 395 + }, + { + "epoch": 0.2816, + "grad_norm": 3.9459939002990723, + "learning_rate": 1.8767772511848342e-05, + "loss": 0.5204, + "step": 396 + }, + { + "epoch": 0.2823111111111111, + "grad_norm": 3.359722375869751, + "learning_rate": 1.881516587677725e-05, + "loss": 0.4653, + "step": 397 + }, + { + "epoch": 0.28302222222222223, + "grad_norm": 3.111274242401123, + "learning_rate": 1.8862559241706163e-05, + "loss": 0.5275, + "step": 398 + }, + { + "epoch": 0.28373333333333334, + "grad_norm": 3.5329575538635254, + "learning_rate": 1.8909952606635075e-05, + "loss": 0.4806, + "step": 399 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 2.1247153282165527, + "learning_rate": 1.8957345971563984e-05, + "loss": 0.5076, + "step": 400 + }, + { + "epoch": 0.28515555555555555, + "grad_norm": 2.3776113986968994, + "learning_rate": 1.9004739336492893e-05, + "loss": 0.4928, + "step": 401 + }, + { + "epoch": 0.28586666666666666, + "grad_norm": 3.134005308151245, + "learning_rate": 1.90521327014218e-05, + "loss": 0.4955, + "step": 402 + }, + { + "epoch": 0.28657777777777776, + "grad_norm": 2.3699586391448975, + "learning_rate": 1.9099526066350713e-05, + "loss": 0.4438, + "step": 403 + }, + { + "epoch": 0.28728888888888887, + "grad_norm": 3.257565975189209, + "learning_rate": 1.9146919431279622e-05, + "loss": 0.5048, + "step": 404 + }, + { + "epoch": 0.288, + "grad_norm": 2.966352939605713, + "learning_rate": 1.919431279620853e-05, + "loss": 0.4887, + "step": 405 + }, + { + "epoch": 0.2887111111111111, + "grad_norm": 2.792395830154419, + "learning_rate": 1.9241706161137443e-05, + "loss": 0.5, + "step": 406 + }, + { + "epoch": 0.28942222222222225, + "grad_norm": 2.134561061859131, + "learning_rate": 1.928909952606635e-05, + "loss": 0.5108, + "step": 407 + }, + { + "epoch": 0.29013333333333335, + "grad_norm": 3.1566803455352783, + "learning_rate": 1.9336492890995263e-05, + "loss": 0.4728, + "step": 408 + }, + { + "epoch": 0.29084444444444446, + "grad_norm": 2.287613868713379, + "learning_rate": 1.9383886255924172e-05, + "loss": 0.4972, + "step": 409 + }, + { + "epoch": 0.29155555555555557, + "grad_norm": 3.161025047302246, + "learning_rate": 1.9431279620853084e-05, + "loss": 0.495, + "step": 410 + }, + { + "epoch": 0.2922666666666667, + "grad_norm": 2.428663730621338, + "learning_rate": 1.9478672985781993e-05, + "loss": 0.4935, + "step": 411 + }, + { + "epoch": 0.2929777777777778, + "grad_norm": 3.6074275970458984, + "learning_rate": 1.95260663507109e-05, + "loss": 0.5075, + "step": 412 + }, + { + "epoch": 0.2936888888888889, + "grad_norm": 3.1715080738067627, + "learning_rate": 1.957345971563981e-05, + "loss": 0.5218, + "step": 413 + }, + { + "epoch": 0.2944, + "grad_norm": 1.7597421407699585, + "learning_rate": 1.9620853080568722e-05, + "loss": 0.4943, + "step": 414 + }, + { + "epoch": 0.2951111111111111, + "grad_norm": 2.357790946960449, + "learning_rate": 1.966824644549763e-05, + "loss": 0.5264, + "step": 415 + }, + { + "epoch": 0.2958222222222222, + "grad_norm": 1.7636620998382568, + "learning_rate": 1.971563981042654e-05, + "loss": 0.4672, + "step": 416 + }, + { + "epoch": 0.2965333333333333, + "grad_norm": 2.314072370529175, + "learning_rate": 1.976303317535545e-05, + "loss": 0.4897, + "step": 417 + }, + { + "epoch": 0.2972444444444444, + "grad_norm": 2.2929835319519043, + "learning_rate": 1.9810426540284364e-05, + "loss": 0.4858, + "step": 418 + }, + { + "epoch": 0.29795555555555553, + "grad_norm": 2.6349401473999023, + "learning_rate": 1.9857819905213272e-05, + "loss": 0.4889, + "step": 419 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 2.15582013130188, + "learning_rate": 1.990521327014218e-05, + "loss": 0.5021, + "step": 420 + }, + { + "epoch": 0.2993777777777778, + "grad_norm": 1.7637985944747925, + "learning_rate": 1.9952606635071093e-05, + "loss": 0.5051, + "step": 421 + }, + { + "epoch": 0.3000888888888889, + "grad_norm": 2.833714246749878, + "learning_rate": 2e-05, + "loss": 0.5258, + "step": 422 + }, + { + "epoch": 0.3008, + "grad_norm": 1.9776523113250732, + "learning_rate": 1.9999996575341136e-05, + "loss": 0.524, + "step": 423 + }, + { + "epoch": 0.3015111111111111, + "grad_norm": 3.507546901702881, + "learning_rate": 1.9999986301366886e-05, + "loss": 0.5336, + "step": 424 + }, + { + "epoch": 0.3022222222222222, + "grad_norm": 2.799515724182129, + "learning_rate": 1.9999969178084283e-05, + "loss": 0.4927, + "step": 425 + }, + { + "epoch": 0.30293333333333333, + "grad_norm": 2.1473653316497803, + "learning_rate": 1.999994520550506e-05, + "loss": 0.4859, + "step": 426 + }, + { + "epoch": 0.30364444444444444, + "grad_norm": 3.0916366577148438, + "learning_rate": 1.9999914383645636e-05, + "loss": 0.4839, + "step": 427 + }, + { + "epoch": 0.30435555555555555, + "grad_norm": 1.6858891248703003, + "learning_rate": 1.9999876712527123e-05, + "loss": 0.5077, + "step": 428 + }, + { + "epoch": 0.30506666666666665, + "grad_norm": 2.501232147216797, + "learning_rate": 1.999983219217532e-05, + "loss": 0.4906, + "step": 429 + }, + { + "epoch": 0.30577777777777776, + "grad_norm": 2.9373323917388916, + "learning_rate": 1.9999780822620726e-05, + "loss": 0.4991, + "step": 430 + }, + { + "epoch": 0.30648888888888887, + "grad_norm": 1.808949589729309, + "learning_rate": 1.9999722603898517e-05, + "loss": 0.4821, + "step": 431 + }, + { + "epoch": 0.3072, + "grad_norm": 1.9162492752075195, + "learning_rate": 1.9999657536048575e-05, + "loss": 0.5335, + "step": 432 + }, + { + "epoch": 0.30791111111111114, + "grad_norm": 2.083402633666992, + "learning_rate": 1.9999585619115466e-05, + "loss": 0.4963, + "step": 433 + }, + { + "epoch": 0.30862222222222224, + "grad_norm": 2.8350741863250732, + "learning_rate": 1.9999506853148447e-05, + "loss": 0.511, + "step": 434 + }, + { + "epoch": 0.30933333333333335, + "grad_norm": 2.0338497161865234, + "learning_rate": 1.9999421238201465e-05, + "loss": 0.5424, + "step": 435 + }, + { + "epoch": 0.31004444444444446, + "grad_norm": 3.66500186920166, + "learning_rate": 1.999932877433317e-05, + "loss": 0.4839, + "step": 436 + }, + { + "epoch": 0.31075555555555556, + "grad_norm": 2.2418339252471924, + "learning_rate": 1.9999229461606884e-05, + "loss": 0.4962, + "step": 437 + }, + { + "epoch": 0.31146666666666667, + "grad_norm": 3.301936626434326, + "learning_rate": 1.9999123300090633e-05, + "loss": 0.4858, + "step": 438 + }, + { + "epoch": 0.3121777777777778, + "grad_norm": 2.6144609451293945, + "learning_rate": 1.999901028985713e-05, + "loss": 0.5326, + "step": 439 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 2.398486375808716, + "learning_rate": 1.9998890430983776e-05, + "loss": 0.49, + "step": 440 + }, + { + "epoch": 0.3136, + "grad_norm": 2.1547908782958984, + "learning_rate": 1.9998763723552672e-05, + "loss": 0.4994, + "step": 441 + }, + { + "epoch": 0.3143111111111111, + "grad_norm": 2.3693995475769043, + "learning_rate": 1.9998630167650603e-05, + "loss": 0.4876, + "step": 442 + }, + { + "epoch": 0.3150222222222222, + "grad_norm": 1.5921911001205444, + "learning_rate": 1.999848976336904e-05, + "loss": 0.4869, + "step": 443 + }, + { + "epoch": 0.3157333333333333, + "grad_norm": 2.184723138809204, + "learning_rate": 1.9998342510804157e-05, + "loss": 0.4508, + "step": 444 + }, + { + "epoch": 0.3164444444444444, + "grad_norm": 2.617661237716675, + "learning_rate": 1.9998188410056807e-05, + "loss": 0.507, + "step": 445 + }, + { + "epoch": 0.3171555555555556, + "grad_norm": 2.1172597408294678, + "learning_rate": 1.9998027461232545e-05, + "loss": 0.5449, + "step": 446 + }, + { + "epoch": 0.3178666666666667, + "grad_norm": 1.4677547216415405, + "learning_rate": 1.99978596644416e-05, + "loss": 0.5118, + "step": 447 + }, + { + "epoch": 0.3185777777777778, + "grad_norm": 1.407429814338684, + "learning_rate": 1.9997685019798913e-05, + "loss": 0.4879, + "step": 448 + }, + { + "epoch": 0.3192888888888889, + "grad_norm": 1.4799028635025024, + "learning_rate": 1.999750352742409e-05, + "loss": 0.4442, + "step": 449 + }, + { + "epoch": 0.32, + "grad_norm": 1.7808533906936646, + "learning_rate": 1.9997315187441452e-05, + "loss": 0.4958, + "step": 450 + }, + { + "epoch": 0.3207111111111111, + "grad_norm": 3.577451705932617, + "learning_rate": 1.9997119999979997e-05, + "loss": 0.5103, + "step": 451 + }, + { + "epoch": 0.3214222222222222, + "grad_norm": 1.7130035161972046, + "learning_rate": 1.9996917965173415e-05, + "loss": 0.4569, + "step": 452 + }, + { + "epoch": 0.3221333333333333, + "grad_norm": 3.889855146408081, + "learning_rate": 1.999670908316008e-05, + "loss": 0.4793, + "step": 453 + }, + { + "epoch": 0.32284444444444443, + "grad_norm": 2.077397584915161, + "learning_rate": 1.9996493354083074e-05, + "loss": 0.4484, + "step": 454 + }, + { + "epoch": 0.32355555555555554, + "grad_norm": 2.8907172679901123, + "learning_rate": 1.9996270778090142e-05, + "loss": 0.5011, + "step": 455 + }, + { + "epoch": 0.32426666666666665, + "grad_norm": 2.484229803085327, + "learning_rate": 1.9996041355333745e-05, + "loss": 0.5652, + "step": 456 + }, + { + "epoch": 0.32497777777777775, + "grad_norm": 1.9867464303970337, + "learning_rate": 1.9995805085971015e-05, + "loss": 0.4915, + "step": 457 + }, + { + "epoch": 0.3256888888888889, + "grad_norm": 2.7284884452819824, + "learning_rate": 1.9995561970163784e-05, + "loss": 0.5177, + "step": 458 + }, + { + "epoch": 0.3264, + "grad_norm": 2.5109829902648926, + "learning_rate": 1.9995312008078568e-05, + "loss": 0.4977, + "step": 459 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 1.859138011932373, + "learning_rate": 1.9995055199886577e-05, + "loss": 0.5284, + "step": 460 + }, + { + "epoch": 0.32782222222222224, + "grad_norm": 2.9489662647247314, + "learning_rate": 1.9994791545763707e-05, + "loss": 0.486, + "step": 461 + }, + { + "epoch": 0.32853333333333334, + "grad_norm": 1.71241295337677, + "learning_rate": 1.9994521045890535e-05, + "loss": 0.5236, + "step": 462 + }, + { + "epoch": 0.32924444444444445, + "grad_norm": 2.31514835357666, + "learning_rate": 1.9994243700452344e-05, + "loss": 0.5215, + "step": 463 + }, + { + "epoch": 0.32995555555555556, + "grad_norm": 2.592459201812744, + "learning_rate": 1.9993959509639094e-05, + "loss": 0.5343, + "step": 464 + }, + { + "epoch": 0.33066666666666666, + "grad_norm": 1.8370646238327026, + "learning_rate": 1.9993668473645436e-05, + "loss": 0.459, + "step": 465 + }, + { + "epoch": 0.33137777777777777, + "grad_norm": 2.7393970489501953, + "learning_rate": 1.999337059267071e-05, + "loss": 0.4921, + "step": 466 + }, + { + "epoch": 0.3320888888888889, + "grad_norm": 2.4636423587799072, + "learning_rate": 1.999306586691894e-05, + "loss": 0.5662, + "step": 467 + }, + { + "epoch": 0.3328, + "grad_norm": 1.70069420337677, + "learning_rate": 1.999275429659885e-05, + "loss": 0.4875, + "step": 468 + }, + { + "epoch": 0.3335111111111111, + "grad_norm": 3.3531947135925293, + "learning_rate": 1.999243588192384e-05, + "loss": 0.5028, + "step": 469 + }, + { + "epoch": 0.3342222222222222, + "grad_norm": 2.392589569091797, + "learning_rate": 1.9992110623112004e-05, + "loss": 0.5005, + "step": 470 + }, + { + "epoch": 0.33493333333333336, + "grad_norm": 2.196497917175293, + "learning_rate": 1.999177852038612e-05, + "loss": 0.5023, + "step": 471 + }, + { + "epoch": 0.33564444444444447, + "grad_norm": 4.078185558319092, + "learning_rate": 1.9991439573973655e-05, + "loss": 0.5217, + "step": 472 + }, + { + "epoch": 0.3363555555555556, + "grad_norm": 2.5604310035705566, + "learning_rate": 1.999109378410677e-05, + "loss": 0.4999, + "step": 473 + }, + { + "epoch": 0.3370666666666667, + "grad_norm": 4.548199653625488, + "learning_rate": 1.9990741151022302e-05, + "loss": 0.4865, + "step": 474 + }, + { + "epoch": 0.3377777777777778, + "grad_norm": 4.355525016784668, + "learning_rate": 1.999038167496178e-05, + "loss": 0.4711, + "step": 475 + }, + { + "epoch": 0.3384888888888889, + "grad_norm": 2.9667112827301025, + "learning_rate": 1.9990015356171426e-05, + "loss": 0.5231, + "step": 476 + }, + { + "epoch": 0.3392, + "grad_norm": 2.90069580078125, + "learning_rate": 1.9989642194902136e-05, + "loss": 0.4851, + "step": 477 + }, + { + "epoch": 0.3399111111111111, + "grad_norm": 2.9968421459198, + "learning_rate": 1.9989262191409505e-05, + "loss": 0.5129, + "step": 478 + }, + { + "epoch": 0.3406222222222222, + "grad_norm": 2.5186960697174072, + "learning_rate": 1.9988875345953808e-05, + "loss": 0.4542, + "step": 479 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 4.55244255065918, + "learning_rate": 1.9988481658800012e-05, + "loss": 0.5042, + "step": 480 + }, + { + "epoch": 0.34204444444444443, + "grad_norm": 3.295694351196289, + "learning_rate": 1.998808113021776e-05, + "loss": 0.5313, + "step": 481 + }, + { + "epoch": 0.34275555555555554, + "grad_norm": 2.84930419921875, + "learning_rate": 1.9987673760481387e-05, + "loss": 0.4825, + "step": 482 + }, + { + "epoch": 0.34346666666666664, + "grad_norm": 2.670814275741577, + "learning_rate": 1.998725954986992e-05, + "loss": 0.4902, + "step": 483 + }, + { + "epoch": 0.3441777777777778, + "grad_norm": 3.8628222942352295, + "learning_rate": 1.9986838498667053e-05, + "loss": 0.4976, + "step": 484 + }, + { + "epoch": 0.3448888888888889, + "grad_norm": 2.688530206680298, + "learning_rate": 1.9986410607161192e-05, + "loss": 0.4882, + "step": 485 + }, + { + "epoch": 0.3456, + "grad_norm": 3.8878402709960938, + "learning_rate": 1.99859758756454e-05, + "loss": 0.5384, + "step": 486 + }, + { + "epoch": 0.3463111111111111, + "grad_norm": 3.900045156478882, + "learning_rate": 1.9985534304417448e-05, + "loss": 0.4958, + "step": 487 + }, + { + "epoch": 0.34702222222222223, + "grad_norm": 2.550661087036133, + "learning_rate": 1.9985085893779776e-05, + "loss": 0.4671, + "step": 488 + }, + { + "epoch": 0.34773333333333334, + "grad_norm": 2.6415789127349854, + "learning_rate": 1.998463064403952e-05, + "loss": 0.5008, + "step": 489 + }, + { + "epoch": 0.34844444444444445, + "grad_norm": 2.9456567764282227, + "learning_rate": 1.9984168555508493e-05, + "loss": 0.4743, + "step": 490 + }, + { + "epoch": 0.34915555555555555, + "grad_norm": 1.9302077293395996, + "learning_rate": 1.998369962850319e-05, + "loss": 0.5084, + "step": 491 + }, + { + "epoch": 0.34986666666666666, + "grad_norm": 3.866057872772217, + "learning_rate": 1.99832238633448e-05, + "loss": 0.47, + "step": 492 + }, + { + "epoch": 0.35057777777777777, + "grad_norm": 2.3373379707336426, + "learning_rate": 1.9982741260359188e-05, + "loss": 0.4421, + "step": 493 + }, + { + "epoch": 0.3512888888888889, + "grad_norm": 3.876507520675659, + "learning_rate": 1.99822518198769e-05, + "loss": 0.4874, + "step": 494 + }, + { + "epoch": 0.352, + "grad_norm": 3.0445845127105713, + "learning_rate": 1.9981755542233175e-05, + "loss": 0.4929, + "step": 495 + }, + { + "epoch": 0.3527111111111111, + "grad_norm": 3.163206100463867, + "learning_rate": 1.998125242776793e-05, + "loss": 0.4583, + "step": 496 + }, + { + "epoch": 0.35342222222222225, + "grad_norm": 2.4368948936462402, + "learning_rate": 1.9980742476825757e-05, + "loss": 0.5351, + "step": 497 + }, + { + "epoch": 0.35413333333333336, + "grad_norm": 3.9216771125793457, + "learning_rate": 1.998022568975594e-05, + "loss": 0.456, + "step": 498 + }, + { + "epoch": 0.35484444444444446, + "grad_norm": 2.852274179458618, + "learning_rate": 1.997970206691245e-05, + "loss": 0.4936, + "step": 499 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 3.0261611938476562, + "learning_rate": 1.9979171608653926e-05, + "loss": 0.4997, + "step": 500 + }, + { + "epoch": 0.3562666666666667, + "grad_norm": 2.3791420459747314, + "learning_rate": 1.9978634315343693e-05, + "loss": 0.4795, + "step": 501 + }, + { + "epoch": 0.3569777777777778, + "grad_norm": 3.0706400871276855, + "learning_rate": 1.9978090187349766e-05, + "loss": 0.493, + "step": 502 + }, + { + "epoch": 0.3576888888888889, + "grad_norm": 2.2512102127075195, + "learning_rate": 1.9977539225044836e-05, + "loss": 0.4623, + "step": 503 + }, + { + "epoch": 0.3584, + "grad_norm": 3.5871524810791016, + "learning_rate": 1.9976981428806272e-05, + "loss": 0.4337, + "step": 504 + }, + { + "epoch": 0.3591111111111111, + "grad_norm": 3.3260562419891357, + "learning_rate": 1.9976416799016128e-05, + "loss": 0.4728, + "step": 505 + }, + { + "epoch": 0.3598222222222222, + "grad_norm": 2.604079246520996, + "learning_rate": 1.9975845336061134e-05, + "loss": 0.4907, + "step": 506 + }, + { + "epoch": 0.3605333333333333, + "grad_norm": 2.720327615737915, + "learning_rate": 1.9975267040332705e-05, + "loss": 0.4722, + "step": 507 + }, + { + "epoch": 0.3612444444444444, + "grad_norm": 2.405660629272461, + "learning_rate": 1.9974681912226934e-05, + "loss": 0.4526, + "step": 508 + }, + { + "epoch": 0.36195555555555553, + "grad_norm": 2.1337242126464844, + "learning_rate": 1.9974089952144595e-05, + "loss": 0.5404, + "step": 509 + }, + { + "epoch": 0.3626666666666667, + "grad_norm": 3.01167893409729, + "learning_rate": 1.9973491160491137e-05, + "loss": 0.488, + "step": 510 + }, + { + "epoch": 0.3633777777777778, + "grad_norm": 2.496649980545044, + "learning_rate": 1.99728855376767e-05, + "loss": 0.4526, + "step": 511 + }, + { + "epoch": 0.3640888888888889, + "grad_norm": 2.5681872367858887, + "learning_rate": 1.9972273084116082e-05, + "loss": 0.4672, + "step": 512 + }, + { + "epoch": 0.3648, + "grad_norm": 2.0203449726104736, + "learning_rate": 1.9971653800228782e-05, + "loss": 0.4056, + "step": 513 + }, + { + "epoch": 0.3655111111111111, + "grad_norm": 2.7935714721679688, + "learning_rate": 1.9971027686438956e-05, + "loss": 0.4549, + "step": 514 + }, + { + "epoch": 0.3662222222222222, + "grad_norm": 2.2784855365753174, + "learning_rate": 1.9970394743175464e-05, + "loss": 0.4461, + "step": 515 + }, + { + "epoch": 0.36693333333333333, + "grad_norm": 3.7940878868103027, + "learning_rate": 1.9969754970871818e-05, + "loss": 0.4867, + "step": 516 + }, + { + "epoch": 0.36764444444444444, + "grad_norm": 3.386700391769409, + "learning_rate": 1.996910836996622e-05, + "loss": 0.5076, + "step": 517 + }, + { + "epoch": 0.36835555555555555, + "grad_norm": 2.5359184741973877, + "learning_rate": 1.996845494090155e-05, + "loss": 0.45, + "step": 518 + }, + { + "epoch": 0.36906666666666665, + "grad_norm": 2.872170925140381, + "learning_rate": 1.9967794684125363e-05, + "loss": 0.5292, + "step": 519 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 2.7346885204315186, + "learning_rate": 1.9967127600089885e-05, + "loss": 0.4809, + "step": 520 + }, + { + "epoch": 0.37048888888888887, + "grad_norm": 2.2059731483459473, + "learning_rate": 1.996645368925203e-05, + "loss": 0.4484, + "step": 521 + }, + { + "epoch": 0.3712, + "grad_norm": 2.3617849349975586, + "learning_rate": 1.9965772952073376e-05, + "loss": 0.4876, + "step": 522 + }, + { + "epoch": 0.37191111111111114, + "grad_norm": 1.888338327407837, + "learning_rate": 1.996508538902018e-05, + "loss": 0.4977, + "step": 523 + }, + { + "epoch": 0.37262222222222224, + "grad_norm": 2.495836019515991, + "learning_rate": 1.9964391000563382e-05, + "loss": 0.4965, + "step": 524 + }, + { + "epoch": 0.37333333333333335, + "grad_norm": 1.7823423147201538, + "learning_rate": 1.9963689787178587e-05, + "loss": 0.5035, + "step": 525 + }, + { + "epoch": 0.37404444444444446, + "grad_norm": 3.7695841789245605, + "learning_rate": 1.996298174934608e-05, + "loss": 0.5057, + "step": 526 + }, + { + "epoch": 0.37475555555555556, + "grad_norm": 3.202395439147949, + "learning_rate": 1.9962266887550816e-05, + "loss": 0.4422, + "step": 527 + }, + { + "epoch": 0.37546666666666667, + "grad_norm": 2.0220539569854736, + "learning_rate": 1.9961545202282427e-05, + "loss": 0.4571, + "step": 528 + }, + { + "epoch": 0.3761777777777778, + "grad_norm": 2.6506786346435547, + "learning_rate": 1.996081669403522e-05, + "loss": 0.4697, + "step": 529 + }, + { + "epoch": 0.3768888888888889, + "grad_norm": 5.04495096206665, + "learning_rate": 1.996008136330817e-05, + "loss": 0.5338, + "step": 530 + }, + { + "epoch": 0.3776, + "grad_norm": 2.8911781311035156, + "learning_rate": 1.9959339210604934e-05, + "loss": 0.4762, + "step": 531 + }, + { + "epoch": 0.3783111111111111, + "grad_norm": 1.308676838874817, + "learning_rate": 1.995859023643383e-05, + "loss": 0.4669, + "step": 532 + }, + { + "epoch": 0.3790222222222222, + "grad_norm": 3.514946460723877, + "learning_rate": 1.995783444130786e-05, + "loss": 0.5027, + "step": 533 + }, + { + "epoch": 0.3797333333333333, + "grad_norm": 1.993985652923584, + "learning_rate": 1.9957071825744685e-05, + "loss": 0.5256, + "step": 534 + }, + { + "epoch": 0.3804444444444444, + "grad_norm": 3.2812976837158203, + "learning_rate": 1.9956302390266652e-05, + "loss": 0.4832, + "step": 535 + }, + { + "epoch": 0.3811555555555556, + "grad_norm": 3.480802536010742, + "learning_rate": 1.9955526135400768e-05, + "loss": 0.4695, + "step": 536 + }, + { + "epoch": 0.3818666666666667, + "grad_norm": 1.8565747737884521, + "learning_rate": 1.9954743061678715e-05, + "loss": 0.5278, + "step": 537 + }, + { + "epoch": 0.3825777777777778, + "grad_norm": 2.8600106239318848, + "learning_rate": 1.995395316963684e-05, + "loss": 0.4467, + "step": 538 + }, + { + "epoch": 0.3832888888888889, + "grad_norm": 1.7738546133041382, + "learning_rate": 1.995315645981618e-05, + "loss": 0.5176, + "step": 539 + }, + { + "epoch": 0.384, + "grad_norm": 2.126570463180542, + "learning_rate": 1.995235293276241e-05, + "loss": 0.4809, + "step": 540 + }, + { + "epoch": 0.3847111111111111, + "grad_norm": 2.0607144832611084, + "learning_rate": 1.99515425890259e-05, + "loss": 0.4797, + "step": 541 + }, + { + "epoch": 0.3854222222222222, + "grad_norm": 1.8785585165023804, + "learning_rate": 1.995072542916168e-05, + "loss": 0.4761, + "step": 542 + }, + { + "epoch": 0.38613333333333333, + "grad_norm": 2.8150289058685303, + "learning_rate": 1.9949901453729444e-05, + "loss": 0.4861, + "step": 543 + }, + { + "epoch": 0.38684444444444444, + "grad_norm": 2.202181577682495, + "learning_rate": 1.9949070663293567e-05, + "loss": 0.4961, + "step": 544 + }, + { + "epoch": 0.38755555555555554, + "grad_norm": 2.639108419418335, + "learning_rate": 1.9948233058423078e-05, + "loss": 0.4651, + "step": 545 + }, + { + "epoch": 0.38826666666666665, + "grad_norm": 2.6120409965515137, + "learning_rate": 1.9947388639691678e-05, + "loss": 0.4495, + "step": 546 + }, + { + "epoch": 0.38897777777777776, + "grad_norm": 3.272209644317627, + "learning_rate": 1.994653740767774e-05, + "loss": 0.4882, + "step": 547 + }, + { + "epoch": 0.38968888888888886, + "grad_norm": 2.62929105758667, + "learning_rate": 1.9945679362964298e-05, + "loss": 0.496, + "step": 548 + }, + { + "epoch": 0.3904, + "grad_norm": 1.9820858240127563, + "learning_rate": 1.9944814506139053e-05, + "loss": 0.4917, + "step": 549 + }, + { + "epoch": 0.39111111111111113, + "grad_norm": 2.071793556213379, + "learning_rate": 1.9943942837794378e-05, + "loss": 0.5148, + "step": 550 + }, + { + "epoch": 0.39182222222222224, + "grad_norm": 2.5890281200408936, + "learning_rate": 1.99430643585273e-05, + "loss": 0.4823, + "step": 551 + }, + { + "epoch": 0.39253333333333335, + "grad_norm": 1.8528071641921997, + "learning_rate": 1.994217906893952e-05, + "loss": 0.4823, + "step": 552 + }, + { + "epoch": 0.39324444444444445, + "grad_norm": 2.8410422801971436, + "learning_rate": 1.9941286969637402e-05, + "loss": 0.5188, + "step": 553 + }, + { + "epoch": 0.39395555555555556, + "grad_norm": 2.1342577934265137, + "learning_rate": 1.994038806123197e-05, + "loss": 0.441, + "step": 554 + }, + { + "epoch": 0.39466666666666667, + "grad_norm": 3.021951198577881, + "learning_rate": 1.9939482344338922e-05, + "loss": 0.4611, + "step": 555 + }, + { + "epoch": 0.3953777777777778, + "grad_norm": 2.306688070297241, + "learning_rate": 1.9938569819578605e-05, + "loss": 0.4555, + "step": 556 + }, + { + "epoch": 0.3960888888888889, + "grad_norm": 2.652862071990967, + "learning_rate": 1.9937650487576038e-05, + "loss": 0.4516, + "step": 557 + }, + { + "epoch": 0.3968, + "grad_norm": 2.0471227169036865, + "learning_rate": 1.9936724348960902e-05, + "loss": 0.4924, + "step": 558 + }, + { + "epoch": 0.3975111111111111, + "grad_norm": 3.0647501945495605, + "learning_rate": 1.9935791404367534e-05, + "loss": 0.4448, + "step": 559 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 1.9600642919540405, + "learning_rate": 1.9934851654434945e-05, + "loss": 0.4688, + "step": 560 + }, + { + "epoch": 0.3989333333333333, + "grad_norm": 2.9962406158447266, + "learning_rate": 1.9933905099806796e-05, + "loss": 0.4847, + "step": 561 + }, + { + "epoch": 0.39964444444444447, + "grad_norm": 2.5012073516845703, + "learning_rate": 1.9932951741131413e-05, + "loss": 0.4791, + "step": 562 + }, + { + "epoch": 0.4003555555555556, + "grad_norm": 2.3866870403289795, + "learning_rate": 1.993199157906178e-05, + "loss": 0.4871, + "step": 563 + }, + { + "epoch": 0.4010666666666667, + "grad_norm": 2.652003288269043, + "learning_rate": 1.9931024614255542e-05, + "loss": 0.4981, + "step": 564 + }, + { + "epoch": 0.4017777777777778, + "grad_norm": 2.260114908218384, + "learning_rate": 1.993005084737501e-05, + "loss": 0.4213, + "step": 565 + }, + { + "epoch": 0.4024888888888889, + "grad_norm": 1.8540630340576172, + "learning_rate": 1.9929070279087137e-05, + "loss": 0.4413, + "step": 566 + }, + { + "epoch": 0.4032, + "grad_norm": 1.7977911233901978, + "learning_rate": 1.9928082910063556e-05, + "loss": 0.4629, + "step": 567 + }, + { + "epoch": 0.4039111111111111, + "grad_norm": 1.8995805978775024, + "learning_rate": 1.992708874098054e-05, + "loss": 0.4697, + "step": 568 + }, + { + "epoch": 0.4046222222222222, + "grad_norm": 2.047802686691284, + "learning_rate": 1.9926087772519035e-05, + "loss": 0.5111, + "step": 569 + }, + { + "epoch": 0.4053333333333333, + "grad_norm": 1.9696111679077148, + "learning_rate": 1.9925080005364628e-05, + "loss": 0.4932, + "step": 570 + }, + { + "epoch": 0.40604444444444443, + "grad_norm": 2.1687138080596924, + "learning_rate": 1.9924065440207575e-05, + "loss": 0.4779, + "step": 571 + }, + { + "epoch": 0.40675555555555554, + "grad_norm": 1.8573496341705322, + "learning_rate": 1.992304407774278e-05, + "loss": 0.539, + "step": 572 + }, + { + "epoch": 0.40746666666666664, + "grad_norm": 2.650284767150879, + "learning_rate": 1.992201591866981e-05, + "loss": 0.4231, + "step": 573 + }, + { + "epoch": 0.40817777777777775, + "grad_norm": 2.0399317741394043, + "learning_rate": 1.9920980963692887e-05, + "loss": 0.4593, + "step": 574 + }, + { + "epoch": 0.4088888888888889, + "grad_norm": 3.5517942905426025, + "learning_rate": 1.991993921352088e-05, + "loss": 0.4851, + "step": 575 + }, + { + "epoch": 0.4096, + "grad_norm": 2.743251085281372, + "learning_rate": 1.9918890668867315e-05, + "loss": 0.4481, + "step": 576 + }, + { + "epoch": 0.4103111111111111, + "grad_norm": 2.3659846782684326, + "learning_rate": 1.991783533045038e-05, + "loss": 0.4564, + "step": 577 + }, + { + "epoch": 0.41102222222222223, + "grad_norm": 2.351651906967163, + "learning_rate": 1.99167731989929e-05, + "loss": 0.4581, + "step": 578 + }, + { + "epoch": 0.41173333333333334, + "grad_norm": 1.9420870542526245, + "learning_rate": 1.9915704275222374e-05, + "loss": 0.4525, + "step": 579 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 2.051187753677368, + "learning_rate": 1.9914628559870933e-05, + "loss": 0.4378, + "step": 580 + }, + { + "epoch": 0.41315555555555555, + "grad_norm": 2.2282252311706543, + "learning_rate": 1.9913546053675374e-05, + "loss": 0.4725, + "step": 581 + }, + { + "epoch": 0.41386666666666666, + "grad_norm": 2.88924241065979, + "learning_rate": 1.9912456757377138e-05, + "loss": 0.5006, + "step": 582 + }, + { + "epoch": 0.41457777777777777, + "grad_norm": 1.9519983530044556, + "learning_rate": 1.991136067172232e-05, + "loss": 0.4589, + "step": 583 + }, + { + "epoch": 0.4152888888888889, + "grad_norm": 2.071173667907715, + "learning_rate": 1.991025779746166e-05, + "loss": 0.526, + "step": 584 + }, + { + "epoch": 0.416, + "grad_norm": 1.7574818134307861, + "learning_rate": 1.9909148135350553e-05, + "loss": 0.4886, + "step": 585 + }, + { + "epoch": 0.4167111111111111, + "grad_norm": 2.7940051555633545, + "learning_rate": 1.9908031686149045e-05, + "loss": 0.5113, + "step": 586 + }, + { + "epoch": 0.4174222222222222, + "grad_norm": 2.0076651573181152, + "learning_rate": 1.9906908450621825e-05, + "loss": 0.4491, + "step": 587 + }, + { + "epoch": 0.41813333333333336, + "grad_norm": 1.6784553527832031, + "learning_rate": 1.9905778429538232e-05, + "loss": 0.4439, + "step": 588 + }, + { + "epoch": 0.41884444444444446, + "grad_norm": 2.2727744579315186, + "learning_rate": 1.9904641623672258e-05, + "loss": 0.5234, + "step": 589 + }, + { + "epoch": 0.41955555555555557, + "grad_norm": 1.6602866649627686, + "learning_rate": 1.990349803380253e-05, + "loss": 0.4592, + "step": 590 + }, + { + "epoch": 0.4202666666666667, + "grad_norm": 1.6065152883529663, + "learning_rate": 1.9902347660712336e-05, + "loss": 0.4943, + "step": 591 + }, + { + "epoch": 0.4209777777777778, + "grad_norm": 1.7765272855758667, + "learning_rate": 1.9901190505189597e-05, + "loss": 0.4604, + "step": 592 + }, + { + "epoch": 0.4216888888888889, + "grad_norm": 2.3617889881134033, + "learning_rate": 1.9900026568026895e-05, + "loss": 0.4694, + "step": 593 + }, + { + "epoch": 0.4224, + "grad_norm": 1.8684749603271484, + "learning_rate": 1.9898855850021436e-05, + "loss": 0.4777, + "step": 594 + }, + { + "epoch": 0.4231111111111111, + "grad_norm": 1.9223345518112183, + "learning_rate": 1.9897678351975086e-05, + "loss": 0.4834, + "step": 595 + }, + { + "epoch": 0.4238222222222222, + "grad_norm": 1.865857481956482, + "learning_rate": 1.9896494074694352e-05, + "loss": 0.4697, + "step": 596 + }, + { + "epoch": 0.4245333333333333, + "grad_norm": 1.280529260635376, + "learning_rate": 1.9895303018990388e-05, + "loss": 0.4549, + "step": 597 + }, + { + "epoch": 0.4252444444444444, + "grad_norm": 1.6691361665725708, + "learning_rate": 1.9894105185678977e-05, + "loss": 0.4847, + "step": 598 + }, + { + "epoch": 0.42595555555555553, + "grad_norm": 1.44730806350708, + "learning_rate": 1.989290057558056e-05, + "loss": 0.4992, + "step": 599 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 1.7356247901916504, + "learning_rate": 1.9891689189520208e-05, + "loss": 0.4985, + "step": 600 + }, + { + "epoch": 0.4273777777777778, + "grad_norm": 1.496366024017334, + "learning_rate": 1.9890471028327637e-05, + "loss": 0.5156, + "step": 601 + }, + { + "epoch": 0.4280888888888889, + "grad_norm": 1.4667977094650269, + "learning_rate": 1.9889246092837208e-05, + "loss": 0.4892, + "step": 602 + }, + { + "epoch": 0.4288, + "grad_norm": 1.5844496488571167, + "learning_rate": 1.988801438388792e-05, + "loss": 0.4932, + "step": 603 + }, + { + "epoch": 0.4295111111111111, + "grad_norm": 1.2338106632232666, + "learning_rate": 1.9886775902323405e-05, + "loss": 0.4964, + "step": 604 + }, + { + "epoch": 0.43022222222222223, + "grad_norm": 1.9749044179916382, + "learning_rate": 1.988553064899194e-05, + "loss": 0.4653, + "step": 605 + }, + { + "epoch": 0.43093333333333333, + "grad_norm": 2.4607295989990234, + "learning_rate": 1.9884278624746442e-05, + "loss": 0.4918, + "step": 606 + }, + { + "epoch": 0.43164444444444444, + "grad_norm": 1.9557958841323853, + "learning_rate": 1.9883019830444455e-05, + "loss": 0.4402, + "step": 607 + }, + { + "epoch": 0.43235555555555555, + "grad_norm": 1.6397583484649658, + "learning_rate": 1.9881754266948174e-05, + "loss": 0.513, + "step": 608 + }, + { + "epoch": 0.43306666666666666, + "grad_norm": 2.2165520191192627, + "learning_rate": 1.988048193512442e-05, + "loss": 0.5186, + "step": 609 + }, + { + "epoch": 0.43377777777777776, + "grad_norm": 1.71074640750885, + "learning_rate": 1.987920283584465e-05, + "loss": 0.4762, + "step": 610 + }, + { + "epoch": 0.43448888888888887, + "grad_norm": 1.482975959777832, + "learning_rate": 1.987791696998497e-05, + "loss": 0.4542, + "step": 611 + }, + { + "epoch": 0.4352, + "grad_norm": 1.6289290189743042, + "learning_rate": 1.9876624338426103e-05, + "loss": 0.4787, + "step": 612 + }, + { + "epoch": 0.43591111111111114, + "grad_norm": 1.2894456386566162, + "learning_rate": 1.987532494205341e-05, + "loss": 0.5078, + "step": 613 + }, + { + "epoch": 0.43662222222222224, + "grad_norm": 2.005520820617676, + "learning_rate": 1.9874018781756898e-05, + "loss": 0.5144, + "step": 614 + }, + { + "epoch": 0.43733333333333335, + "grad_norm": 1.734578251838684, + "learning_rate": 1.9872705858431196e-05, + "loss": 0.5304, + "step": 615 + }, + { + "epoch": 0.43804444444444446, + "grad_norm": 1.6002241373062134, + "learning_rate": 1.987138617297556e-05, + "loss": 0.5027, + "step": 616 + }, + { + "epoch": 0.43875555555555557, + "grad_norm": 1.5120142698287964, + "learning_rate": 1.9870059726293892e-05, + "loss": 0.4942, + "step": 617 + }, + { + "epoch": 0.43946666666666667, + "grad_norm": 1.3984020948410034, + "learning_rate": 1.986872651929471e-05, + "loss": 0.4728, + "step": 618 + }, + { + "epoch": 0.4401777777777778, + "grad_norm": 1.5515751838684082, + "learning_rate": 1.9867386552891175e-05, + "loss": 0.5102, + "step": 619 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 1.30840265750885, + "learning_rate": 1.9866039828001075e-05, + "loss": 0.4663, + "step": 620 + }, + { + "epoch": 0.4416, + "grad_norm": 1.6964130401611328, + "learning_rate": 1.9864686345546818e-05, + "loss": 0.4605, + "step": 621 + }, + { + "epoch": 0.4423111111111111, + "grad_norm": 1.7013435363769531, + "learning_rate": 1.986332610645545e-05, + "loss": 0.4353, + "step": 622 + }, + { + "epoch": 0.4430222222222222, + "grad_norm": 1.2881067991256714, + "learning_rate": 1.9861959111658646e-05, + "loss": 0.4439, + "step": 623 + }, + { + "epoch": 0.4437333333333333, + "grad_norm": 1.44131338596344, + "learning_rate": 1.98605853620927e-05, + "loss": 0.4156, + "step": 624 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 1.5430924892425537, + "learning_rate": 1.9859204858698535e-05, + "loss": 0.4474, + "step": 625 + }, + { + "epoch": 0.4451555555555556, + "grad_norm": 1.3087620735168457, + "learning_rate": 1.9857817602421705e-05, + "loss": 0.499, + "step": 626 + }, + { + "epoch": 0.4458666666666667, + "grad_norm": 2.715606927871704, + "learning_rate": 1.9856423594212388e-05, + "loss": 0.4521, + "step": 627 + }, + { + "epoch": 0.4465777777777778, + "grad_norm": 1.944486379623413, + "learning_rate": 1.9855022835025373e-05, + "loss": 0.5222, + "step": 628 + }, + { + "epoch": 0.4472888888888889, + "grad_norm": 3.2671902179718018, + "learning_rate": 1.98536153258201e-05, + "loss": 0.5209, + "step": 629 + }, + { + "epoch": 0.448, + "grad_norm": 3.191406011581421, + "learning_rate": 1.9852201067560607e-05, + "loss": 0.4741, + "step": 630 + }, + { + "epoch": 0.4487111111111111, + "grad_norm": 2.3659451007843018, + "learning_rate": 1.9850780061215567e-05, + "loss": 0.4488, + "step": 631 + }, + { + "epoch": 0.4494222222222222, + "grad_norm": 2.004542827606201, + "learning_rate": 1.9849352307758277e-05, + "loss": 0.4916, + "step": 632 + }, + { + "epoch": 0.45013333333333333, + "grad_norm": 2.104529857635498, + "learning_rate": 1.984791780816664e-05, + "loss": 0.466, + "step": 633 + }, + { + "epoch": 0.45084444444444444, + "grad_norm": 1.7619432210922241, + "learning_rate": 1.98464765634232e-05, + "loss": 0.4922, + "step": 634 + }, + { + "epoch": 0.45155555555555554, + "grad_norm": 2.3756513595581055, + "learning_rate": 1.9845028574515106e-05, + "loss": 0.4669, + "step": 635 + }, + { + "epoch": 0.45226666666666665, + "grad_norm": 2.0950958728790283, + "learning_rate": 1.9843573842434135e-05, + "loss": 0.5049, + "step": 636 + }, + { + "epoch": 0.45297777777777776, + "grad_norm": 2.6618034839630127, + "learning_rate": 1.9842112368176674e-05, + "loss": 0.4641, + "step": 637 + }, + { + "epoch": 0.45368888888888886, + "grad_norm": 2.1522297859191895, + "learning_rate": 1.9840644152743742e-05, + "loss": 0.4852, + "step": 638 + }, + { + "epoch": 0.4544, + "grad_norm": 2.720611333847046, + "learning_rate": 1.983916919714096e-05, + "loss": 0.4854, + "step": 639 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 1.783340334892273, + "learning_rate": 1.983768750237857e-05, + "loss": 0.5101, + "step": 640 + }, + { + "epoch": 0.45582222222222224, + "grad_norm": 2.9449591636657715, + "learning_rate": 1.983619906947144e-05, + "loss": 0.4905, + "step": 641 + }, + { + "epoch": 0.45653333333333335, + "grad_norm": 2.9385428428649902, + "learning_rate": 1.9834703899439035e-05, + "loss": 0.5073, + "step": 642 + }, + { + "epoch": 0.45724444444444445, + "grad_norm": 1.5325900316238403, + "learning_rate": 1.9833201993305452e-05, + "loss": 0.4758, + "step": 643 + }, + { + "epoch": 0.45795555555555556, + "grad_norm": 2.5896053314208984, + "learning_rate": 1.983169335209939e-05, + "loss": 0.5035, + "step": 644 + }, + { + "epoch": 0.45866666666666667, + "grad_norm": 2.4543726444244385, + "learning_rate": 1.983017797685417e-05, + "loss": 0.4814, + "step": 645 + }, + { + "epoch": 0.4593777777777778, + "grad_norm": 1.4862459897994995, + "learning_rate": 1.9828655868607712e-05, + "loss": 0.4667, + "step": 646 + }, + { + "epoch": 0.4600888888888889, + "grad_norm": 2.348222017288208, + "learning_rate": 1.9827127028402566e-05, + "loss": 0.5277, + "step": 647 + }, + { + "epoch": 0.4608, + "grad_norm": 2.456897497177124, + "learning_rate": 1.982559145728588e-05, + "loss": 0.5178, + "step": 648 + }, + { + "epoch": 0.4615111111111111, + "grad_norm": 1.9403817653656006, + "learning_rate": 1.9824049156309413e-05, + "loss": 0.4899, + "step": 649 + }, + { + "epoch": 0.4622222222222222, + "grad_norm": 1.6274908781051636, + "learning_rate": 1.9822500126529536e-05, + "loss": 0.4594, + "step": 650 + }, + { + "epoch": 0.4629333333333333, + "grad_norm": 1.6605608463287354, + "learning_rate": 1.9820944369007233e-05, + "loss": 0.4223, + "step": 651 + }, + { + "epoch": 0.46364444444444447, + "grad_norm": 1.5630549192428589, + "learning_rate": 1.981938188480809e-05, + "loss": 0.4852, + "step": 652 + }, + { + "epoch": 0.4643555555555556, + "grad_norm": 1.5902000665664673, + "learning_rate": 1.98178126750023e-05, + "loss": 0.4825, + "step": 653 + }, + { + "epoch": 0.4650666666666667, + "grad_norm": 1.5126309394836426, + "learning_rate": 1.9816236740664664e-05, + "loss": 0.5061, + "step": 654 + }, + { + "epoch": 0.4657777777777778, + "grad_norm": 1.2672244310379028, + "learning_rate": 1.981465408287459e-05, + "loss": 0.4378, + "step": 655 + }, + { + "epoch": 0.4664888888888889, + "grad_norm": 1.913783311843872, + "learning_rate": 1.9813064702716094e-05, + "loss": 0.4538, + "step": 656 + }, + { + "epoch": 0.4672, + "grad_norm": 1.4407871961593628, + "learning_rate": 1.9811468601277794e-05, + "loss": 0.4392, + "step": 657 + }, + { + "epoch": 0.4679111111111111, + "grad_norm": 1.6555932760238647, + "learning_rate": 1.9809865779652902e-05, + "loss": 0.5064, + "step": 658 + }, + { + "epoch": 0.4686222222222222, + "grad_norm": 1.2054680585861206, + "learning_rate": 1.980825623893925e-05, + "loss": 0.4726, + "step": 659 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 2.2315988540649414, + "learning_rate": 1.9806639980239257e-05, + "loss": 0.4925, + "step": 660 + }, + { + "epoch": 0.47004444444444443, + "grad_norm": 2.266817092895508, + "learning_rate": 1.980501700465996e-05, + "loss": 0.4782, + "step": 661 + }, + { + "epoch": 0.47075555555555554, + "grad_norm": 1.3658310174942017, + "learning_rate": 1.9803387313312973e-05, + "loss": 0.4649, + "step": 662 + }, + { + "epoch": 0.47146666666666665, + "grad_norm": 1.4084585905075073, + "learning_rate": 1.980175090731453e-05, + "loss": 0.493, + "step": 663 + }, + { + "epoch": 0.47217777777777775, + "grad_norm": 1.5978797674179077, + "learning_rate": 1.9800107787785462e-05, + "loss": 0.4321, + "step": 664 + }, + { + "epoch": 0.4728888888888889, + "grad_norm": 1.4581161737442017, + "learning_rate": 1.9798457955851188e-05, + "loss": 0.4271, + "step": 665 + }, + { + "epoch": 0.4736, + "grad_norm": 1.741991639137268, + "learning_rate": 1.9796801412641735e-05, + "loss": 0.5082, + "step": 666 + }, + { + "epoch": 0.47431111111111113, + "grad_norm": 1.1308116912841797, + "learning_rate": 1.9795138159291715e-05, + "loss": 0.4575, + "step": 667 + }, + { + "epoch": 0.47502222222222223, + "grad_norm": 1.4086958169937134, + "learning_rate": 1.979346819694035e-05, + "loss": 0.442, + "step": 668 + }, + { + "epoch": 0.47573333333333334, + "grad_norm": 1.373207926750183, + "learning_rate": 1.9791791526731444e-05, + "loss": 0.4638, + "step": 669 + }, + { + "epoch": 0.47644444444444445, + "grad_norm": 1.5367385149002075, + "learning_rate": 1.9790108149813407e-05, + "loss": 0.4693, + "step": 670 + }, + { + "epoch": 0.47715555555555556, + "grad_norm": 2.021167516708374, + "learning_rate": 1.9788418067339236e-05, + "loss": 0.4381, + "step": 671 + }, + { + "epoch": 0.47786666666666666, + "grad_norm": 2.4753341674804688, + "learning_rate": 1.978672128046652e-05, + "loss": 0.4928, + "step": 672 + }, + { + "epoch": 0.47857777777777777, + "grad_norm": 1.3335649967193604, + "learning_rate": 1.9785017790357442e-05, + "loss": 0.4882, + "step": 673 + }, + { + "epoch": 0.4792888888888889, + "grad_norm": 3.0004873275756836, + "learning_rate": 1.9783307598178784e-05, + "loss": 0.5456, + "step": 674 + }, + { + "epoch": 0.48, + "grad_norm": 2.1591508388519287, + "learning_rate": 1.9781590705101897e-05, + "loss": 0.4579, + "step": 675 + }, + { + "epoch": 0.4807111111111111, + "grad_norm": 2.9075701236724854, + "learning_rate": 1.977986711230275e-05, + "loss": 0.4622, + "step": 676 + }, + { + "epoch": 0.4814222222222222, + "grad_norm": 2.5242908000946045, + "learning_rate": 1.977813682096188e-05, + "loss": 0.4821, + "step": 677 + }, + { + "epoch": 0.48213333333333336, + "grad_norm": 2.0561559200286865, + "learning_rate": 1.9776399832264416e-05, + "loss": 0.5035, + "step": 678 + }, + { + "epoch": 0.48284444444444446, + "grad_norm": 3.0792107582092285, + "learning_rate": 1.977465614740008e-05, + "loss": 0.4808, + "step": 679 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 2.655033826828003, + "learning_rate": 1.9772905767563176e-05, + "loss": 0.4941, + "step": 680 + }, + { + "epoch": 0.4842666666666667, + "grad_norm": 2.093510627746582, + "learning_rate": 1.97711486939526e-05, + "loss": 0.4889, + "step": 681 + }, + { + "epoch": 0.4849777777777778, + "grad_norm": 2.2596147060394287, + "learning_rate": 1.976938492777182e-05, + "loss": 0.4638, + "step": 682 + }, + { + "epoch": 0.4856888888888889, + "grad_norm": 1.4487544298171997, + "learning_rate": 1.9767614470228896e-05, + "loss": 0.4547, + "step": 683 + }, + { + "epoch": 0.4864, + "grad_norm": 2.3444149494171143, + "learning_rate": 1.9765837322536476e-05, + "loss": 0.4919, + "step": 684 + }, + { + "epoch": 0.4871111111111111, + "grad_norm": 2.0651936531066895, + "learning_rate": 1.9764053485911784e-05, + "loss": 0.4554, + "step": 685 + }, + { + "epoch": 0.4878222222222222, + "grad_norm": 2.5883572101593018, + "learning_rate": 1.976226296157662e-05, + "loss": 0.494, + "step": 686 + }, + { + "epoch": 0.4885333333333333, + "grad_norm": 2.1958274841308594, + "learning_rate": 1.976046575075738e-05, + "loss": 0.5004, + "step": 687 + }, + { + "epoch": 0.4892444444444444, + "grad_norm": 3.0227925777435303, + "learning_rate": 1.975866185468502e-05, + "loss": 0.4707, + "step": 688 + }, + { + "epoch": 0.48995555555555553, + "grad_norm": 2.0557680130004883, + "learning_rate": 1.9756851274595098e-05, + "loss": 0.4443, + "step": 689 + }, + { + "epoch": 0.49066666666666664, + "grad_norm": 1.7336000204086304, + "learning_rate": 1.975503401172773e-05, + "loss": 0.4835, + "step": 690 + }, + { + "epoch": 0.4913777777777778, + "grad_norm": 2.4375674724578857, + "learning_rate": 1.9753210067327616e-05, + "loss": 0.4742, + "step": 691 + }, + { + "epoch": 0.4920888888888889, + "grad_norm": 2.33661150932312, + "learning_rate": 1.975137944264404e-05, + "loss": 0.4902, + "step": 692 + }, + { + "epoch": 0.4928, + "grad_norm": 1.9277093410491943, + "learning_rate": 1.974954213893085e-05, + "loss": 0.4257, + "step": 693 + }, + { + "epoch": 0.4935111111111111, + "grad_norm": 2.019904851913452, + "learning_rate": 1.974769815744648e-05, + "loss": 0.4568, + "step": 694 + }, + { + "epoch": 0.49422222222222223, + "grad_norm": 2.0493781566619873, + "learning_rate": 1.9745847499453918e-05, + "loss": 0.4711, + "step": 695 + }, + { + "epoch": 0.49493333333333334, + "grad_norm": 1.9954863786697388, + "learning_rate": 1.974399016622075e-05, + "loss": 0.4355, + "step": 696 + }, + { + "epoch": 0.49564444444444444, + "grad_norm": 1.877583384513855, + "learning_rate": 1.9742126159019123e-05, + "loss": 0.4479, + "step": 697 + }, + { + "epoch": 0.49635555555555555, + "grad_norm": 1.6715912818908691, + "learning_rate": 1.9740255479125747e-05, + "loss": 0.4947, + "step": 698 + }, + { + "epoch": 0.49706666666666666, + "grad_norm": 1.7335442304611206, + "learning_rate": 1.9738378127821917e-05, + "loss": 0.5152, + "step": 699 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 1.4539364576339722, + "learning_rate": 1.9736494106393488e-05, + "loss": 0.4718, + "step": 700 + }, + { + "epoch": 0.49848888888888887, + "grad_norm": 2.3710975646972656, + "learning_rate": 1.9734603416130886e-05, + "loss": 0.4649, + "step": 701 + }, + { + "epoch": 0.4992, + "grad_norm": 1.68340003490448, + "learning_rate": 1.9732706058329106e-05, + "loss": 0.4724, + "step": 702 + }, + { + "epoch": 0.4999111111111111, + "grad_norm": 2.670262336730957, + "learning_rate": 1.9730802034287704e-05, + "loss": 0.4746, + "step": 703 + }, + { + "epoch": 0.5006222222222222, + "grad_norm": 2.2915215492248535, + "learning_rate": 1.972889134531081e-05, + "loss": 0.4478, + "step": 704 + }, + { + "epoch": 0.5013333333333333, + "grad_norm": 2.602689266204834, + "learning_rate": 1.972697399270712e-05, + "loss": 0.473, + "step": 705 + }, + { + "epoch": 0.5020444444444444, + "grad_norm": 2.6824212074279785, + "learning_rate": 1.9725049977789883e-05, + "loss": 0.4547, + "step": 706 + }, + { + "epoch": 0.5027555555555555, + "grad_norm": 2.716599941253662, + "learning_rate": 1.9723119301876924e-05, + "loss": 0.4367, + "step": 707 + }, + { + "epoch": 0.5034666666666666, + "grad_norm": 2.1787993907928467, + "learning_rate": 1.9721181966290614e-05, + "loss": 0.4606, + "step": 708 + }, + { + "epoch": 0.5041777777777777, + "grad_norm": 1.7827023267745972, + "learning_rate": 1.9719237972357906e-05, + "loss": 0.4522, + "step": 709 + }, + { + "epoch": 0.5048888888888889, + "grad_norm": 1.5366742610931396, + "learning_rate": 1.97172873214103e-05, + "loss": 0.507, + "step": 710 + }, + { + "epoch": 0.5056, + "grad_norm": 1.5911592245101929, + "learning_rate": 1.971533001478386e-05, + "loss": 0.4303, + "step": 711 + }, + { + "epoch": 0.5063111111111112, + "grad_norm": 1.4469748735427856, + "learning_rate": 1.9713366053819203e-05, + "loss": 0.4815, + "step": 712 + }, + { + "epoch": 0.5070222222222223, + "grad_norm": 1.9155778884887695, + "learning_rate": 1.971139543986151e-05, + "loss": 0.4589, + "step": 713 + }, + { + "epoch": 0.5077333333333334, + "grad_norm": 1.6283910274505615, + "learning_rate": 1.9709418174260523e-05, + "loss": 0.4463, + "step": 714 + }, + { + "epoch": 0.5084444444444445, + "grad_norm": 1.8075711727142334, + "learning_rate": 1.9707434258370524e-05, + "loss": 0.4438, + "step": 715 + }, + { + "epoch": 0.5091555555555556, + "grad_norm": 1.9674094915390015, + "learning_rate": 1.970544369355037e-05, + "loss": 0.4435, + "step": 716 + }, + { + "epoch": 0.5098666666666667, + "grad_norm": 1.7431280612945557, + "learning_rate": 1.9703446481163457e-05, + "loss": 0.4007, + "step": 717 + }, + { + "epoch": 0.5105777777777778, + "grad_norm": 1.7266861200332642, + "learning_rate": 1.9701442622577736e-05, + "loss": 0.4405, + "step": 718 + }, + { + "epoch": 0.5112888888888889, + "grad_norm": 1.3482186794281006, + "learning_rate": 1.969943211916572e-05, + "loss": 0.4504, + "step": 719 + }, + { + "epoch": 0.512, + "grad_norm": 1.5705890655517578, + "learning_rate": 1.9697414972304462e-05, + "loss": 0.5327, + "step": 720 + }, + { + "epoch": 0.5127111111111111, + "grad_norm": 2.042382001876831, + "learning_rate": 1.9695391183375575e-05, + "loss": 0.4442, + "step": 721 + }, + { + "epoch": 0.5134222222222222, + "grad_norm": 1.424829125404358, + "learning_rate": 1.969336075376521e-05, + "loss": 0.4762, + "step": 722 + }, + { + "epoch": 0.5141333333333333, + "grad_norm": 1.4352731704711914, + "learning_rate": 1.9691323684864077e-05, + "loss": 0.5107, + "step": 723 + }, + { + "epoch": 0.5148444444444444, + "grad_norm": 1.5125515460968018, + "learning_rate": 1.9689279978067427e-05, + "loss": 0.444, + "step": 724 + }, + { + "epoch": 0.5155555555555555, + "grad_norm": 1.719496250152588, + "learning_rate": 1.9687229634775062e-05, + "loss": 0.4587, + "step": 725 + }, + { + "epoch": 0.5162666666666667, + "grad_norm": 1.7198173999786377, + "learning_rate": 1.9685172656391326e-05, + "loss": 0.5565, + "step": 726 + }, + { + "epoch": 0.5169777777777778, + "grad_norm": 1.6755539178848267, + "learning_rate": 1.9683109044325108e-05, + "loss": 0.4673, + "step": 727 + }, + { + "epoch": 0.5176888888888889, + "grad_norm": 1.7488329410552979, + "learning_rate": 1.9681038799989843e-05, + "loss": 0.4775, + "step": 728 + }, + { + "epoch": 0.5184, + "grad_norm": 1.6711724996566772, + "learning_rate": 1.9678961924803507e-05, + "loss": 0.4939, + "step": 729 + }, + { + "epoch": 0.5191111111111111, + "grad_norm": 1.7405672073364258, + "learning_rate": 1.9676878420188618e-05, + "loss": 0.4491, + "step": 730 + }, + { + "epoch": 0.5198222222222222, + "grad_norm": 1.714692234992981, + "learning_rate": 1.967478828757223e-05, + "loss": 0.5024, + "step": 731 + }, + { + "epoch": 0.5205333333333333, + "grad_norm": 1.3055658340454102, + "learning_rate": 1.9672691528385947e-05, + "loss": 0.4586, + "step": 732 + }, + { + "epoch": 0.5212444444444444, + "grad_norm": 1.768305778503418, + "learning_rate": 1.9670588144065906e-05, + "loss": 0.4738, + "step": 733 + }, + { + "epoch": 0.5219555555555555, + "grad_norm": 1.825606107711792, + "learning_rate": 1.9668478136052776e-05, + "loss": 0.4502, + "step": 734 + }, + { + "epoch": 0.5226666666666666, + "grad_norm": 1.5163905620574951, + "learning_rate": 1.9666361505791774e-05, + "loss": 0.5128, + "step": 735 + }, + { + "epoch": 0.5233777777777778, + "grad_norm": 1.931495189666748, + "learning_rate": 1.9664238254732648e-05, + "loss": 0.4616, + "step": 736 + }, + { + "epoch": 0.5240888888888889, + "grad_norm": 1.4629310369491577, + "learning_rate": 1.9662108384329675e-05, + "loss": 0.5109, + "step": 737 + }, + { + "epoch": 0.5248, + "grad_norm": 1.632228970527649, + "learning_rate": 1.9659971896041673e-05, + "loss": 0.4942, + "step": 738 + }, + { + "epoch": 0.5255111111111112, + "grad_norm": 1.6017383337020874, + "learning_rate": 1.9657828791331993e-05, + "loss": 0.4986, + "step": 739 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 1.5995383262634277, + "learning_rate": 1.9655679071668516e-05, + "loss": 0.4552, + "step": 740 + }, + { + "epoch": 0.5269333333333334, + "grad_norm": 1.4951246976852417, + "learning_rate": 1.9653522738523647e-05, + "loss": 0.4885, + "step": 741 + }, + { + "epoch": 0.5276444444444445, + "grad_norm": 1.663974642753601, + "learning_rate": 1.9651359793374332e-05, + "loss": 0.4585, + "step": 742 + }, + { + "epoch": 0.5283555555555556, + "grad_norm": 3.838059663772583, + "learning_rate": 1.9649190237702044e-05, + "loss": 0.4927, + "step": 743 + }, + { + "epoch": 0.5290666666666667, + "grad_norm": 1.6570461988449097, + "learning_rate": 1.9647014072992775e-05, + "loss": 0.4588, + "step": 744 + }, + { + "epoch": 0.5297777777777778, + "grad_norm": 1.1205484867095947, + "learning_rate": 1.964483130073705e-05, + "loss": 0.4314, + "step": 745 + }, + { + "epoch": 0.5304888888888889, + "grad_norm": 1.745342493057251, + "learning_rate": 1.9642641922429918e-05, + "loss": 0.4628, + "step": 746 + }, + { + "epoch": 0.5312, + "grad_norm": 1.9148955345153809, + "learning_rate": 1.964044593957096e-05, + "loss": 0.4214, + "step": 747 + }, + { + "epoch": 0.5319111111111111, + "grad_norm": 1.2199004888534546, + "learning_rate": 1.9638243353664267e-05, + "loss": 0.4845, + "step": 748 + }, + { + "epoch": 0.5326222222222222, + "grad_norm": 1.5879738330841064, + "learning_rate": 1.9636034166218465e-05, + "loss": 0.4344, + "step": 749 + }, + { + "epoch": 0.5333333333333333, + "grad_norm": 1.3638322353363037, + "learning_rate": 1.9633818378746694e-05, + "loss": 0.4875, + "step": 750 + }, + { + "epoch": 0.5340444444444444, + "grad_norm": 1.8749130964279175, + "learning_rate": 1.963159599276662e-05, + "loss": 0.4895, + "step": 751 + }, + { + "epoch": 0.5347555555555555, + "grad_norm": 2.122419595718384, + "learning_rate": 1.962936700980042e-05, + "loss": 0.4651, + "step": 752 + }, + { + "epoch": 0.5354666666666666, + "grad_norm": 1.4296207427978516, + "learning_rate": 1.9627131431374807e-05, + "loss": 0.4579, + "step": 753 + }, + { + "epoch": 0.5361777777777778, + "grad_norm": 1.8916964530944824, + "learning_rate": 1.9624889259020987e-05, + "loss": 0.4935, + "step": 754 + }, + { + "epoch": 0.5368888888888889, + "grad_norm": 1.980924129486084, + "learning_rate": 1.96226404942747e-05, + "loss": 0.4449, + "step": 755 + }, + { + "epoch": 0.5376, + "grad_norm": 2.5847346782684326, + "learning_rate": 1.96203851386762e-05, + "loss": 0.4587, + "step": 756 + }, + { + "epoch": 0.5383111111111111, + "grad_norm": 1.9231551885604858, + "learning_rate": 1.9618123193770243e-05, + "loss": 0.4962, + "step": 757 + }, + { + "epoch": 0.5390222222222222, + "grad_norm": 3.63643217086792, + "learning_rate": 1.9615854661106116e-05, + "loss": 0.4562, + "step": 758 + }, + { + "epoch": 0.5397333333333333, + "grad_norm": 3.399381637573242, + "learning_rate": 1.96135795422376e-05, + "loss": 0.4975, + "step": 759 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 1.8508740663528442, + "learning_rate": 1.961129783872301e-05, + "loss": 0.4761, + "step": 760 + }, + { + "epoch": 0.5411555555555555, + "grad_norm": 2.1600069999694824, + "learning_rate": 1.960900955212514e-05, + "loss": 0.4816, + "step": 761 + }, + { + "epoch": 0.5418666666666667, + "grad_norm": 2.308814525604248, + "learning_rate": 1.9606714684011328e-05, + "loss": 0.4669, + "step": 762 + }, + { + "epoch": 0.5425777777777778, + "grad_norm": 1.2682002782821655, + "learning_rate": 1.9604413235953388e-05, + "loss": 0.4637, + "step": 763 + }, + { + "epoch": 0.5432888888888889, + "grad_norm": 2.299311637878418, + "learning_rate": 1.960210520952766e-05, + "loss": 0.4392, + "step": 764 + }, + { + "epoch": 0.544, + "grad_norm": 2.1286842823028564, + "learning_rate": 1.9599790606314984e-05, + "loss": 0.5065, + "step": 765 + }, + { + "epoch": 0.5447111111111111, + "grad_norm": 3.194489002227783, + "learning_rate": 1.9597469427900704e-05, + "loss": 0.4534, + "step": 766 + }, + { + "epoch": 0.5454222222222223, + "grad_norm": 2.469780921936035, + "learning_rate": 1.9595141675874677e-05, + "loss": 0.3903, + "step": 767 + }, + { + "epoch": 0.5461333333333334, + "grad_norm": 2.312621593475342, + "learning_rate": 1.9592807351831244e-05, + "loss": 0.4354, + "step": 768 + }, + { + "epoch": 0.5468444444444445, + "grad_norm": 2.5844264030456543, + "learning_rate": 1.9590466457369262e-05, + "loss": 0.4941, + "step": 769 + }, + { + "epoch": 0.5475555555555556, + "grad_norm": 1.9075970649719238, + "learning_rate": 1.9588118994092086e-05, + "loss": 0.4426, + "step": 770 + }, + { + "epoch": 0.5482666666666667, + "grad_norm": 1.335552453994751, + "learning_rate": 1.958576496360756e-05, + "loss": 0.4653, + "step": 771 + }, + { + "epoch": 0.5489777777777778, + "grad_norm": 1.8688048124313354, + "learning_rate": 1.9583404367528044e-05, + "loss": 0.4842, + "step": 772 + }, + { + "epoch": 0.5496888888888889, + "grad_norm": 1.1013209819793701, + "learning_rate": 1.9581037207470382e-05, + "loss": 0.4264, + "step": 773 + }, + { + "epoch": 0.5504, + "grad_norm": 1.7236747741699219, + "learning_rate": 1.957866348505592e-05, + "loss": 0.4364, + "step": 774 + }, + { + "epoch": 0.5511111111111111, + "grad_norm": 1.7861499786376953, + "learning_rate": 1.9576283201910487e-05, + "loss": 0.457, + "step": 775 + }, + { + "epoch": 0.5518222222222222, + "grad_norm": 1.6979371309280396, + "learning_rate": 1.957389635966442e-05, + "loss": 0.4439, + "step": 776 + }, + { + "epoch": 0.5525333333333333, + "grad_norm": 2.4139387607574463, + "learning_rate": 1.9571502959952545e-05, + "loss": 0.4258, + "step": 777 + }, + { + "epoch": 0.5532444444444444, + "grad_norm": 1.559971809387207, + "learning_rate": 1.9569103004414174e-05, + "loss": 0.455, + "step": 778 + }, + { + "epoch": 0.5539555555555555, + "grad_norm": 1.7602064609527588, + "learning_rate": 1.9566696494693116e-05, + "loss": 0.4428, + "step": 779 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 1.927603006362915, + "learning_rate": 1.9564283432437664e-05, + "loss": 0.4777, + "step": 780 + }, + { + "epoch": 0.5553777777777777, + "grad_norm": 1.7086844444274902, + "learning_rate": 1.9561863819300603e-05, + "loss": 0.4418, + "step": 781 + }, + { + "epoch": 0.5560888888888889, + "grad_norm": 1.3004646301269531, + "learning_rate": 1.9559437656939197e-05, + "loss": 0.4425, + "step": 782 + }, + { + "epoch": 0.5568, + "grad_norm": 1.7352361679077148, + "learning_rate": 1.955700494701521e-05, + "loss": 0.4519, + "step": 783 + }, + { + "epoch": 0.5575111111111111, + "grad_norm": 1.731749415397644, + "learning_rate": 1.9554565691194875e-05, + "loss": 0.47, + "step": 784 + }, + { + "epoch": 0.5582222222222222, + "grad_norm": 1.9568352699279785, + "learning_rate": 1.955211989114892e-05, + "loss": 0.4445, + "step": 785 + }, + { + "epoch": 0.5589333333333333, + "grad_norm": 1.4895111322402954, + "learning_rate": 1.9549667548552557e-05, + "loss": 0.5101, + "step": 786 + }, + { + "epoch": 0.5596444444444445, + "grad_norm": 2.313359022140503, + "learning_rate": 1.954720866508546e-05, + "loss": 0.4331, + "step": 787 + }, + { + "epoch": 0.5603555555555556, + "grad_norm": 1.8560254573822021, + "learning_rate": 1.9544743242431804e-05, + "loss": 0.4601, + "step": 788 + }, + { + "epoch": 0.5610666666666667, + "grad_norm": 1.8457664251327515, + "learning_rate": 1.9542271282280233e-05, + "loss": 0.4749, + "step": 789 + }, + { + "epoch": 0.5617777777777778, + "grad_norm": 1.7003227472305298, + "learning_rate": 1.9539792786323874e-05, + "loss": 0.4387, + "step": 790 + }, + { + "epoch": 0.5624888888888889, + "grad_norm": 1.2294131517410278, + "learning_rate": 1.9537307756260325e-05, + "loss": 0.4794, + "step": 791 + }, + { + "epoch": 0.5632, + "grad_norm": 1.3084807395935059, + "learning_rate": 1.9534816193791664e-05, + "loss": 0.5007, + "step": 792 + }, + { + "epoch": 0.5639111111111111, + "grad_norm": 1.5658988952636719, + "learning_rate": 1.953231810062444e-05, + "loss": 0.443, + "step": 793 + }, + { + "epoch": 0.5646222222222222, + "grad_norm": 1.6942716836929321, + "learning_rate": 1.952981347846968e-05, + "loss": 0.492, + "step": 794 + }, + { + "epoch": 0.5653333333333334, + "grad_norm": 1.4663857221603394, + "learning_rate": 1.952730232904287e-05, + "loss": 0.404, + "step": 795 + }, + { + "epoch": 0.5660444444444445, + "grad_norm": 1.3431514501571655, + "learning_rate": 1.9524784654063988e-05, + "loss": 0.5002, + "step": 796 + }, + { + "epoch": 0.5667555555555556, + "grad_norm": 1.972238302230835, + "learning_rate": 1.952226045525746e-05, + "loss": 0.4784, + "step": 797 + }, + { + "epoch": 0.5674666666666667, + "grad_norm": 1.1562806367874146, + "learning_rate": 1.9519729734352196e-05, + "loss": 0.4207, + "step": 798 + }, + { + "epoch": 0.5681777777777778, + "grad_norm": 1.4820719957351685, + "learning_rate": 1.9517192493081563e-05, + "loss": 0.3909, + "step": 799 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 1.7653021812438965, + "learning_rate": 1.95146487331834e-05, + "loss": 0.5032, + "step": 800 + }, + { + "epoch": 0.5696, + "grad_norm": 1.094610333442688, + "learning_rate": 1.9512098456400014e-05, + "loss": 0.3929, + "step": 801 + }, + { + "epoch": 0.5703111111111111, + "grad_norm": 1.5147162675857544, + "learning_rate": 1.950954166447816e-05, + "loss": 0.4252, + "step": 802 + }, + { + "epoch": 0.5710222222222222, + "grad_norm": 1.3646975755691528, + "learning_rate": 1.9506978359169073e-05, + "loss": 0.4652, + "step": 803 + }, + { + "epoch": 0.5717333333333333, + "grad_norm": 1.617104411125183, + "learning_rate": 1.950440854222844e-05, + "loss": 0.4428, + "step": 804 + }, + { + "epoch": 0.5724444444444444, + "grad_norm": 1.418869972229004, + "learning_rate": 1.950183221541641e-05, + "loss": 0.4622, + "step": 805 + }, + { + "epoch": 0.5731555555555555, + "grad_norm": 2.4287898540496826, + "learning_rate": 1.9499249380497597e-05, + "loss": 0.4666, + "step": 806 + }, + { + "epoch": 0.5738666666666666, + "grad_norm": 1.583718180656433, + "learning_rate": 1.9496660039241058e-05, + "loss": 0.4432, + "step": 807 + }, + { + "epoch": 0.5745777777777777, + "grad_norm": 2.139763116836548, + "learning_rate": 1.949406419342032e-05, + "loss": 0.4449, + "step": 808 + }, + { + "epoch": 0.5752888888888888, + "grad_norm": 1.3066848516464233, + "learning_rate": 1.949146184481336e-05, + "loss": 0.475, + "step": 809 + }, + { + "epoch": 0.576, + "grad_norm": 2.056758403778076, + "learning_rate": 1.948885299520261e-05, + "loss": 0.478, + "step": 810 + }, + { + "epoch": 0.5767111111111111, + "grad_norm": 1.5637251138687134, + "learning_rate": 1.948623764637495e-05, + "loss": 0.4709, + "step": 811 + }, + { + "epoch": 0.5774222222222222, + "grad_norm": 1.8826402425765991, + "learning_rate": 1.9483615800121717e-05, + "loss": 0.4288, + "step": 812 + }, + { + "epoch": 0.5781333333333334, + "grad_norm": 1.6989904642105103, + "learning_rate": 1.94809874582387e-05, + "loss": 0.4215, + "step": 813 + }, + { + "epoch": 0.5788444444444445, + "grad_norm": 2.341240406036377, + "learning_rate": 1.947835262252613e-05, + "loss": 0.4657, + "step": 814 + }, + { + "epoch": 0.5795555555555556, + "grad_norm": 2.4367377758026123, + "learning_rate": 1.9475711294788693e-05, + "loss": 0.4585, + "step": 815 + }, + { + "epoch": 0.5802666666666667, + "grad_norm": 1.5105916261672974, + "learning_rate": 1.9473063476835518e-05, + "loss": 0.48, + "step": 816 + }, + { + "epoch": 0.5809777777777778, + "grad_norm": 2.2668662071228027, + "learning_rate": 1.9470409170480177e-05, + "loss": 0.5379, + "step": 817 + }, + { + "epoch": 0.5816888888888889, + "grad_norm": 1.82729971408844, + "learning_rate": 1.946774837754069e-05, + "loss": 0.4159, + "step": 818 + }, + { + "epoch": 0.5824, + "grad_norm": 2.493115186691284, + "learning_rate": 1.9465081099839515e-05, + "loss": 0.4597, + "step": 819 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 2.144582509994507, + "learning_rate": 1.946240733920356e-05, + "loss": 0.456, + "step": 820 + }, + { + "epoch": 0.5838222222222222, + "grad_norm": 1.660337209701538, + "learning_rate": 1.945972709746417e-05, + "loss": 0.4881, + "step": 821 + }, + { + "epoch": 0.5845333333333333, + "grad_norm": 2.153820276260376, + "learning_rate": 1.945704037645713e-05, + "loss": 0.4595, + "step": 822 + }, + { + "epoch": 0.5852444444444445, + "grad_norm": 1.123346209526062, + "learning_rate": 1.945434717802265e-05, + "loss": 0.4271, + "step": 823 + }, + { + "epoch": 0.5859555555555556, + "grad_norm": 1.775033950805664, + "learning_rate": 1.9451647504005394e-05, + "loss": 0.4288, + "step": 824 + }, + { + "epoch": 0.5866666666666667, + "grad_norm": 1.4474109411239624, + "learning_rate": 1.9448941356254453e-05, + "loss": 0.4459, + "step": 825 + }, + { + "epoch": 0.5873777777777778, + "grad_norm": 1.958325743675232, + "learning_rate": 1.9446228736623355e-05, + "loss": 0.4765, + "step": 826 + }, + { + "epoch": 0.5880888888888889, + "grad_norm": 1.5237672328948975, + "learning_rate": 1.944350964697006e-05, + "loss": 0.5029, + "step": 827 + }, + { + "epoch": 0.5888, + "grad_norm": 1.5609651803970337, + "learning_rate": 1.9440784089156955e-05, + "loss": 0.4606, + "step": 828 + }, + { + "epoch": 0.5895111111111111, + "grad_norm": 1.8578894138336182, + "learning_rate": 1.9438052065050865e-05, + "loss": 0.4573, + "step": 829 + }, + { + "epoch": 0.5902222222222222, + "grad_norm": 2.7874748706817627, + "learning_rate": 1.9435313576523037e-05, + "loss": 0.4419, + "step": 830 + }, + { + "epoch": 0.5909333333333333, + "grad_norm": 2.4316794872283936, + "learning_rate": 1.9432568625449152e-05, + "loss": 0.4472, + "step": 831 + }, + { + "epoch": 0.5916444444444444, + "grad_norm": 2.0532658100128174, + "learning_rate": 1.942981721370931e-05, + "loss": 0.4485, + "step": 832 + }, + { + "epoch": 0.5923555555555555, + "grad_norm": 1.96194589138031, + "learning_rate": 1.9427059343188047e-05, + "loss": 0.4849, + "step": 833 + }, + { + "epoch": 0.5930666666666666, + "grad_norm": 1.9596260786056519, + "learning_rate": 1.942429501577431e-05, + "loss": 0.4362, + "step": 834 + }, + { + "epoch": 0.5937777777777777, + "grad_norm": 2.413278579711914, + "learning_rate": 1.942152423336148e-05, + "loss": 0.4294, + "step": 835 + }, + { + "epoch": 0.5944888888888888, + "grad_norm": 1.996556043624878, + "learning_rate": 1.9418746997847347e-05, + "loss": 0.4661, + "step": 836 + }, + { + "epoch": 0.5952, + "grad_norm": 2.076242208480835, + "learning_rate": 1.9415963311134135e-05, + "loss": 0.4287, + "step": 837 + }, + { + "epoch": 0.5959111111111111, + "grad_norm": 1.9697989225387573, + "learning_rate": 1.9413173175128472e-05, + "loss": 0.4834, + "step": 838 + }, + { + "epoch": 0.5966222222222223, + "grad_norm": 1.7743202447891235, + "learning_rate": 1.941037659174142e-05, + "loss": 0.4447, + "step": 839 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 2.5803024768829346, + "learning_rate": 1.9407573562888437e-05, + "loss": 0.4111, + "step": 840 + }, + { + "epoch": 0.5980444444444445, + "grad_norm": 2.189613103866577, + "learning_rate": 1.9404764090489413e-05, + "loss": 0.4597, + "step": 841 + }, + { + "epoch": 0.5987555555555556, + "grad_norm": 2.3707635402679443, + "learning_rate": 1.9401948176468645e-05, + "loss": 0.445, + "step": 842 + }, + { + "epoch": 0.5994666666666667, + "grad_norm": 2.33351731300354, + "learning_rate": 1.9399125822754845e-05, + "loss": 0.4673, + "step": 843 + }, + { + "epoch": 0.6001777777777778, + "grad_norm": 1.930680751800537, + "learning_rate": 1.9396297031281124e-05, + "loss": 0.4196, + "step": 844 + }, + { + "epoch": 0.6008888888888889, + "grad_norm": 1.7494875192642212, + "learning_rate": 1.939346180398502e-05, + "loss": 0.4288, + "step": 845 + }, + { + "epoch": 0.6016, + "grad_norm": 1.5277602672576904, + "learning_rate": 1.9390620142808462e-05, + "loss": 0.4343, + "step": 846 + }, + { + "epoch": 0.6023111111111111, + "grad_norm": 1.7591135501861572, + "learning_rate": 1.9387772049697803e-05, + "loss": 0.4995, + "step": 847 + }, + { + "epoch": 0.6030222222222222, + "grad_norm": 2.5099804401397705, + "learning_rate": 1.9384917526603783e-05, + "loss": 0.4286, + "step": 848 + }, + { + "epoch": 0.6037333333333333, + "grad_norm": 1.6381641626358032, + "learning_rate": 1.9382056575481565e-05, + "loss": 0.4628, + "step": 849 + }, + { + "epoch": 0.6044444444444445, + "grad_norm": 2.7657461166381836, + "learning_rate": 1.9379189198290697e-05, + "loss": 0.4635, + "step": 850 + }, + { + "epoch": 0.6051555555555556, + "grad_norm": 2.8177220821380615, + "learning_rate": 1.937631539699514e-05, + "loss": 0.4459, + "step": 851 + }, + { + "epoch": 0.6058666666666667, + "grad_norm": 2.3989200592041016, + "learning_rate": 1.9373435173563257e-05, + "loss": 0.4341, + "step": 852 + }, + { + "epoch": 0.6065777777777778, + "grad_norm": 2.4620869159698486, + "learning_rate": 1.9370548529967795e-05, + "loss": 0.4633, + "step": 853 + }, + { + "epoch": 0.6072888888888889, + "grad_norm": 1.4641337394714355, + "learning_rate": 1.9367655468185913e-05, + "loss": 0.5015, + "step": 854 + }, + { + "epoch": 0.608, + "grad_norm": 1.6383060216903687, + "learning_rate": 1.9364755990199162e-05, + "loss": 0.4292, + "step": 855 + }, + { + "epoch": 0.6087111111111111, + "grad_norm": 1.447767972946167, + "learning_rate": 1.9361850097993487e-05, + "loss": 0.4793, + "step": 856 + }, + { + "epoch": 0.6094222222222222, + "grad_norm": 1.5150604248046875, + "learning_rate": 1.9358937793559222e-05, + "loss": 0.4411, + "step": 857 + }, + { + "epoch": 0.6101333333333333, + "grad_norm": 1.715248465538025, + "learning_rate": 1.9356019078891098e-05, + "loss": 0.3821, + "step": 858 + }, + { + "epoch": 0.6108444444444444, + "grad_norm": 1.4054685831069946, + "learning_rate": 1.935309395598824e-05, + "loss": 0.4388, + "step": 859 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 2.3306474685668945, + "learning_rate": 1.9350162426854152e-05, + "loss": 0.4701, + "step": 860 + }, + { + "epoch": 0.6122666666666666, + "grad_norm": 1.4369913339614868, + "learning_rate": 1.934722449349673e-05, + "loss": 0.4108, + "step": 861 + }, + { + "epoch": 0.6129777777777777, + "grad_norm": 2.7008185386657715, + "learning_rate": 1.9344280157928265e-05, + "loss": 0.4246, + "step": 862 + }, + { + "epoch": 0.6136888888888888, + "grad_norm": 2.294527292251587, + "learning_rate": 1.9341329422165422e-05, + "loss": 0.42, + "step": 863 + }, + { + "epoch": 0.6144, + "grad_norm": 2.661700487136841, + "learning_rate": 1.9338372288229253e-05, + "loss": 0.4536, + "step": 864 + }, + { + "epoch": 0.6151111111111112, + "grad_norm": 1.9521516561508179, + "learning_rate": 1.9335408758145194e-05, + "loss": 0.4566, + "step": 865 + }, + { + "epoch": 0.6158222222222223, + "grad_norm": 1.8434220552444458, + "learning_rate": 1.9332438833943065e-05, + "loss": 0.425, + "step": 866 + }, + { + "epoch": 0.6165333333333334, + "grad_norm": 1.7006642818450928, + "learning_rate": 1.932946251765705e-05, + "loss": 0.4022, + "step": 867 + }, + { + "epoch": 0.6172444444444445, + "grad_norm": 1.8852732181549072, + "learning_rate": 1.9326479811325736e-05, + "loss": 0.4438, + "step": 868 + }, + { + "epoch": 0.6179555555555556, + "grad_norm": 1.7331607341766357, + "learning_rate": 1.9323490716992067e-05, + "loss": 0.4327, + "step": 869 + }, + { + "epoch": 0.6186666666666667, + "grad_norm": 2.475703239440918, + "learning_rate": 1.9320495236703366e-05, + "loss": 0.4369, + "step": 870 + }, + { + "epoch": 0.6193777777777778, + "grad_norm": 1.9660125970840454, + "learning_rate": 1.9317493372511335e-05, + "loss": 0.423, + "step": 871 + }, + { + "epoch": 0.6200888888888889, + "grad_norm": 1.8893154859542847, + "learning_rate": 1.931448512647205e-05, + "loss": 0.4152, + "step": 872 + }, + { + "epoch": 0.6208, + "grad_norm": 2.011042356491089, + "learning_rate": 1.9311470500645946e-05, + "loss": 0.4174, + "step": 873 + }, + { + "epoch": 0.6215111111111111, + "grad_norm": 1.7617449760437012, + "learning_rate": 1.9308449497097847e-05, + "loss": 0.4694, + "step": 874 + }, + { + "epoch": 0.6222222222222222, + "grad_norm": 2.0529448986053467, + "learning_rate": 1.9305422117896927e-05, + "loss": 0.4577, + "step": 875 + }, + { + "epoch": 0.6229333333333333, + "grad_norm": 2.244443893432617, + "learning_rate": 1.9302388365116734e-05, + "loss": 0.4501, + "step": 876 + }, + { + "epoch": 0.6236444444444444, + "grad_norm": 1.9051746129989624, + "learning_rate": 1.9299348240835182e-05, + "loss": 0.4597, + "step": 877 + }, + { + "epoch": 0.6243555555555556, + "grad_norm": 2.227942943572998, + "learning_rate": 1.9296301747134555e-05, + "loss": 0.4283, + "step": 878 + }, + { + "epoch": 0.6250666666666667, + "grad_norm": 2.2462565898895264, + "learning_rate": 1.929324888610148e-05, + "loss": 0.4798, + "step": 879 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 1.4855154752731323, + "learning_rate": 1.9290189659826974e-05, + "loss": 0.4484, + "step": 880 + }, + { + "epoch": 0.6264888888888889, + "grad_norm": 1.8193817138671875, + "learning_rate": 1.9287124070406388e-05, + "loss": 0.4572, + "step": 881 + }, + { + "epoch": 0.6272, + "grad_norm": 1.4335157871246338, + "learning_rate": 1.9284052119939448e-05, + "loss": 0.4576, + "step": 882 + }, + { + "epoch": 0.6279111111111111, + "grad_norm": 1.762441873550415, + "learning_rate": 1.9280973810530225e-05, + "loss": 0.4541, + "step": 883 + }, + { + "epoch": 0.6286222222222222, + "grad_norm": 1.4353560209274292, + "learning_rate": 1.927788914428715e-05, + "loss": 0.4088, + "step": 884 + }, + { + "epoch": 0.6293333333333333, + "grad_norm": 1.4797428846359253, + "learning_rate": 1.9274798123323015e-05, + "loss": 0.4445, + "step": 885 + }, + { + "epoch": 0.6300444444444444, + "grad_norm": 1.7261810302734375, + "learning_rate": 1.9271700749754954e-05, + "loss": 0.438, + "step": 886 + }, + { + "epoch": 0.6307555555555555, + "grad_norm": 1.579777717590332, + "learning_rate": 1.9268597025704464e-05, + "loss": 0.4811, + "step": 887 + }, + { + "epoch": 0.6314666666666666, + "grad_norm": 1.8855563402175903, + "learning_rate": 1.9265486953297373e-05, + "loss": 0.4822, + "step": 888 + }, + { + "epoch": 0.6321777777777777, + "grad_norm": 2.176384925842285, + "learning_rate": 1.9262370534663874e-05, + "loss": 0.4075, + "step": 889 + }, + { + "epoch": 0.6328888888888888, + "grad_norm": 1.5426148176193237, + "learning_rate": 1.92592477719385e-05, + "loss": 0.4245, + "step": 890 + }, + { + "epoch": 0.6336, + "grad_norm": 2.021475315093994, + "learning_rate": 1.9256118667260134e-05, + "loss": 0.4001, + "step": 891 + }, + { + "epoch": 0.6343111111111112, + "grad_norm": 1.7395906448364258, + "learning_rate": 1.9252983222771996e-05, + "loss": 0.4037, + "step": 892 + }, + { + "epoch": 0.6350222222222223, + "grad_norm": 2.6725871562957764, + "learning_rate": 1.924984144062165e-05, + "loss": 0.4135, + "step": 893 + }, + { + "epoch": 0.6357333333333334, + "grad_norm": 2.0060768127441406, + "learning_rate": 1.9246693322961002e-05, + "loss": 0.4615, + "step": 894 + }, + { + "epoch": 0.6364444444444445, + "grad_norm": 1.893784999847412, + "learning_rate": 1.9243538871946304e-05, + "loss": 0.3953, + "step": 895 + }, + { + "epoch": 0.6371555555555556, + "grad_norm": 1.6854790449142456, + "learning_rate": 1.9240378089738136e-05, + "loss": 0.4278, + "step": 896 + }, + { + "epoch": 0.6378666666666667, + "grad_norm": 1.9277033805847168, + "learning_rate": 1.923721097850142e-05, + "loss": 0.4438, + "step": 897 + }, + { + "epoch": 0.6385777777777778, + "grad_norm": 1.7453292608261108, + "learning_rate": 1.9234037540405405e-05, + "loss": 0.4288, + "step": 898 + }, + { + "epoch": 0.6392888888888889, + "grad_norm": 1.0124443769454956, + "learning_rate": 1.9230857777623684e-05, + "loss": 0.4447, + "step": 899 + }, + { + "epoch": 0.64, + "grad_norm": 1.6840322017669678, + "learning_rate": 1.922767169233418e-05, + "loss": 0.4578, + "step": 900 + }, + { + "epoch": 0.6407111111111111, + "grad_norm": 1.3345437049865723, + "learning_rate": 1.9224479286719143e-05, + "loss": 0.4474, + "step": 901 + }, + { + "epoch": 0.6414222222222222, + "grad_norm": 2.176096200942993, + "learning_rate": 1.9221280562965145e-05, + "loss": 0.4004, + "step": 902 + }, + { + "epoch": 0.6421333333333333, + "grad_norm": 1.5793989896774292, + "learning_rate": 1.9218075523263104e-05, + "loss": 0.4332, + "step": 903 + }, + { + "epoch": 0.6428444444444444, + "grad_norm": 1.7723636627197266, + "learning_rate": 1.9214864169808252e-05, + "loss": 0.4103, + "step": 904 + }, + { + "epoch": 0.6435555555555555, + "grad_norm": 2.10990309715271, + "learning_rate": 1.921164650480014e-05, + "loss": 0.4035, + "step": 905 + }, + { + "epoch": 0.6442666666666667, + "grad_norm": 1.7074127197265625, + "learning_rate": 1.9208422530442658e-05, + "loss": 0.4887, + "step": 906 + }, + { + "epoch": 0.6449777777777778, + "grad_norm": 2.368265390396118, + "learning_rate": 1.9205192248944e-05, + "loss": 0.4433, + "step": 907 + }, + { + "epoch": 0.6456888888888889, + "grad_norm": 1.8453022241592407, + "learning_rate": 1.9201955662516693e-05, + "loss": 0.4348, + "step": 908 + }, + { + "epoch": 0.6464, + "grad_norm": 2.074856758117676, + "learning_rate": 1.919871277337758e-05, + "loss": 0.4784, + "step": 909 + }, + { + "epoch": 0.6471111111111111, + "grad_norm": 2.4104316234588623, + "learning_rate": 1.9195463583747814e-05, + "loss": 0.4287, + "step": 910 + }, + { + "epoch": 0.6478222222222222, + "grad_norm": 1.402001976966858, + "learning_rate": 1.919220809585287e-05, + "loss": 0.4434, + "step": 911 + }, + { + "epoch": 0.6485333333333333, + "grad_norm": 2.2844417095184326, + "learning_rate": 1.9188946311922537e-05, + "loss": 0.4073, + "step": 912 + }, + { + "epoch": 0.6492444444444444, + "grad_norm": 1.7766413688659668, + "learning_rate": 1.9185678234190913e-05, + "loss": 0.3975, + "step": 913 + }, + { + "epoch": 0.6499555555555555, + "grad_norm": 1.4851187467575073, + "learning_rate": 1.9182403864896407e-05, + "loss": 0.4479, + "step": 914 + }, + { + "epoch": 0.6506666666666666, + "grad_norm": 1.996512532234192, + "learning_rate": 1.917912320628174e-05, + "loss": 0.4252, + "step": 915 + }, + { + "epoch": 0.6513777777777778, + "grad_norm": 1.3163974285125732, + "learning_rate": 1.9175836260593937e-05, + "loss": 0.4133, + "step": 916 + }, + { + "epoch": 0.6520888888888889, + "grad_norm": 2.029017925262451, + "learning_rate": 1.9172543030084338e-05, + "loss": 0.4323, + "step": 917 + }, + { + "epoch": 0.6528, + "grad_norm": 1.315398097038269, + "learning_rate": 1.9169243517008572e-05, + "loss": 0.444, + "step": 918 + }, + { + "epoch": 0.6535111111111112, + "grad_norm": 1.390018105506897, + "learning_rate": 1.916593772362659e-05, + "loss": 0.4201, + "step": 919 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 1.7951534986495972, + "learning_rate": 1.916262565220263e-05, + "loss": 0.4494, + "step": 920 + }, + { + "epoch": 0.6549333333333334, + "grad_norm": 1.1677720546722412, + "learning_rate": 1.915930730500523e-05, + "loss": 0.4109, + "step": 921 + }, + { + "epoch": 0.6556444444444445, + "grad_norm": 1.52410888671875, + "learning_rate": 1.915598268430724e-05, + "loss": 0.4797, + "step": 922 + }, + { + "epoch": 0.6563555555555556, + "grad_norm": 1.768891453742981, + "learning_rate": 1.9152651792385793e-05, + "loss": 0.411, + "step": 923 + }, + { + "epoch": 0.6570666666666667, + "grad_norm": 1.185375452041626, + "learning_rate": 1.9149314631522325e-05, + "loss": 0.3989, + "step": 924 + }, + { + "epoch": 0.6577777777777778, + "grad_norm": 2.615072727203369, + "learning_rate": 1.9145971204002566e-05, + "loss": 0.4412, + "step": 925 + }, + { + "epoch": 0.6584888888888889, + "grad_norm": 2.025893449783325, + "learning_rate": 1.914262151211653e-05, + "loss": 0.4347, + "step": 926 + }, + { + "epoch": 0.6592, + "grad_norm": 1.7542346715927124, + "learning_rate": 1.913926555815853e-05, + "loss": 0.4226, + "step": 927 + }, + { + "epoch": 0.6599111111111111, + "grad_norm": 1.8960744142532349, + "learning_rate": 1.9135903344427163e-05, + "loss": 0.4239, + "step": 928 + }, + { + "epoch": 0.6606222222222222, + "grad_norm": 2.8602285385131836, + "learning_rate": 1.9132534873225323e-05, + "loss": 0.459, + "step": 929 + }, + { + "epoch": 0.6613333333333333, + "grad_norm": 1.839680790901184, + "learning_rate": 1.912916014686018e-05, + "loss": 0.4893, + "step": 930 + }, + { + "epoch": 0.6620444444444444, + "grad_norm": 2.215729236602783, + "learning_rate": 1.9125779167643187e-05, + "loss": 0.4627, + "step": 931 + }, + { + "epoch": 0.6627555555555555, + "grad_norm": 2.2629570960998535, + "learning_rate": 1.912239193789009e-05, + "loss": 0.3997, + "step": 932 + }, + { + "epoch": 0.6634666666666666, + "grad_norm": 1.862696647644043, + "learning_rate": 1.91189984599209e-05, + "loss": 0.4149, + "step": 933 + }, + { + "epoch": 0.6641777777777778, + "grad_norm": 1.7599492073059082, + "learning_rate": 1.9115598736059934e-05, + "loss": 0.4298, + "step": 934 + }, + { + "epoch": 0.6648888888888889, + "grad_norm": 1.593726634979248, + "learning_rate": 1.9112192768635756e-05, + "loss": 0.3743, + "step": 935 + }, + { + "epoch": 0.6656, + "grad_norm": 1.4605954885482788, + "learning_rate": 1.910878055998123e-05, + "loss": 0.3774, + "step": 936 + }, + { + "epoch": 0.6663111111111111, + "grad_norm": 2.173814296722412, + "learning_rate": 1.9105362112433484e-05, + "loss": 0.4735, + "step": 937 + }, + { + "epoch": 0.6670222222222222, + "grad_norm": 1.4773623943328857, + "learning_rate": 1.9101937428333918e-05, + "loss": 0.4671, + "step": 938 + }, + { + "epoch": 0.6677333333333333, + "grad_norm": 2.286327838897705, + "learning_rate": 1.909850651002821e-05, + "loss": 0.4427, + "step": 939 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 1.6355377435684204, + "learning_rate": 1.909506935986631e-05, + "loss": 0.4576, + "step": 940 + }, + { + "epoch": 0.6691555555555555, + "grad_norm": 2.4499120712280273, + "learning_rate": 1.909162598020242e-05, + "loss": 0.4673, + "step": 941 + }, + { + "epoch": 0.6698666666666667, + "grad_norm": 1.7143490314483643, + "learning_rate": 1.908817637339503e-05, + "loss": 0.4748, + "step": 942 + }, + { + "epoch": 0.6705777777777778, + "grad_norm": 1.854475736618042, + "learning_rate": 1.9084720541806883e-05, + "loss": 0.4705, + "step": 943 + }, + { + "epoch": 0.6712888888888889, + "grad_norm": 1.5117998123168945, + "learning_rate": 1.908125848780498e-05, + "loss": 0.4678, + "step": 944 + }, + { + "epoch": 0.672, + "grad_norm": 2.2441301345825195, + "learning_rate": 1.9077790213760604e-05, + "loss": 0.4555, + "step": 945 + }, + { + "epoch": 0.6727111111111111, + "grad_norm": 1.8586398363113403, + "learning_rate": 1.9074315722049278e-05, + "loss": 0.4249, + "step": 946 + }, + { + "epoch": 0.6734222222222223, + "grad_norm": 2.2055532932281494, + "learning_rate": 1.9070835015050796e-05, + "loss": 0.4361, + "step": 947 + }, + { + "epoch": 0.6741333333333334, + "grad_norm": 1.7764239311218262, + "learning_rate": 1.90673480951492e-05, + "loss": 0.4726, + "step": 948 + }, + { + "epoch": 0.6748444444444445, + "grad_norm": 2.383765459060669, + "learning_rate": 1.9063854964732796e-05, + "loss": 0.4028, + "step": 949 + }, + { + "epoch": 0.6755555555555556, + "grad_norm": 2.156425952911377, + "learning_rate": 1.906035562619414e-05, + "loss": 0.5258, + "step": 950 + }, + { + "epoch": 0.6762666666666667, + "grad_norm": 1.8455520868301392, + "learning_rate": 1.905685008193004e-05, + "loss": 0.4638, + "step": 951 + }, + { + "epoch": 0.6769777777777778, + "grad_norm": 1.9742159843444824, + "learning_rate": 1.9053338334341553e-05, + "loss": 0.4581, + "step": 952 + }, + { + "epoch": 0.6776888888888889, + "grad_norm": 2.0623605251312256, + "learning_rate": 1.9049820385833987e-05, + "loss": 0.4045, + "step": 953 + }, + { + "epoch": 0.6784, + "grad_norm": 1.8264994621276855, + "learning_rate": 1.90462962388169e-05, + "loss": 0.4052, + "step": 954 + }, + { + "epoch": 0.6791111111111111, + "grad_norm": 2.335390567779541, + "learning_rate": 1.9042765895704085e-05, + "loss": 0.4239, + "step": 955 + }, + { + "epoch": 0.6798222222222222, + "grad_norm": 1.4708797931671143, + "learning_rate": 1.9039229358913594e-05, + "loss": 0.4363, + "step": 956 + }, + { + "epoch": 0.6805333333333333, + "grad_norm": 1.9882786273956299, + "learning_rate": 1.9035686630867706e-05, + "loss": 0.4208, + "step": 957 + }, + { + "epoch": 0.6812444444444444, + "grad_norm": 2.051907777786255, + "learning_rate": 1.9032137713992957e-05, + "loss": 0.4547, + "step": 958 + }, + { + "epoch": 0.6819555555555555, + "grad_norm": 1.9634442329406738, + "learning_rate": 1.9028582610720106e-05, + "loss": 0.4253, + "step": 959 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 2.4471778869628906, + "learning_rate": 1.9025021323484155e-05, + "loss": 0.4354, + "step": 960 + }, + { + "epoch": 0.6833777777777778, + "grad_norm": 1.745047688484192, + "learning_rate": 1.902145385472435e-05, + "loss": 0.4409, + "step": 961 + }, + { + "epoch": 0.6840888888888889, + "grad_norm": 1.9081916809082031, + "learning_rate": 1.9017880206884157e-05, + "loss": 0.4221, + "step": 962 + }, + { + "epoch": 0.6848, + "grad_norm": 1.5096672773361206, + "learning_rate": 1.9014300382411285e-05, + "loss": 0.4138, + "step": 963 + }, + { + "epoch": 0.6855111111111111, + "grad_norm": 1.3299483060836792, + "learning_rate": 1.9010714383757668e-05, + "loss": 0.4742, + "step": 964 + }, + { + "epoch": 0.6862222222222222, + "grad_norm": 1.9163708686828613, + "learning_rate": 1.900712221337947e-05, + "loss": 0.4075, + "step": 965 + }, + { + "epoch": 0.6869333333333333, + "grad_norm": 1.413456916809082, + "learning_rate": 1.9003523873737082e-05, + "loss": 0.4022, + "step": 966 + }, + { + "epoch": 0.6876444444444444, + "grad_norm": 2.050042152404785, + "learning_rate": 1.8999919367295122e-05, + "loss": 0.4154, + "step": 967 + }, + { + "epoch": 0.6883555555555556, + "grad_norm": 1.859709620475769, + "learning_rate": 1.8996308696522435e-05, + "loss": 0.4598, + "step": 968 + }, + { + "epoch": 0.6890666666666667, + "grad_norm": 1.269230604171753, + "learning_rate": 1.8992691863892078e-05, + "loss": 0.418, + "step": 969 + }, + { + "epoch": 0.6897777777777778, + "grad_norm": 1.8144010305404663, + "learning_rate": 1.8989068871881336e-05, + "loss": 0.4442, + "step": 970 + }, + { + "epoch": 0.6904888888888889, + "grad_norm": 1.8606047630310059, + "learning_rate": 1.8985439722971714e-05, + "loss": 0.4069, + "step": 971 + }, + { + "epoch": 0.6912, + "grad_norm": 1.3215489387512207, + "learning_rate": 1.898180441964893e-05, + "loss": 0.4363, + "step": 972 + }, + { + "epoch": 0.6919111111111111, + "grad_norm": 1.7768440246582031, + "learning_rate": 1.8978162964402918e-05, + "loss": 0.4069, + "step": 973 + }, + { + "epoch": 0.6926222222222223, + "grad_norm": 1.1064733266830444, + "learning_rate": 1.897451535972783e-05, + "loss": 0.4333, + "step": 974 + }, + { + "epoch": 0.6933333333333334, + "grad_norm": 1.669946312904358, + "learning_rate": 1.8970861608122017e-05, + "loss": 0.4447, + "step": 975 + }, + { + "epoch": 0.6940444444444445, + "grad_norm": 1.6347194910049438, + "learning_rate": 1.8967201712088056e-05, + "loss": 0.4385, + "step": 976 + }, + { + "epoch": 0.6947555555555556, + "grad_norm": 1.5063124895095825, + "learning_rate": 1.896353567413273e-05, + "loss": 0.457, + "step": 977 + }, + { + "epoch": 0.6954666666666667, + "grad_norm": 1.4115867614746094, + "learning_rate": 1.8959863496767022e-05, + "loss": 0.4101, + "step": 978 + }, + { + "epoch": 0.6961777777777778, + "grad_norm": 1.5335878133773804, + "learning_rate": 1.895618518250612e-05, + "loss": 0.3662, + "step": 979 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 1.502942442893982, + "learning_rate": 1.8952500733869415e-05, + "loss": 0.439, + "step": 980 + }, + { + "epoch": 0.6976, + "grad_norm": 1.5190778970718384, + "learning_rate": 1.8948810153380514e-05, + "loss": 0.4712, + "step": 981 + }, + { + "epoch": 0.6983111111111111, + "grad_norm": 1.1864885091781616, + "learning_rate": 1.8945113443567202e-05, + "loss": 0.4468, + "step": 982 + }, + { + "epoch": 0.6990222222222222, + "grad_norm": 1.8010454177856445, + "learning_rate": 1.894141060696148e-05, + "loss": 0.4636, + "step": 983 + }, + { + "epoch": 0.6997333333333333, + "grad_norm": 1.5485213994979858, + "learning_rate": 1.8937701646099537e-05, + "loss": 0.4625, + "step": 984 + }, + { + "epoch": 0.7004444444444444, + "grad_norm": 1.4470927715301514, + "learning_rate": 1.8933986563521757e-05, + "loss": 0.4356, + "step": 985 + }, + { + "epoch": 0.7011555555555555, + "grad_norm": 1.584431767463684, + "learning_rate": 1.893026536177272e-05, + "loss": 0.4111, + "step": 986 + }, + { + "epoch": 0.7018666666666666, + "grad_norm": 1.130873680114746, + "learning_rate": 1.8926538043401192e-05, + "loss": 0.4831, + "step": 987 + }, + { + "epoch": 0.7025777777777777, + "grad_norm": 1.3628320693969727, + "learning_rate": 1.8922804610960134e-05, + "loss": 0.4439, + "step": 988 + }, + { + "epoch": 0.7032888888888889, + "grad_norm": 1.4469481706619263, + "learning_rate": 1.8919065067006694e-05, + "loss": 0.459, + "step": 989 + }, + { + "epoch": 0.704, + "grad_norm": 1.561838984489441, + "learning_rate": 1.8915319414102197e-05, + "loss": 0.463, + "step": 990 + }, + { + "epoch": 0.7047111111111111, + "grad_norm": 1.3215110301971436, + "learning_rate": 1.891156765481217e-05, + "loss": 0.3927, + "step": 991 + }, + { + "epoch": 0.7054222222222222, + "grad_norm": 1.8927325010299683, + "learning_rate": 1.8907809791706304e-05, + "loss": 0.399, + "step": 992 + }, + { + "epoch": 0.7061333333333333, + "grad_norm": 1.3398423194885254, + "learning_rate": 1.8904045827358483e-05, + "loss": 0.4403, + "step": 993 + }, + { + "epoch": 0.7068444444444445, + "grad_norm": 1.7695531845092773, + "learning_rate": 1.890027576434677e-05, + "loss": 0.4879, + "step": 994 + }, + { + "epoch": 0.7075555555555556, + "grad_norm": 1.4437052011489868, + "learning_rate": 1.889649960525339e-05, + "loss": 0.4274, + "step": 995 + }, + { + "epoch": 0.7082666666666667, + "grad_norm": 1.5803529024124146, + "learning_rate": 1.8892717352664762e-05, + "loss": 0.4405, + "step": 996 + }, + { + "epoch": 0.7089777777777778, + "grad_norm": 1.7902425527572632, + "learning_rate": 1.888892900917147e-05, + "loss": 0.4159, + "step": 997 + }, + { + "epoch": 0.7096888888888889, + "grad_norm": 2.4440481662750244, + "learning_rate": 1.8885134577368268e-05, + "loss": 0.4555, + "step": 998 + }, + { + "epoch": 0.7104, + "grad_norm": 2.088907241821289, + "learning_rate": 1.8881334059854087e-05, + "loss": 0.4386, + "step": 999 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 1.4719020128250122, + "learning_rate": 1.887752745923202e-05, + "loss": 0.4117, + "step": 1000 + }, + { + "epoch": 0.7118222222222222, + "grad_norm": 1.991410255432129, + "learning_rate": 1.8873714778109328e-05, + "loss": 0.4627, + "step": 1001 + }, + { + "epoch": 0.7125333333333334, + "grad_norm": 2.5293521881103516, + "learning_rate": 1.886989601909744e-05, + "loss": 0.4666, + "step": 1002 + }, + { + "epoch": 0.7132444444444445, + "grad_norm": 1.5991255044937134, + "learning_rate": 1.8866071184811943e-05, + "loss": 0.4348, + "step": 1003 + }, + { + "epoch": 0.7139555555555556, + "grad_norm": 2.116929769515991, + "learning_rate": 1.8862240277872587e-05, + "loss": 0.4627, + "step": 1004 + }, + { + "epoch": 0.7146666666666667, + "grad_norm": 2.2743546962738037, + "learning_rate": 1.8858403300903285e-05, + "loss": 0.4341, + "step": 1005 + }, + { + "epoch": 0.7153777777777778, + "grad_norm": 1.2609387636184692, + "learning_rate": 1.8854560256532098e-05, + "loss": 0.4361, + "step": 1006 + }, + { + "epoch": 0.7160888888888889, + "grad_norm": 2.8257741928100586, + "learning_rate": 1.8850711147391257e-05, + "loss": 0.4323, + "step": 1007 + }, + { + "epoch": 0.7168, + "grad_norm": 1.5496368408203125, + "learning_rate": 1.8846855976117137e-05, + "loss": 0.4199, + "step": 1008 + }, + { + "epoch": 0.7175111111111111, + "grad_norm": 1.5379061698913574, + "learning_rate": 1.8842994745350266e-05, + "loss": 0.4135, + "step": 1009 + }, + { + "epoch": 0.7182222222222222, + "grad_norm": 1.4375029802322388, + "learning_rate": 1.8839127457735325e-05, + "loss": 0.3715, + "step": 1010 + }, + { + "epoch": 0.7189333333333333, + "grad_norm": 1.321212887763977, + "learning_rate": 1.883525411592114e-05, + "loss": 0.436, + "step": 1011 + }, + { + "epoch": 0.7196444444444444, + "grad_norm": 1.5866754055023193, + "learning_rate": 1.8831374722560686e-05, + "loss": 0.4133, + "step": 1012 + }, + { + "epoch": 0.7203555555555555, + "grad_norm": 1.1085195541381836, + "learning_rate": 1.8827489280311082e-05, + "loss": 0.4163, + "step": 1013 + }, + { + "epoch": 0.7210666666666666, + "grad_norm": 1.5205706357955933, + "learning_rate": 1.8823597791833595e-05, + "loss": 0.4122, + "step": 1014 + }, + { + "epoch": 0.7217777777777777, + "grad_norm": 1.3843448162078857, + "learning_rate": 1.8819700259793623e-05, + "loss": 0.3977, + "step": 1015 + }, + { + "epoch": 0.7224888888888888, + "grad_norm": 1.4499127864837646, + "learning_rate": 1.8815796686860717e-05, + "loss": 0.4217, + "step": 1016 + }, + { + "epoch": 0.7232, + "grad_norm": 2.1234042644500732, + "learning_rate": 1.8811887075708553e-05, + "loss": 0.4572, + "step": 1017 + }, + { + "epoch": 0.7239111111111111, + "grad_norm": 1.7839239835739136, + "learning_rate": 1.880797142901495e-05, + "loss": 0.4353, + "step": 1018 + }, + { + "epoch": 0.7246222222222222, + "grad_norm": 2.1299095153808594, + "learning_rate": 1.8804049749461856e-05, + "loss": 0.3729, + "step": 1019 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 1.7157394886016846, + "learning_rate": 1.880012203973536e-05, + "loss": 0.4059, + "step": 1020 + }, + { + "epoch": 0.7260444444444445, + "grad_norm": 1.8961623907089233, + "learning_rate": 1.8796188302525666e-05, + "loss": 0.4152, + "step": 1021 + }, + { + "epoch": 0.7267555555555556, + "grad_norm": 1.9264575242996216, + "learning_rate": 1.8792248540527124e-05, + "loss": 0.3999, + "step": 1022 + }, + { + "epoch": 0.7274666666666667, + "grad_norm": 1.7541251182556152, + "learning_rate": 1.8788302756438196e-05, + "loss": 0.4672, + "step": 1023 + }, + { + "epoch": 0.7281777777777778, + "grad_norm": 1.6375668048858643, + "learning_rate": 1.878435095296148e-05, + "loss": 0.4264, + "step": 1024 + }, + { + "epoch": 0.7288888888888889, + "grad_norm": 1.590945839881897, + "learning_rate": 1.878039313280369e-05, + "loss": 0.4423, + "step": 1025 + }, + { + "epoch": 0.7296, + "grad_norm": 1.4146884679794312, + "learning_rate": 1.877642929867566e-05, + "loss": 0.4123, + "step": 1026 + }, + { + "epoch": 0.7303111111111111, + "grad_norm": 1.6223822832107544, + "learning_rate": 1.877245945329235e-05, + "loss": 0.4284, + "step": 1027 + }, + { + "epoch": 0.7310222222222222, + "grad_norm": 1.368173599243164, + "learning_rate": 1.876848359937283e-05, + "loss": 0.4151, + "step": 1028 + }, + { + "epoch": 0.7317333333333333, + "grad_norm": 1.9520292282104492, + "learning_rate": 1.8764501739640292e-05, + "loss": 0.4265, + "step": 1029 + }, + { + "epoch": 0.7324444444444445, + "grad_norm": 1.3561264276504517, + "learning_rate": 1.876051387682204e-05, + "loss": 0.3976, + "step": 1030 + }, + { + "epoch": 0.7331555555555556, + "grad_norm": 2.033113956451416, + "learning_rate": 1.8756520013649476e-05, + "loss": 0.4145, + "step": 1031 + }, + { + "epoch": 0.7338666666666667, + "grad_norm": 1.5047447681427002, + "learning_rate": 1.8752520152858136e-05, + "loss": 0.3843, + "step": 1032 + }, + { + "epoch": 0.7345777777777778, + "grad_norm": 1.6205891370773315, + "learning_rate": 1.874851429718765e-05, + "loss": 0.4105, + "step": 1033 + }, + { + "epoch": 0.7352888888888889, + "grad_norm": 1.6487525701522827, + "learning_rate": 1.8744502449381748e-05, + "loss": 0.4218, + "step": 1034 + }, + { + "epoch": 0.736, + "grad_norm": 1.7168282270431519, + "learning_rate": 1.8740484612188283e-05, + "loss": 0.4308, + "step": 1035 + }, + { + "epoch": 0.7367111111111111, + "grad_norm": 1.5157570838928223, + "learning_rate": 1.8736460788359195e-05, + "loss": 0.4234, + "step": 1036 + }, + { + "epoch": 0.7374222222222222, + "grad_norm": 1.8201406002044678, + "learning_rate": 1.8732430980650524e-05, + "loss": 0.426, + "step": 1037 + }, + { + "epoch": 0.7381333333333333, + "grad_norm": 1.4510910511016846, + "learning_rate": 1.872839519182242e-05, + "loss": 0.4479, + "step": 1038 + }, + { + "epoch": 0.7388444444444444, + "grad_norm": 2.0350656509399414, + "learning_rate": 1.872435342463912e-05, + "loss": 0.437, + "step": 1039 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 1.526462197303772, + "learning_rate": 1.8720305681868954e-05, + "loss": 0.4207, + "step": 1040 + }, + { + "epoch": 0.7402666666666666, + "grad_norm": 2.466059684753418, + "learning_rate": 1.871625196628436e-05, + "loss": 0.4766, + "step": 1041 + }, + { + "epoch": 0.7409777777777777, + "grad_norm": 2.1220085620880127, + "learning_rate": 1.871219228066185e-05, + "loss": 0.4842, + "step": 1042 + }, + { + "epoch": 0.7416888888888888, + "grad_norm": 1.3863115310668945, + "learning_rate": 1.8708126627782032e-05, + "loss": 0.4146, + "step": 1043 + }, + { + "epoch": 0.7424, + "grad_norm": 1.9809832572937012, + "learning_rate": 1.8704055010429604e-05, + "loss": 0.4255, + "step": 1044 + }, + { + "epoch": 0.7431111111111111, + "grad_norm": 1.3458855152130127, + "learning_rate": 1.8699977431393342e-05, + "loss": 0.4064, + "step": 1045 + }, + { + "epoch": 0.7438222222222223, + "grad_norm": 1.7454112768173218, + "learning_rate": 1.869589389346611e-05, + "loss": 0.412, + "step": 1046 + }, + { + "epoch": 0.7445333333333334, + "grad_norm": 1.3203306198120117, + "learning_rate": 1.8691804399444858e-05, + "loss": 0.4274, + "step": 1047 + }, + { + "epoch": 0.7452444444444445, + "grad_norm": 2.0510079860687256, + "learning_rate": 1.8687708952130602e-05, + "loss": 0.4258, + "step": 1048 + }, + { + "epoch": 0.7459555555555556, + "grad_norm": 1.2821767330169678, + "learning_rate": 1.8683607554328455e-05, + "loss": 0.4099, + "step": 1049 + }, + { + "epoch": 0.7466666666666667, + "grad_norm": 1.8915725946426392, + "learning_rate": 1.867950020884758e-05, + "loss": 0.4189, + "step": 1050 + }, + { + "epoch": 0.7473777777777778, + "grad_norm": 1.5722265243530273, + "learning_rate": 1.8675386918501243e-05, + "loss": 0.424, + "step": 1051 + }, + { + "epoch": 0.7480888888888889, + "grad_norm": 1.827351450920105, + "learning_rate": 1.867126768610676e-05, + "loss": 0.3752, + "step": 1052 + }, + { + "epoch": 0.7488, + "grad_norm": 1.8520047664642334, + "learning_rate": 1.866714251448552e-05, + "loss": 0.427, + "step": 1053 + }, + { + "epoch": 0.7495111111111111, + "grad_norm": 2.272862195968628, + "learning_rate": 1.866301140646299e-05, + "loss": 0.4052, + "step": 1054 + }, + { + "epoch": 0.7502222222222222, + "grad_norm": 1.5204271078109741, + "learning_rate": 1.8658874364868703e-05, + "loss": 0.4863, + "step": 1055 + }, + { + "epoch": 0.7509333333333333, + "grad_norm": 1.3142095804214478, + "learning_rate": 1.8654731392536238e-05, + "loss": 0.3994, + "step": 1056 + }, + { + "epoch": 0.7516444444444444, + "grad_norm": 1.33653724193573, + "learning_rate": 1.865058249230326e-05, + "loss": 0.3905, + "step": 1057 + }, + { + "epoch": 0.7523555555555556, + "grad_norm": 1.439945101737976, + "learning_rate": 1.8646427667011465e-05, + "loss": 0.4204, + "step": 1058 + }, + { + "epoch": 0.7530666666666667, + "grad_norm": 1.0971317291259766, + "learning_rate": 1.8642266919506644e-05, + "loss": 0.4232, + "step": 1059 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 1.4764541387557983, + "learning_rate": 1.863810025263862e-05, + "loss": 0.403, + "step": 1060 + }, + { + "epoch": 0.7544888888888889, + "grad_norm": 1.4701802730560303, + "learning_rate": 1.863392766926127e-05, + "loss": 0.4495, + "step": 1061 + }, + { + "epoch": 0.7552, + "grad_norm": 2.035507917404175, + "learning_rate": 1.8629749172232528e-05, + "loss": 0.3936, + "step": 1062 + }, + { + "epoch": 0.7559111111111111, + "grad_norm": 1.9365460872650146, + "learning_rate": 1.862556476441439e-05, + "loss": 0.3875, + "step": 1063 + }, + { + "epoch": 0.7566222222222222, + "grad_norm": 1.6877012252807617, + "learning_rate": 1.8621374448672878e-05, + "loss": 0.4458, + "step": 1064 + }, + { + "epoch": 0.7573333333333333, + "grad_norm": 2.3012890815734863, + "learning_rate": 1.861717822787808e-05, + "loss": 0.4439, + "step": 1065 + }, + { + "epoch": 0.7580444444444444, + "grad_norm": 1.586560845375061, + "learning_rate": 1.861297610490412e-05, + "loss": 0.4304, + "step": 1066 + }, + { + "epoch": 0.7587555555555555, + "grad_norm": 1.9860994815826416, + "learning_rate": 1.860876808262916e-05, + "loss": 0.3835, + "step": 1067 + }, + { + "epoch": 0.7594666666666666, + "grad_norm": 1.5063390731811523, + "learning_rate": 1.8604554163935416e-05, + "loss": 0.4393, + "step": 1068 + }, + { + "epoch": 0.7601777777777777, + "grad_norm": 1.2355026006698608, + "learning_rate": 1.860033435170913e-05, + "loss": 0.4108, + "step": 1069 + }, + { + "epoch": 0.7608888888888888, + "grad_norm": 1.4176344871520996, + "learning_rate": 1.859610864884059e-05, + "loss": 0.4527, + "step": 1070 + }, + { + "epoch": 0.7616, + "grad_norm": 1.4012192487716675, + "learning_rate": 1.8591877058224104e-05, + "loss": 0.3816, + "step": 1071 + }, + { + "epoch": 0.7623111111111112, + "grad_norm": 1.4236618280410767, + "learning_rate": 1.8587639582758032e-05, + "loss": 0.4605, + "step": 1072 + }, + { + "epoch": 0.7630222222222223, + "grad_norm": 1.695212483406067, + "learning_rate": 1.8583396225344755e-05, + "loss": 0.3552, + "step": 1073 + }, + { + "epoch": 0.7637333333333334, + "grad_norm": 1.2142409086227417, + "learning_rate": 1.8579146988890677e-05, + "loss": 0.4213, + "step": 1074 + }, + { + "epoch": 0.7644444444444445, + "grad_norm": 1.3483442068099976, + "learning_rate": 1.8574891876306243e-05, + "loss": 0.4242, + "step": 1075 + }, + { + "epoch": 0.7651555555555556, + "grad_norm": 1.6383816003799438, + "learning_rate": 1.857063089050591e-05, + "loss": 0.4119, + "step": 1076 + }, + { + "epoch": 0.7658666666666667, + "grad_norm": 1.2471303939819336, + "learning_rate": 1.8566364034408166e-05, + "loss": 0.4061, + "step": 1077 + }, + { + "epoch": 0.7665777777777778, + "grad_norm": 1.3698431253433228, + "learning_rate": 1.8562091310935514e-05, + "loss": 0.4061, + "step": 1078 + }, + { + "epoch": 0.7672888888888889, + "grad_norm": 1.082532525062561, + "learning_rate": 1.8557812723014476e-05, + "loss": 0.4317, + "step": 1079 + }, + { + "epoch": 0.768, + "grad_norm": 1.2265276908874512, + "learning_rate": 1.8553528273575597e-05, + "loss": 0.4413, + "step": 1080 + }, + { + "epoch": 0.7687111111111111, + "grad_norm": 1.4205242395401, + "learning_rate": 1.854923796555343e-05, + "loss": 0.4419, + "step": 1081 + }, + { + "epoch": 0.7694222222222222, + "grad_norm": 1.2588516473770142, + "learning_rate": 1.8544941801886543e-05, + "loss": 0.4085, + "step": 1082 + }, + { + "epoch": 0.7701333333333333, + "grad_norm": 1.3697237968444824, + "learning_rate": 1.8540639785517518e-05, + "loss": 0.4402, + "step": 1083 + }, + { + "epoch": 0.7708444444444444, + "grad_norm": 1.367194652557373, + "learning_rate": 1.8536331919392938e-05, + "loss": 0.3914, + "step": 1084 + }, + { + "epoch": 0.7715555555555556, + "grad_norm": 1.1375547647476196, + "learning_rate": 1.8532018206463404e-05, + "loss": 0.4322, + "step": 1085 + }, + { + "epoch": 0.7722666666666667, + "grad_norm": 1.6496233940124512, + "learning_rate": 1.8527698649683507e-05, + "loss": 0.4076, + "step": 1086 + }, + { + "epoch": 0.7729777777777778, + "grad_norm": 1.1615852117538452, + "learning_rate": 1.8523373252011854e-05, + "loss": 0.3852, + "step": 1087 + }, + { + "epoch": 0.7736888888888889, + "grad_norm": 1.8026020526885986, + "learning_rate": 1.8519042016411045e-05, + "loss": 0.4392, + "step": 1088 + }, + { + "epoch": 0.7744, + "grad_norm": 1.2586928606033325, + "learning_rate": 1.8514704945847684e-05, + "loss": 0.4211, + "step": 1089 + }, + { + "epoch": 0.7751111111111111, + "grad_norm": 1.3373411893844604, + "learning_rate": 1.8510362043292363e-05, + "loss": 0.4533, + "step": 1090 + }, + { + "epoch": 0.7758222222222222, + "grad_norm": 1.22789466381073, + "learning_rate": 1.850601331171968e-05, + "loss": 0.3921, + "step": 1091 + }, + { + "epoch": 0.7765333333333333, + "grad_norm": 1.070896863937378, + "learning_rate": 1.850165875410822e-05, + "loss": 0.4233, + "step": 1092 + }, + { + "epoch": 0.7772444444444444, + "grad_norm": 1.267790675163269, + "learning_rate": 1.8497298373440548e-05, + "loss": 0.4499, + "step": 1093 + }, + { + "epoch": 0.7779555555555555, + "grad_norm": 1.0554709434509277, + "learning_rate": 1.8492932172703236e-05, + "loss": 0.451, + "step": 1094 + }, + { + "epoch": 0.7786666666666666, + "grad_norm": 1.0826747417449951, + "learning_rate": 1.8488560154886834e-05, + "loss": 0.4144, + "step": 1095 + }, + { + "epoch": 0.7793777777777777, + "grad_norm": 1.2507866621017456, + "learning_rate": 1.848418232298587e-05, + "loss": 0.4088, + "step": 1096 + }, + { + "epoch": 0.7800888888888889, + "grad_norm": 1.1860218048095703, + "learning_rate": 1.8479798679998863e-05, + "loss": 0.4439, + "step": 1097 + }, + { + "epoch": 0.7808, + "grad_norm": 1.3823771476745605, + "learning_rate": 1.8475409228928314e-05, + "loss": 0.4466, + "step": 1098 + }, + { + "epoch": 0.7815111111111112, + "grad_norm": 1.2125331163406372, + "learning_rate": 1.8471013972780693e-05, + "loss": 0.3914, + "step": 1099 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 1.092445731163025, + "learning_rate": 1.8466612914566447e-05, + "loss": 0.3931, + "step": 1100 + }, + { + "epoch": 0.7829333333333334, + "grad_norm": 1.1295580863952637, + "learning_rate": 1.8462206057300004e-05, + "loss": 0.3873, + "step": 1101 + }, + { + "epoch": 0.7836444444444445, + "grad_norm": 1.199808120727539, + "learning_rate": 1.8457793403999766e-05, + "loss": 0.415, + "step": 1102 + }, + { + "epoch": 0.7843555555555556, + "grad_norm": 1.190609097480774, + "learning_rate": 1.8453374957688092e-05, + "loss": 0.4595, + "step": 1103 + }, + { + "epoch": 0.7850666666666667, + "grad_norm": 1.159568428993225, + "learning_rate": 1.844895072139132e-05, + "loss": 0.4717, + "step": 1104 + }, + { + "epoch": 0.7857777777777778, + "grad_norm": 1.2271127700805664, + "learning_rate": 1.8444520698139748e-05, + "loss": 0.4498, + "step": 1105 + }, + { + "epoch": 0.7864888888888889, + "grad_norm": 1.6676779985427856, + "learning_rate": 1.844008489096764e-05, + "loss": 0.4398, + "step": 1106 + }, + { + "epoch": 0.7872, + "grad_norm": 1.34696364402771, + "learning_rate": 1.8435643302913223e-05, + "loss": 0.4514, + "step": 1107 + }, + { + "epoch": 0.7879111111111111, + "grad_norm": 2.860999584197998, + "learning_rate": 1.843119593701868e-05, + "loss": 0.4656, + "step": 1108 + }, + { + "epoch": 0.7886222222222222, + "grad_norm": 1.4551297426223755, + "learning_rate": 1.8426742796330154e-05, + "loss": 0.3989, + "step": 1109 + }, + { + "epoch": 0.7893333333333333, + "grad_norm": 1.299858808517456, + "learning_rate": 1.8422283883897744e-05, + "loss": 0.5017, + "step": 1110 + }, + { + "epoch": 0.7900444444444444, + "grad_norm": 1.4028828144073486, + "learning_rate": 1.8417819202775497e-05, + "loss": 0.4394, + "step": 1111 + }, + { + "epoch": 0.7907555555555555, + "grad_norm": 1.3205115795135498, + "learning_rate": 1.8413348756021415e-05, + "loss": 0.4198, + "step": 1112 + }, + { + "epoch": 0.7914666666666667, + "grad_norm": 1.231268286705017, + "learning_rate": 1.8408872546697452e-05, + "loss": 0.412, + "step": 1113 + }, + { + "epoch": 0.7921777777777778, + "grad_norm": 1.589005947113037, + "learning_rate": 1.8404390577869508e-05, + "loss": 0.4883, + "step": 1114 + }, + { + "epoch": 0.7928888888888889, + "grad_norm": 1.1024540662765503, + "learning_rate": 1.8399902852607422e-05, + "loss": 0.4335, + "step": 1115 + }, + { + "epoch": 0.7936, + "grad_norm": 1.22498619556427, + "learning_rate": 1.839540937398498e-05, + "loss": 0.4436, + "step": 1116 + }, + { + "epoch": 0.7943111111111111, + "grad_norm": 1.3043051958084106, + "learning_rate": 1.839091014507991e-05, + "loss": 0.4212, + "step": 1117 + }, + { + "epoch": 0.7950222222222222, + "grad_norm": 1.4414454698562622, + "learning_rate": 1.8386405168973875e-05, + "loss": 0.3898, + "step": 1118 + }, + { + "epoch": 0.7957333333333333, + "grad_norm": 1.076413869857788, + "learning_rate": 1.8381894448752476e-05, + "loss": 0.4383, + "step": 1119 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 1.3791922330856323, + "learning_rate": 1.8377377987505252e-05, + "loss": 0.3958, + "step": 1120 + }, + { + "epoch": 0.7971555555555555, + "grad_norm": 1.2400132417678833, + "learning_rate": 1.8372855788325665e-05, + "loss": 0.4738, + "step": 1121 + }, + { + "epoch": 0.7978666666666666, + "grad_norm": 1.3730205297470093, + "learning_rate": 1.836832785431112e-05, + "loss": 0.4148, + "step": 1122 + }, + { + "epoch": 0.7985777777777778, + "grad_norm": 1.2289845943450928, + "learning_rate": 1.8363794188562936e-05, + "loss": 0.408, + "step": 1123 + }, + { + "epoch": 0.7992888888888889, + "grad_norm": 1.3101403713226318, + "learning_rate": 1.8359254794186368e-05, + "loss": 0.4212, + "step": 1124 + }, + { + "epoch": 0.8, + "grad_norm": 1.2930467128753662, + "learning_rate": 1.8354709674290597e-05, + "loss": 0.4415, + "step": 1125 + }, + { + "epoch": 0.8007111111111112, + "grad_norm": 1.7991876602172852, + "learning_rate": 1.8350158831988708e-05, + "loss": 0.3847, + "step": 1126 + }, + { + "epoch": 0.8014222222222223, + "grad_norm": 1.5983142852783203, + "learning_rate": 1.8345602270397726e-05, + "loss": 0.4262, + "step": 1127 + }, + { + "epoch": 0.8021333333333334, + "grad_norm": 1.615351676940918, + "learning_rate": 1.8341039992638585e-05, + "loss": 0.4127, + "step": 1128 + }, + { + "epoch": 0.8028444444444445, + "grad_norm": 1.4755810499191284, + "learning_rate": 1.8336472001836127e-05, + "loss": 0.4303, + "step": 1129 + }, + { + "epoch": 0.8035555555555556, + "grad_norm": 1.8304355144500732, + "learning_rate": 1.8331898301119123e-05, + "loss": 0.4117, + "step": 1130 + }, + { + "epoch": 0.8042666666666667, + "grad_norm": 1.3305652141571045, + "learning_rate": 1.8327318893620236e-05, + "loss": 0.4017, + "step": 1131 + }, + { + "epoch": 0.8049777777777778, + "grad_norm": 1.5291268825531006, + "learning_rate": 1.832273378247606e-05, + "loss": 0.4243, + "step": 1132 + }, + { + "epoch": 0.8056888888888889, + "grad_norm": 1.4422999620437622, + "learning_rate": 1.831814297082707e-05, + "loss": 0.4499, + "step": 1133 + }, + { + "epoch": 0.8064, + "grad_norm": 1.387300729751587, + "learning_rate": 1.831354646181767e-05, + "loss": 0.4515, + "step": 1134 + }, + { + "epoch": 0.8071111111111111, + "grad_norm": 1.3656572103500366, + "learning_rate": 1.830894425859615e-05, + "loss": 0.4027, + "step": 1135 + }, + { + "epoch": 0.8078222222222222, + "grad_norm": 1.2059826850891113, + "learning_rate": 1.8304336364314707e-05, + "loss": 0.4765, + "step": 1136 + }, + { + "epoch": 0.8085333333333333, + "grad_norm": 1.278936743736267, + "learning_rate": 1.8299722782129428e-05, + "loss": 0.3966, + "step": 1137 + }, + { + "epoch": 0.8092444444444444, + "grad_norm": 1.312134027481079, + "learning_rate": 1.829510351520031e-05, + "loss": 0.3858, + "step": 1138 + }, + { + "epoch": 0.8099555555555555, + "grad_norm": 1.211637258529663, + "learning_rate": 1.8290478566691236e-05, + "loss": 0.4335, + "step": 1139 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 1.328520655632019, + "learning_rate": 1.828584793976997e-05, + "loss": 0.4103, + "step": 1140 + }, + { + "epoch": 0.8113777777777778, + "grad_norm": 1.4415886402130127, + "learning_rate": 1.828121163760819e-05, + "loss": 0.4009, + "step": 1141 + }, + { + "epoch": 0.8120888888888889, + "grad_norm": 1.1619144678115845, + "learning_rate": 1.827656966338143e-05, + "loss": 0.4321, + "step": 1142 + }, + { + "epoch": 0.8128, + "grad_norm": 1.1977046728134155, + "learning_rate": 1.827192202026914e-05, + "loss": 0.4336, + "step": 1143 + }, + { + "epoch": 0.8135111111111111, + "grad_norm": 1.3298897743225098, + "learning_rate": 1.8267268711454634e-05, + "loss": 0.4067, + "step": 1144 + }, + { + "epoch": 0.8142222222222222, + "grad_norm": 1.0704607963562012, + "learning_rate": 1.826260974012511e-05, + "loss": 0.4147, + "step": 1145 + }, + { + "epoch": 0.8149333333333333, + "grad_norm": 1.5060964822769165, + "learning_rate": 1.825794510947164e-05, + "loss": 0.4228, + "step": 1146 + }, + { + "epoch": 0.8156444444444444, + "grad_norm": 1.5884753465652466, + "learning_rate": 1.825327482268919e-05, + "loss": 0.4077, + "step": 1147 + }, + { + "epoch": 0.8163555555555555, + "grad_norm": 1.1785101890563965, + "learning_rate": 1.8248598882976575e-05, + "loss": 0.4597, + "step": 1148 + }, + { + "epoch": 0.8170666666666667, + "grad_norm": 1.0555037260055542, + "learning_rate": 1.8243917293536503e-05, + "loss": 0.4366, + "step": 1149 + }, + { + "epoch": 0.8177777777777778, + "grad_norm": 1.5978187322616577, + "learning_rate": 1.8239230057575542e-05, + "loss": 0.4402, + "step": 1150 + }, + { + "epoch": 0.8184888888888889, + "grad_norm": 1.2311723232269287, + "learning_rate": 1.823453717830413e-05, + "loss": 0.4282, + "step": 1151 + }, + { + "epoch": 0.8192, + "grad_norm": 1.6638100147247314, + "learning_rate": 1.8229838658936566e-05, + "loss": 0.4709, + "step": 1152 + }, + { + "epoch": 0.8199111111111111, + "grad_norm": 1.766432285308838, + "learning_rate": 1.8225134502691014e-05, + "loss": 0.4823, + "step": 1153 + }, + { + "epoch": 0.8206222222222223, + "grad_norm": 1.6033650636672974, + "learning_rate": 1.8220424712789502e-05, + "loss": 0.4449, + "step": 1154 + }, + { + "epoch": 0.8213333333333334, + "grad_norm": 1.5729529857635498, + "learning_rate": 1.8215709292457922e-05, + "loss": 0.4213, + "step": 1155 + }, + { + "epoch": 0.8220444444444445, + "grad_norm": 1.4979512691497803, + "learning_rate": 1.8210988244926003e-05, + "loss": 0.3949, + "step": 1156 + }, + { + "epoch": 0.8227555555555556, + "grad_norm": 1.8964827060699463, + "learning_rate": 1.8206261573427345e-05, + "loss": 0.4496, + "step": 1157 + }, + { + "epoch": 0.8234666666666667, + "grad_norm": 1.332676887512207, + "learning_rate": 1.8201529281199398e-05, + "loss": 0.3881, + "step": 1158 + }, + { + "epoch": 0.8241777777777778, + "grad_norm": 1.5395827293395996, + "learning_rate": 1.819679137148346e-05, + "loss": 0.4209, + "step": 1159 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 1.467753529548645, + "learning_rate": 1.8192047847524667e-05, + "loss": 0.4173, + "step": 1160 + }, + { + "epoch": 0.8256, + "grad_norm": 1.6400642395019531, + "learning_rate": 1.8187298712572022e-05, + "loss": 0.5017, + "step": 1161 + }, + { + "epoch": 0.8263111111111111, + "grad_norm": 1.8867915868759155, + "learning_rate": 1.818254396987835e-05, + "loss": 0.4425, + "step": 1162 + }, + { + "epoch": 0.8270222222222222, + "grad_norm": 1.435748815536499, + "learning_rate": 1.8177783622700328e-05, + "loss": 0.4422, + "step": 1163 + }, + { + "epoch": 0.8277333333333333, + "grad_norm": 1.725231647491455, + "learning_rate": 1.8173017674298467e-05, + "loss": 0.4151, + "step": 1164 + }, + { + "epoch": 0.8284444444444444, + "grad_norm": 2.1153364181518555, + "learning_rate": 1.816824612793712e-05, + "loss": 0.4355, + "step": 1165 + }, + { + "epoch": 0.8291555555555555, + "grad_norm": 1.5387612581253052, + "learning_rate": 1.8163468986884466e-05, + "loss": 0.4329, + "step": 1166 + }, + { + "epoch": 0.8298666666666666, + "grad_norm": 2.2984468936920166, + "learning_rate": 1.8158686254412527e-05, + "loss": 0.3879, + "step": 1167 + }, + { + "epoch": 0.8305777777777777, + "grad_norm": 2.0996482372283936, + "learning_rate": 1.8153897933797145e-05, + "loss": 0.3821, + "step": 1168 + }, + { + "epoch": 0.8312888888888889, + "grad_norm": 1.6913201808929443, + "learning_rate": 1.8149104028317988e-05, + "loss": 0.4234, + "step": 1169 + }, + { + "epoch": 0.832, + "grad_norm": 2.258856773376465, + "learning_rate": 1.8144304541258562e-05, + "loss": 0.4321, + "step": 1170 + }, + { + "epoch": 0.8327111111111111, + "grad_norm": 1.2349025011062622, + "learning_rate": 1.813949947590619e-05, + "loss": 0.4258, + "step": 1171 + }, + { + "epoch": 0.8334222222222222, + "grad_norm": 2.3016772270202637, + "learning_rate": 1.8134688835552003e-05, + "loss": 0.3813, + "step": 1172 + }, + { + "epoch": 0.8341333333333333, + "grad_norm": 2.1430838108062744, + "learning_rate": 1.812987262349097e-05, + "loss": 0.4444, + "step": 1173 + }, + { + "epoch": 0.8348444444444444, + "grad_norm": 2.378312349319458, + "learning_rate": 1.8125050843021864e-05, + "loss": 0.4237, + "step": 1174 + }, + { + "epoch": 0.8355555555555556, + "grad_norm": 2.001107692718506, + "learning_rate": 1.8120223497447277e-05, + "loss": 0.3899, + "step": 1175 + }, + { + "epoch": 0.8362666666666667, + "grad_norm": 1.6105848550796509, + "learning_rate": 1.8115390590073612e-05, + "loss": 0.4311, + "step": 1176 + }, + { + "epoch": 0.8369777777777778, + "grad_norm": 1.2973062992095947, + "learning_rate": 1.811055212421108e-05, + "loss": 0.3933, + "step": 1177 + }, + { + "epoch": 0.8376888888888889, + "grad_norm": 1.8080798387527466, + "learning_rate": 1.81057081031737e-05, + "loss": 0.3846, + "step": 1178 + }, + { + "epoch": 0.8384, + "grad_norm": 1.231115698814392, + "learning_rate": 1.8100858530279295e-05, + "loss": 0.4066, + "step": 1179 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 1.821086049079895, + "learning_rate": 1.8096003408849494e-05, + "loss": 0.425, + "step": 1180 + }, + { + "epoch": 0.8398222222222222, + "grad_norm": 1.499575138092041, + "learning_rate": 1.809114274220972e-05, + "loss": 0.4264, + "step": 1181 + }, + { + "epoch": 0.8405333333333334, + "grad_norm": 1.6319888830184937, + "learning_rate": 1.8086276533689203e-05, + "loss": 0.3905, + "step": 1182 + }, + { + "epoch": 0.8412444444444445, + "grad_norm": 1.6146550178527832, + "learning_rate": 1.8081404786620962e-05, + "loss": 0.3625, + "step": 1183 + }, + { + "epoch": 0.8419555555555556, + "grad_norm": 1.6807446479797363, + "learning_rate": 1.807652750434181e-05, + "loss": 0.3755, + "step": 1184 + }, + { + "epoch": 0.8426666666666667, + "grad_norm": 1.1371365785598755, + "learning_rate": 1.8071644690192354e-05, + "loss": 0.4051, + "step": 1185 + }, + { + "epoch": 0.8433777777777778, + "grad_norm": 1.8514411449432373, + "learning_rate": 1.8066756347516985e-05, + "loss": 0.4355, + "step": 1186 + }, + { + "epoch": 0.8440888888888889, + "grad_norm": 1.2479661703109741, + "learning_rate": 1.806186247966389e-05, + "loss": 0.4471, + "step": 1187 + }, + { + "epoch": 0.8448, + "grad_norm": 1.3693217039108276, + "learning_rate": 1.805696308998503e-05, + "loss": 0.3849, + "step": 1188 + }, + { + "epoch": 0.8455111111111111, + "grad_norm": 1.483383297920227, + "learning_rate": 1.8052058181836152e-05, + "loss": 0.4485, + "step": 1189 + }, + { + "epoch": 0.8462222222222222, + "grad_norm": 1.689863920211792, + "learning_rate": 1.804714775857679e-05, + "loss": 0.4217, + "step": 1190 + }, + { + "epoch": 0.8469333333333333, + "grad_norm": 1.5615885257720947, + "learning_rate": 1.8042231823570243e-05, + "loss": 0.4476, + "step": 1191 + }, + { + "epoch": 0.8476444444444444, + "grad_norm": 1.9366493225097656, + "learning_rate": 1.803731038018359e-05, + "loss": 0.4209, + "step": 1192 + }, + { + "epoch": 0.8483555555555555, + "grad_norm": 1.316591501235962, + "learning_rate": 1.8032383431787693e-05, + "loss": 0.4477, + "step": 1193 + }, + { + "epoch": 0.8490666666666666, + "grad_norm": 1.2623172998428345, + "learning_rate": 1.8027450981757162e-05, + "loss": 0.4167, + "step": 1194 + }, + { + "epoch": 0.8497777777777777, + "grad_norm": 1.7907108068466187, + "learning_rate": 1.80225130334704e-05, + "loss": 0.3959, + "step": 1195 + }, + { + "epoch": 0.8504888888888888, + "grad_norm": 1.190742015838623, + "learning_rate": 1.8017569590309564e-05, + "loss": 0.4209, + "step": 1196 + }, + { + "epoch": 0.8512, + "grad_norm": 1.4494214057922363, + "learning_rate": 1.8012620655660566e-05, + "loss": 0.3942, + "step": 1197 + }, + { + "epoch": 0.8519111111111111, + "grad_norm": 1.4642387628555298, + "learning_rate": 1.8007666232913095e-05, + "loss": 0.4031, + "step": 1198 + }, + { + "epoch": 0.8526222222222222, + "grad_norm": 1.1215721368789673, + "learning_rate": 1.8002706325460595e-05, + "loss": 0.369, + "step": 1199 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 1.6557698249816895, + "learning_rate": 1.799774093670026e-05, + "loss": 0.4063, + "step": 1200 + }, + { + "epoch": 0.8540444444444445, + "grad_norm": 1.1156922578811646, + "learning_rate": 1.7992770070033043e-05, + "loss": 0.3826, + "step": 1201 + }, + { + "epoch": 0.8547555555555556, + "grad_norm": 1.1856187582015991, + "learning_rate": 1.798779372886365e-05, + "loss": 0.412, + "step": 1202 + }, + { + "epoch": 0.8554666666666667, + "grad_norm": 1.460816740989685, + "learning_rate": 1.7982811916600536e-05, + "loss": 0.4036, + "step": 1203 + }, + { + "epoch": 0.8561777777777778, + "grad_norm": 1.240149736404419, + "learning_rate": 1.7977824636655897e-05, + "loss": 0.4166, + "step": 1204 + }, + { + "epoch": 0.8568888888888889, + "grad_norm": 1.1796374320983887, + "learning_rate": 1.7972831892445687e-05, + "loss": 0.4101, + "step": 1205 + }, + { + "epoch": 0.8576, + "grad_norm": 1.6394150257110596, + "learning_rate": 1.796783368738959e-05, + "loss": 0.363, + "step": 1206 + }, + { + "epoch": 0.8583111111111111, + "grad_norm": 1.6120822429656982, + "learning_rate": 1.7962830024911038e-05, + "loss": 0.3748, + "step": 1207 + }, + { + "epoch": 0.8590222222222222, + "grad_norm": 1.304007887840271, + "learning_rate": 1.79578209084372e-05, + "loss": 0.4103, + "step": 1208 + }, + { + "epoch": 0.8597333333333333, + "grad_norm": 1.5500869750976562, + "learning_rate": 1.7952806341398972e-05, + "loss": 0.4197, + "step": 1209 + }, + { + "epoch": 0.8604444444444445, + "grad_norm": 1.3432271480560303, + "learning_rate": 1.7947786327230996e-05, + "loss": 0.4758, + "step": 1210 + }, + { + "epoch": 0.8611555555555556, + "grad_norm": 1.6576471328735352, + "learning_rate": 1.794276086937164e-05, + "loss": 0.4236, + "step": 1211 + }, + { + "epoch": 0.8618666666666667, + "grad_norm": 1.3624591827392578, + "learning_rate": 1.7937729971262992e-05, + "loss": 0.4181, + "step": 1212 + }, + { + "epoch": 0.8625777777777778, + "grad_norm": 2.357318878173828, + "learning_rate": 1.7932693636350886e-05, + "loss": 0.394, + "step": 1213 + }, + { + "epoch": 0.8632888888888889, + "grad_norm": 2.1654164791107178, + "learning_rate": 1.7927651868084857e-05, + "loss": 0.408, + "step": 1214 + }, + { + "epoch": 0.864, + "grad_norm": 1.9448400735855103, + "learning_rate": 1.792260466991818e-05, + "loss": 0.4597, + "step": 1215 + }, + { + "epoch": 0.8647111111111111, + "grad_norm": 1.9477276802062988, + "learning_rate": 1.7917552045307838e-05, + "loss": 0.4082, + "step": 1216 + }, + { + "epoch": 0.8654222222222222, + "grad_norm": 1.702185034751892, + "learning_rate": 1.7912493997714534e-05, + "loss": 0.4341, + "step": 1217 + }, + { + "epoch": 0.8661333333333333, + "grad_norm": 1.7070279121398926, + "learning_rate": 1.7907430530602688e-05, + "loss": 0.4231, + "step": 1218 + }, + { + "epoch": 0.8668444444444444, + "grad_norm": 1.4015100002288818, + "learning_rate": 1.7902361647440424e-05, + "loss": 0.431, + "step": 1219 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 1.1931381225585938, + "learning_rate": 1.7897287351699586e-05, + "loss": 0.4202, + "step": 1220 + }, + { + "epoch": 0.8682666666666666, + "grad_norm": 1.5053554773330688, + "learning_rate": 1.7892207646855718e-05, + "loss": 0.4341, + "step": 1221 + }, + { + "epoch": 0.8689777777777777, + "grad_norm": 1.2424976825714111, + "learning_rate": 1.7887122536388073e-05, + "loss": 0.4281, + "step": 1222 + }, + { + "epoch": 0.8696888888888888, + "grad_norm": 1.3937859535217285, + "learning_rate": 1.78820320237796e-05, + "loss": 0.38, + "step": 1223 + }, + { + "epoch": 0.8704, + "grad_norm": 1.4240094423294067, + "learning_rate": 1.7876936112516963e-05, + "loss": 0.3888, + "step": 1224 + }, + { + "epoch": 0.8711111111111111, + "grad_norm": 1.6840920448303223, + "learning_rate": 1.7871834806090502e-05, + "loss": 0.439, + "step": 1225 + }, + { + "epoch": 0.8718222222222223, + "grad_norm": 1.4851680994033813, + "learning_rate": 1.786672810799427e-05, + "loss": 0.425, + "step": 1226 + }, + { + "epoch": 0.8725333333333334, + "grad_norm": 1.086274266242981, + "learning_rate": 1.7861616021726008e-05, + "loss": 0.4316, + "step": 1227 + }, + { + "epoch": 0.8732444444444445, + "grad_norm": 1.4956555366516113, + "learning_rate": 1.7856498550787144e-05, + "loss": 0.424, + "step": 1228 + }, + { + "epoch": 0.8739555555555556, + "grad_norm": 1.1356701850891113, + "learning_rate": 1.7851375698682795e-05, + "loss": 0.403, + "step": 1229 + }, + { + "epoch": 0.8746666666666667, + "grad_norm": 1.7328509092330933, + "learning_rate": 1.784624746892177e-05, + "loss": 0.3968, + "step": 1230 + }, + { + "epoch": 0.8753777777777778, + "grad_norm": 1.3245089054107666, + "learning_rate": 1.784111386501655e-05, + "loss": 0.3738, + "step": 1231 + }, + { + "epoch": 0.8760888888888889, + "grad_norm": 1.8584436178207397, + "learning_rate": 1.7835974890483307e-05, + "loss": 0.4049, + "step": 1232 + }, + { + "epoch": 0.8768, + "grad_norm": 1.6986314058303833, + "learning_rate": 1.7830830548841887e-05, + "loss": 0.3907, + "step": 1233 + }, + { + "epoch": 0.8775111111111111, + "grad_norm": 1.3446714878082275, + "learning_rate": 1.782568084361582e-05, + "loss": 0.3971, + "step": 1234 + }, + { + "epoch": 0.8782222222222222, + "grad_norm": 1.332805871963501, + "learning_rate": 1.782052577833229e-05, + "loss": 0.3856, + "step": 1235 + }, + { + "epoch": 0.8789333333333333, + "grad_norm": 1.5187757015228271, + "learning_rate": 1.781536535652217e-05, + "loss": 0.4156, + "step": 1236 + }, + { + "epoch": 0.8796444444444445, + "grad_norm": 1.6471110582351685, + "learning_rate": 1.7810199581720004e-05, + "loss": 0.4373, + "step": 1237 + }, + { + "epoch": 0.8803555555555556, + "grad_norm": 1.536326289176941, + "learning_rate": 1.7805028457463985e-05, + "loss": 0.4031, + "step": 1238 + }, + { + "epoch": 0.8810666666666667, + "grad_norm": 1.503057599067688, + "learning_rate": 1.7799851987295987e-05, + "loss": 0.4253, + "step": 1239 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 1.625705361366272, + "learning_rate": 1.7794670174761537e-05, + "loss": 0.3939, + "step": 1240 + }, + { + "epoch": 0.8824888888888889, + "grad_norm": 1.446251630783081, + "learning_rate": 1.7789483023409822e-05, + "loss": 0.4387, + "step": 1241 + }, + { + "epoch": 0.8832, + "grad_norm": 2.0353822708129883, + "learning_rate": 1.7784290536793687e-05, + "loss": 0.461, + "step": 1242 + }, + { + "epoch": 0.8839111111111111, + "grad_norm": 1.825329303741455, + "learning_rate": 1.7779092718469635e-05, + "loss": 0.3977, + "step": 1243 + }, + { + "epoch": 0.8846222222222222, + "grad_norm": 2.0081870555877686, + "learning_rate": 1.777388957199781e-05, + "loss": 0.4311, + "step": 1244 + }, + { + "epoch": 0.8853333333333333, + "grad_norm": 2.4709835052490234, + "learning_rate": 1.7768681100942017e-05, + "loss": 0.4242, + "step": 1245 + }, + { + "epoch": 0.8860444444444444, + "grad_norm": 1.4503601789474487, + "learning_rate": 1.77634673088697e-05, + "loss": 0.4169, + "step": 1246 + }, + { + "epoch": 0.8867555555555555, + "grad_norm": 2.243291139602661, + "learning_rate": 1.775824819935196e-05, + "loss": 0.4646, + "step": 1247 + }, + { + "epoch": 0.8874666666666666, + "grad_norm": 1.369869351387024, + "learning_rate": 1.7753023775963516e-05, + "loss": 0.4538, + "step": 1248 + }, + { + "epoch": 0.8881777777777777, + "grad_norm": 1.8786637783050537, + "learning_rate": 1.7747794042282756e-05, + "loss": 0.4006, + "step": 1249 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 1.370174765586853, + "learning_rate": 1.774255900189168e-05, + "loss": 0.4248, + "step": 1250 + }, + { + "epoch": 0.8896, + "grad_norm": 1.0141006708145142, + "learning_rate": 1.7737318658375938e-05, + "loss": 0.3563, + "step": 1251 + }, + { + "epoch": 0.8903111111111112, + "grad_norm": 1.5416967868804932, + "learning_rate": 1.773207301532481e-05, + "loss": 0.4117, + "step": 1252 + }, + { + "epoch": 0.8910222222222223, + "grad_norm": 1.4559931755065918, + "learning_rate": 1.77268220763312e-05, + "loss": 0.3864, + "step": 1253 + }, + { + "epoch": 0.8917333333333334, + "grad_norm": 1.2244762182235718, + "learning_rate": 1.7721565844991643e-05, + "loss": 0.4142, + "step": 1254 + }, + { + "epoch": 0.8924444444444445, + "grad_norm": 1.7288707494735718, + "learning_rate": 1.77163043249063e-05, + "loss": 0.3698, + "step": 1255 + }, + { + "epoch": 0.8931555555555556, + "grad_norm": 1.4744073152542114, + "learning_rate": 1.7711037519678957e-05, + "loss": 0.3826, + "step": 1256 + }, + { + "epoch": 0.8938666666666667, + "grad_norm": 1.393553376197815, + "learning_rate": 1.770576543291701e-05, + "loss": 0.414, + "step": 1257 + }, + { + "epoch": 0.8945777777777778, + "grad_norm": 2.1229217052459717, + "learning_rate": 1.770048806823148e-05, + "loss": 0.3929, + "step": 1258 + }, + { + "epoch": 0.8952888888888889, + "grad_norm": 1.114048719406128, + "learning_rate": 1.769520542923701e-05, + "loss": 0.415, + "step": 1259 + }, + { + "epoch": 0.896, + "grad_norm": 2.398008346557617, + "learning_rate": 1.7689917519551834e-05, + "loss": 0.4307, + "step": 1260 + }, + { + "epoch": 0.8967111111111111, + "grad_norm": 1.3663525581359863, + "learning_rate": 1.7684624342797817e-05, + "loss": 0.4022, + "step": 1261 + }, + { + "epoch": 0.8974222222222222, + "grad_norm": 1.7757196426391602, + "learning_rate": 1.767932590260042e-05, + "loss": 0.4683, + "step": 1262 + }, + { + "epoch": 0.8981333333333333, + "grad_norm": 1.3820863962173462, + "learning_rate": 1.767402220258872e-05, + "loss": 0.3426, + "step": 1263 + }, + { + "epoch": 0.8988444444444444, + "grad_norm": 1.4242292642593384, + "learning_rate": 1.766871324639538e-05, + "loss": 0.4087, + "step": 1264 + }, + { + "epoch": 0.8995555555555556, + "grad_norm": 1.2358907461166382, + "learning_rate": 1.766339903765668e-05, + "loss": 0.3732, + "step": 1265 + }, + { + "epoch": 0.9002666666666667, + "grad_norm": 1.0238115787506104, + "learning_rate": 1.7658079580012494e-05, + "loss": 0.3133, + "step": 1266 + }, + { + "epoch": 0.9009777777777778, + "grad_norm": 1.130966305732727, + "learning_rate": 1.7652754877106275e-05, + "loss": 0.3465, + "step": 1267 + }, + { + "epoch": 0.9016888888888889, + "grad_norm": 1.0897200107574463, + "learning_rate": 1.7647424932585093e-05, + "loss": 0.4133, + "step": 1268 + }, + { + "epoch": 0.9024, + "grad_norm": 1.0519295930862427, + "learning_rate": 1.7642089750099593e-05, + "loss": 0.3958, + "step": 1269 + }, + { + "epoch": 0.9031111111111111, + "grad_norm": 1.8196890354156494, + "learning_rate": 1.7636749333304003e-05, + "loss": 0.4057, + "step": 1270 + }, + { + "epoch": 0.9038222222222222, + "grad_norm": 1.1471799612045288, + "learning_rate": 1.7631403685856154e-05, + "loss": 0.4284, + "step": 1271 + }, + { + "epoch": 0.9045333333333333, + "grad_norm": 1.4079738855361938, + "learning_rate": 1.762605281141745e-05, + "loss": 0.3879, + "step": 1272 + }, + { + "epoch": 0.9052444444444444, + "grad_norm": 1.4291226863861084, + "learning_rate": 1.7620696713652866e-05, + "loss": 0.429, + "step": 1273 + }, + { + "epoch": 0.9059555555555555, + "grad_norm": 1.1242671012878418, + "learning_rate": 1.7615335396230974e-05, + "loss": 0.3652, + "step": 1274 + }, + { + "epoch": 0.9066666666666666, + "grad_norm": 1.2438406944274902, + "learning_rate": 1.7609968862823903e-05, + "loss": 0.4024, + "step": 1275 + }, + { + "epoch": 0.9073777777777777, + "grad_norm": 1.1444185972213745, + "learning_rate": 1.7604597117107363e-05, + "loss": 0.3942, + "step": 1276 + }, + { + "epoch": 0.9080888888888888, + "grad_norm": 1.5018045902252197, + "learning_rate": 1.7599220162760637e-05, + "loss": 0.3747, + "step": 1277 + }, + { + "epoch": 0.9088, + "grad_norm": 1.3934212923049927, + "learning_rate": 1.7593838003466572e-05, + "loss": 0.4313, + "step": 1278 + }, + { + "epoch": 0.9095111111111112, + "grad_norm": 1.5952686071395874, + "learning_rate": 1.7588450642911576e-05, + "loss": 0.4051, + "step": 1279 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 1.1785926818847656, + "learning_rate": 1.7583058084785626e-05, + "loss": 0.3652, + "step": 1280 + }, + { + "epoch": 0.9109333333333334, + "grad_norm": 1.6149747371673584, + "learning_rate": 1.7577660332782257e-05, + "loss": 0.4266, + "step": 1281 + }, + { + "epoch": 0.9116444444444445, + "grad_norm": 1.4176816940307617, + "learning_rate": 1.7572257390598558e-05, + "loss": 0.3738, + "step": 1282 + }, + { + "epoch": 0.9123555555555556, + "grad_norm": 1.6408040523529053, + "learning_rate": 1.7566849261935178e-05, + "loss": 0.431, + "step": 1283 + }, + { + "epoch": 0.9130666666666667, + "grad_norm": 1.5157158374786377, + "learning_rate": 1.7561435950496318e-05, + "loss": 0.3895, + "step": 1284 + }, + { + "epoch": 0.9137777777777778, + "grad_norm": 1.4455909729003906, + "learning_rate": 1.755601745998972e-05, + "loss": 0.3997, + "step": 1285 + }, + { + "epoch": 0.9144888888888889, + "grad_norm": 1.3749761581420898, + "learning_rate": 1.755059379412669e-05, + "loss": 0.375, + "step": 1286 + }, + { + "epoch": 0.9152, + "grad_norm": 1.4814507961273193, + "learning_rate": 1.7545164956622057e-05, + "loss": 0.3931, + "step": 1287 + }, + { + "epoch": 0.9159111111111111, + "grad_norm": 1.1882237195968628, + "learning_rate": 1.7539730951194215e-05, + "loss": 0.3947, + "step": 1288 + }, + { + "epoch": 0.9166222222222222, + "grad_norm": 1.3136929273605347, + "learning_rate": 1.7534291781565084e-05, + "loss": 0.3936, + "step": 1289 + }, + { + "epoch": 0.9173333333333333, + "grad_norm": 1.23666250705719, + "learning_rate": 1.7528847451460124e-05, + "loss": 0.3606, + "step": 1290 + }, + { + "epoch": 0.9180444444444444, + "grad_norm": 1.6042792797088623, + "learning_rate": 1.752339796460833e-05, + "loss": 0.4089, + "step": 1291 + }, + { + "epoch": 0.9187555555555555, + "grad_norm": 1.1069750785827637, + "learning_rate": 1.7517943324742225e-05, + "loss": 0.4026, + "step": 1292 + }, + { + "epoch": 0.9194666666666667, + "grad_norm": 1.6648081541061401, + "learning_rate": 1.7512483535597868e-05, + "loss": 0.4742, + "step": 1293 + }, + { + "epoch": 0.9201777777777778, + "grad_norm": 1.1973766088485718, + "learning_rate": 1.7507018600914846e-05, + "loss": 0.4344, + "step": 1294 + }, + { + "epoch": 0.9208888888888889, + "grad_norm": 1.2127573490142822, + "learning_rate": 1.7501548524436258e-05, + "loss": 0.4235, + "step": 1295 + }, + { + "epoch": 0.9216, + "grad_norm": 1.1892508268356323, + "learning_rate": 1.7496073309908738e-05, + "loss": 0.4187, + "step": 1296 + }, + { + "epoch": 0.9223111111111111, + "grad_norm": 1.694645643234253, + "learning_rate": 1.7490592961082437e-05, + "loss": 0.4001, + "step": 1297 + }, + { + "epoch": 0.9230222222222222, + "grad_norm": 1.0223631858825684, + "learning_rate": 1.7485107481711014e-05, + "loss": 0.3889, + "step": 1298 + }, + { + "epoch": 0.9237333333333333, + "grad_norm": 1.5177525281906128, + "learning_rate": 1.747961687555165e-05, + "loss": 0.3736, + "step": 1299 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.9916736483573914, + "learning_rate": 1.7474121146365037e-05, + "loss": 0.3702, + "step": 1300 + }, + { + "epoch": 0.9251555555555555, + "grad_norm": 1.165610909461975, + "learning_rate": 1.7468620297915378e-05, + "loss": 0.3903, + "step": 1301 + }, + { + "epoch": 0.9258666666666666, + "grad_norm": 1.4972482919692993, + "learning_rate": 1.7463114333970374e-05, + "loss": 0.4422, + "step": 1302 + }, + { + "epoch": 0.9265777777777777, + "grad_norm": 1.4982749223709106, + "learning_rate": 1.7457603258301232e-05, + "loss": 0.3788, + "step": 1303 + }, + { + "epoch": 0.9272888888888889, + "grad_norm": 1.0285675525665283, + "learning_rate": 1.745208707468267e-05, + "loss": 0.418, + "step": 1304 + }, + { + "epoch": 0.928, + "grad_norm": 1.2624262571334839, + "learning_rate": 1.7446565786892894e-05, + "loss": 0.3366, + "step": 1305 + }, + { + "epoch": 0.9287111111111112, + "grad_norm": 1.0978126525878906, + "learning_rate": 1.744103939871361e-05, + "loss": 0.3581, + "step": 1306 + }, + { + "epoch": 0.9294222222222223, + "grad_norm": 1.4762237071990967, + "learning_rate": 1.7435507913930015e-05, + "loss": 0.3711, + "step": 1307 + }, + { + "epoch": 0.9301333333333334, + "grad_norm": 1.527138113975525, + "learning_rate": 1.7429971336330798e-05, + "loss": 0.4038, + "step": 1308 + }, + { + "epoch": 0.9308444444444445, + "grad_norm": 2.1097588539123535, + "learning_rate": 1.742442966970814e-05, + "loss": 0.3979, + "step": 1309 + }, + { + "epoch": 0.9315555555555556, + "grad_norm": 1.5619949102401733, + "learning_rate": 1.7418882917857706e-05, + "loss": 0.3869, + "step": 1310 + }, + { + "epoch": 0.9322666666666667, + "grad_norm": 2.100048780441284, + "learning_rate": 1.7413331084578638e-05, + "loss": 0.3779, + "step": 1311 + }, + { + "epoch": 0.9329777777777778, + "grad_norm": 1.4714446067810059, + "learning_rate": 1.7407774173673567e-05, + "loss": 0.4061, + "step": 1312 + }, + { + "epoch": 0.9336888888888889, + "grad_norm": 1.6887257099151611, + "learning_rate": 1.7402212188948592e-05, + "loss": 0.4383, + "step": 1313 + }, + { + "epoch": 0.9344, + "grad_norm": 1.736033320426941, + "learning_rate": 1.73966451342133e-05, + "loss": 0.4563, + "step": 1314 + }, + { + "epoch": 0.9351111111111111, + "grad_norm": 0.9721370339393616, + "learning_rate": 1.7391073013280738e-05, + "loss": 0.3648, + "step": 1315 + }, + { + "epoch": 0.9358222222222222, + "grad_norm": 1.9666762351989746, + "learning_rate": 1.7385495829967437e-05, + "loss": 0.4278, + "step": 1316 + }, + { + "epoch": 0.9365333333333333, + "grad_norm": 1.1610902547836304, + "learning_rate": 1.7379913588093374e-05, + "loss": 0.3805, + "step": 1317 + }, + { + "epoch": 0.9372444444444444, + "grad_norm": 1.8322278261184692, + "learning_rate": 1.7374326291482016e-05, + "loss": 0.4386, + "step": 1318 + }, + { + "epoch": 0.9379555555555555, + "grad_norm": 1.586854100227356, + "learning_rate": 1.7368733943960278e-05, + "loss": 0.4415, + "step": 1319 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 1.0262722969055176, + "learning_rate": 1.7363136549358527e-05, + "loss": 0.4, + "step": 1320 + }, + { + "epoch": 0.9393777777777778, + "grad_norm": 1.7050113677978516, + "learning_rate": 1.7357534111510604e-05, + "loss": 0.4163, + "step": 1321 + }, + { + "epoch": 0.9400888888888889, + "grad_norm": 1.2788245677947998, + "learning_rate": 1.73519266342538e-05, + "loss": 0.4085, + "step": 1322 + }, + { + "epoch": 0.9408, + "grad_norm": 1.673746109008789, + "learning_rate": 1.7346314121428846e-05, + "loss": 0.4144, + "step": 1323 + }, + { + "epoch": 0.9415111111111111, + "grad_norm": 1.1339434385299683, + "learning_rate": 1.7340696576879938e-05, + "loss": 0.3918, + "step": 1324 + }, + { + "epoch": 0.9422222222222222, + "grad_norm": 1.32035493850708, + "learning_rate": 1.73350740044547e-05, + "loss": 0.4189, + "step": 1325 + }, + { + "epoch": 0.9429333333333333, + "grad_norm": 1.342937707901001, + "learning_rate": 1.7329446408004224e-05, + "loss": 0.3509, + "step": 1326 + }, + { + "epoch": 0.9436444444444444, + "grad_norm": 1.3522717952728271, + "learning_rate": 1.7323813791383026e-05, + "loss": 0.4213, + "step": 1327 + }, + { + "epoch": 0.9443555555555555, + "grad_norm": 1.051612377166748, + "learning_rate": 1.7318176158449057e-05, + "loss": 0.3731, + "step": 1328 + }, + { + "epoch": 0.9450666666666667, + "grad_norm": 1.4662954807281494, + "learning_rate": 1.7312533513063714e-05, + "loss": 0.4172, + "step": 1329 + }, + { + "epoch": 0.9457777777777778, + "grad_norm": 1.1178747415542603, + "learning_rate": 1.7306885859091828e-05, + "loss": 0.4362, + "step": 1330 + }, + { + "epoch": 0.9464888888888889, + "grad_norm": 1.0331188440322876, + "learning_rate": 1.7301233200401656e-05, + "loss": 0.4391, + "step": 1331 + }, + { + "epoch": 0.9472, + "grad_norm": 1.1247471570968628, + "learning_rate": 1.7295575540864878e-05, + "loss": 0.4003, + "step": 1332 + }, + { + "epoch": 0.9479111111111111, + "grad_norm": 0.8947275280952454, + "learning_rate": 1.728991288435661e-05, + "loss": 0.4488, + "step": 1333 + }, + { + "epoch": 0.9486222222222223, + "grad_norm": 1.2471835613250732, + "learning_rate": 1.728424523475538e-05, + "loss": 0.3968, + "step": 1334 + }, + { + "epoch": 0.9493333333333334, + "grad_norm": 1.421459436416626, + "learning_rate": 1.727857259594315e-05, + "loss": 0.4439, + "step": 1335 + }, + { + "epoch": 0.9500444444444445, + "grad_norm": 1.1662904024124146, + "learning_rate": 1.7272894971805283e-05, + "loss": 0.4293, + "step": 1336 + }, + { + "epoch": 0.9507555555555556, + "grad_norm": 0.9959128499031067, + "learning_rate": 1.7267212366230567e-05, + "loss": 0.4267, + "step": 1337 + }, + { + "epoch": 0.9514666666666667, + "grad_norm": 1.364781141281128, + "learning_rate": 1.7261524783111202e-05, + "loss": 0.4079, + "step": 1338 + }, + { + "epoch": 0.9521777777777778, + "grad_norm": 1.6053006649017334, + "learning_rate": 1.725583222634279e-05, + "loss": 0.4266, + "step": 1339 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 1.0602662563323975, + "learning_rate": 1.7250134699824343e-05, + "loss": 0.3438, + "step": 1340 + }, + { + "epoch": 0.9536, + "grad_norm": 1.4582699537277222, + "learning_rate": 1.724443220745828e-05, + "loss": 0.399, + "step": 1341 + }, + { + "epoch": 0.9543111111111111, + "grad_norm": 1.011934757232666, + "learning_rate": 1.723872475315042e-05, + "loss": 0.369, + "step": 1342 + }, + { + "epoch": 0.9550222222222222, + "grad_norm": 1.2250741720199585, + "learning_rate": 1.7233012340809983e-05, + "loss": 0.396, + "step": 1343 + }, + { + "epoch": 0.9557333333333333, + "grad_norm": 0.9501094818115234, + "learning_rate": 1.7227294974349576e-05, + "loss": 0.3862, + "step": 1344 + }, + { + "epoch": 0.9564444444444444, + "grad_norm": 1.3885756731033325, + "learning_rate": 1.7221572657685205e-05, + "loss": 0.416, + "step": 1345 + }, + { + "epoch": 0.9571555555555555, + "grad_norm": 1.0543265342712402, + "learning_rate": 1.7215845394736268e-05, + "loss": 0.44, + "step": 1346 + }, + { + "epoch": 0.9578666666666666, + "grad_norm": 1.3239190578460693, + "learning_rate": 1.721011318942555e-05, + "loss": 0.4437, + "step": 1347 + }, + { + "epoch": 0.9585777777777778, + "grad_norm": 1.045580506324768, + "learning_rate": 1.7204376045679225e-05, + "loss": 0.4529, + "step": 1348 + }, + { + "epoch": 0.9592888888888889, + "grad_norm": 1.4350066184997559, + "learning_rate": 1.7198633967426838e-05, + "loss": 0.4176, + "step": 1349 + }, + { + "epoch": 0.96, + "grad_norm": 1.126920223236084, + "learning_rate": 1.719288695860132e-05, + "loss": 0.4002, + "step": 1350 + }, + { + "epoch": 0.9607111111111111, + "grad_norm": 1.4288395643234253, + "learning_rate": 1.7187135023138984e-05, + "loss": 0.4255, + "step": 1351 + }, + { + "epoch": 0.9614222222222222, + "grad_norm": 1.8058830499649048, + "learning_rate": 1.7181378164979515e-05, + "loss": 0.4297, + "step": 1352 + }, + { + "epoch": 0.9621333333333333, + "grad_norm": 1.2865937948226929, + "learning_rate": 1.7175616388065965e-05, + "loss": 0.3584, + "step": 1353 + }, + { + "epoch": 0.9628444444444444, + "grad_norm": 1.891649603843689, + "learning_rate": 1.7169849696344756e-05, + "loss": 0.422, + "step": 1354 + }, + { + "epoch": 0.9635555555555556, + "grad_norm": 1.5911121368408203, + "learning_rate": 1.7164078093765684e-05, + "loss": 0.4198, + "step": 1355 + }, + { + "epoch": 0.9642666666666667, + "grad_norm": 1.1915384531021118, + "learning_rate": 1.7158301584281898e-05, + "loss": 0.3965, + "step": 1356 + }, + { + "epoch": 0.9649777777777778, + "grad_norm": 1.9912608861923218, + "learning_rate": 1.7152520171849917e-05, + "loss": 0.4507, + "step": 1357 + }, + { + "epoch": 0.9656888888888889, + "grad_norm": 1.0680032968521118, + "learning_rate": 1.7146733860429614e-05, + "loss": 0.4383, + "step": 1358 + }, + { + "epoch": 0.9664, + "grad_norm": 2.6099066734313965, + "learning_rate": 1.714094265398421e-05, + "loss": 0.4273, + "step": 1359 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 1.5917364358901978, + "learning_rate": 1.7135146556480294e-05, + "loss": 0.4312, + "step": 1360 + }, + { + "epoch": 0.9678222222222223, + "grad_norm": 2.3587615489959717, + "learning_rate": 1.7129345571887795e-05, + "loss": 0.3873, + "step": 1361 + }, + { + "epoch": 0.9685333333333334, + "grad_norm": 1.9594316482543945, + "learning_rate": 1.7123539704179993e-05, + "loss": 0.4214, + "step": 1362 + }, + { + "epoch": 0.9692444444444445, + "grad_norm": 1.2614094018936157, + "learning_rate": 1.7117728957333507e-05, + "loss": 0.415, + "step": 1363 + }, + { + "epoch": 0.9699555555555556, + "grad_norm": 1.8267403841018677, + "learning_rate": 1.711191333532831e-05, + "loss": 0.4276, + "step": 1364 + }, + { + "epoch": 0.9706666666666667, + "grad_norm": 0.9699957370758057, + "learning_rate": 1.71060928421477e-05, + "loss": 0.3393, + "step": 1365 + }, + { + "epoch": 0.9713777777777778, + "grad_norm": 1.7892223596572876, + "learning_rate": 1.7100267481778316e-05, + "loss": 0.4044, + "step": 1366 + }, + { + "epoch": 0.9720888888888889, + "grad_norm": 1.554416537284851, + "learning_rate": 1.709443725821014e-05, + "loss": 0.3765, + "step": 1367 + }, + { + "epoch": 0.9728, + "grad_norm": 1.63836669921875, + "learning_rate": 1.7088602175436468e-05, + "loss": 0.43, + "step": 1368 + }, + { + "epoch": 0.9735111111111111, + "grad_norm": 2.079576253890991, + "learning_rate": 1.708276223745394e-05, + "loss": 0.3408, + "step": 1369 + }, + { + "epoch": 0.9742222222222222, + "grad_norm": 1.1428107023239136, + "learning_rate": 1.7076917448262516e-05, + "loss": 0.3637, + "step": 1370 + }, + { + "epoch": 0.9749333333333333, + "grad_norm": 1.3114690780639648, + "learning_rate": 1.7071067811865477e-05, + "loss": 0.3979, + "step": 1371 + }, + { + "epoch": 0.9756444444444444, + "grad_norm": 1.1474499702453613, + "learning_rate": 1.7065213332269422e-05, + "loss": 0.3692, + "step": 1372 + }, + { + "epoch": 0.9763555555555555, + "grad_norm": 1.8865656852722168, + "learning_rate": 1.705935401348427e-05, + "loss": 0.3894, + "step": 1373 + }, + { + "epoch": 0.9770666666666666, + "grad_norm": 1.8120735883712769, + "learning_rate": 1.7053489859523256e-05, + "loss": 0.4023, + "step": 1374 + }, + { + "epoch": 0.9777777777777777, + "grad_norm": 1.4487820863723755, + "learning_rate": 1.7047620874402927e-05, + "loss": 0.3906, + "step": 1375 + }, + { + "epoch": 0.9784888888888889, + "grad_norm": 1.5597162246704102, + "learning_rate": 1.7041747062143136e-05, + "loss": 0.4307, + "step": 1376 + }, + { + "epoch": 0.9792, + "grad_norm": 1.0199686288833618, + "learning_rate": 1.703586842676704e-05, + "loss": 0.3401, + "step": 1377 + }, + { + "epoch": 0.9799111111111111, + "grad_norm": 1.8015869855880737, + "learning_rate": 1.7029984972301108e-05, + "loss": 0.3693, + "step": 1378 + }, + { + "epoch": 0.9806222222222222, + "grad_norm": 1.2282450199127197, + "learning_rate": 1.7024096702775102e-05, + "loss": 0.3651, + "step": 1379 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 1.080920934677124, + "learning_rate": 1.701820362222209e-05, + "loss": 0.3886, + "step": 1380 + }, + { + "epoch": 0.9820444444444445, + "grad_norm": 1.6381980180740356, + "learning_rate": 1.7012305734678425e-05, + "loss": 0.3754, + "step": 1381 + }, + { + "epoch": 0.9827555555555556, + "grad_norm": 1.258268117904663, + "learning_rate": 1.7006403044183758e-05, + "loss": 0.3651, + "step": 1382 + }, + { + "epoch": 0.9834666666666667, + "grad_norm": 1.1745879650115967, + "learning_rate": 1.700049555478103e-05, + "loss": 0.4301, + "step": 1383 + }, + { + "epoch": 0.9841777777777778, + "grad_norm": 1.3173675537109375, + "learning_rate": 1.699458327051647e-05, + "loss": 0.4047, + "step": 1384 + }, + { + "epoch": 0.9848888888888889, + "grad_norm": 1.65962815284729, + "learning_rate": 1.698866619543959e-05, + "loss": 0.3942, + "step": 1385 + }, + { + "epoch": 0.9856, + "grad_norm": 1.2486056089401245, + "learning_rate": 1.6982744333603177e-05, + "loss": 0.4138, + "step": 1386 + }, + { + "epoch": 0.9863111111111111, + "grad_norm": 2.052551031112671, + "learning_rate": 1.6976817689063308e-05, + "loss": 0.417, + "step": 1387 + }, + { + "epoch": 0.9870222222222222, + "grad_norm": 1.5529996156692505, + "learning_rate": 1.6970886265879328e-05, + "loss": 0.3966, + "step": 1388 + }, + { + "epoch": 0.9877333333333334, + "grad_norm": 2.0040957927703857, + "learning_rate": 1.6964950068113854e-05, + "loss": 0.3883, + "step": 1389 + }, + { + "epoch": 0.9884444444444445, + "grad_norm": 1.4844411611557007, + "learning_rate": 1.6959009099832787e-05, + "loss": 0.4316, + "step": 1390 + }, + { + "epoch": 0.9891555555555556, + "grad_norm": 1.1441562175750732, + "learning_rate": 1.6953063365105272e-05, + "loss": 0.3807, + "step": 1391 + }, + { + "epoch": 0.9898666666666667, + "grad_norm": 1.5123248100280762, + "learning_rate": 1.6947112868003737e-05, + "loss": 0.3587, + "step": 1392 + }, + { + "epoch": 0.9905777777777778, + "grad_norm": 1.3816200494766235, + "learning_rate": 1.694115761260387e-05, + "loss": 0.4131, + "step": 1393 + }, + { + "epoch": 0.9912888888888889, + "grad_norm": 1.3448748588562012, + "learning_rate": 1.6935197602984607e-05, + "loss": 0.404, + "step": 1394 + }, + { + "epoch": 0.992, + "grad_norm": 2.2939116954803467, + "learning_rate": 1.6929232843228155e-05, + "loss": 0.3966, + "step": 1395 + }, + { + "epoch": 0.9927111111111111, + "grad_norm": 1.8636271953582764, + "learning_rate": 1.6923263337419965e-05, + "loss": 0.4133, + "step": 1396 + }, + { + "epoch": 0.9934222222222222, + "grad_norm": 1.955564022064209, + "learning_rate": 1.6917289089648742e-05, + "loss": 0.3679, + "step": 1397 + }, + { + "epoch": 0.9941333333333333, + "grad_norm": 1.9916272163391113, + "learning_rate": 1.6911310104006433e-05, + "loss": 0.4407, + "step": 1398 + }, + { + "epoch": 0.9948444444444444, + "grad_norm": 1.0206936597824097, + "learning_rate": 1.6905326384588242e-05, + "loss": 0.4056, + "step": 1399 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 3.044020652770996, + "learning_rate": 1.6899337935492607e-05, + "loss": 0.3429, + "step": 1400 + }, + { + "epoch": 0.9962666666666666, + "grad_norm": 2.9919373989105225, + "learning_rate": 1.6893344760821205e-05, + "loss": 0.3791, + "step": 1401 + }, + { + "epoch": 0.9969777777777777, + "grad_norm": 1.34165620803833, + "learning_rate": 1.6887346864678953e-05, + "loss": 0.4283, + "step": 1402 + }, + { + "epoch": 0.9976888888888888, + "grad_norm": 1.7959727048873901, + "learning_rate": 1.6881344251174004e-05, + "loss": 0.4006, + "step": 1403 + }, + { + "epoch": 0.9984, + "grad_norm": 1.748821496963501, + "learning_rate": 1.687533692441773e-05, + "loss": 0.4014, + "step": 1404 + }, + { + "epoch": 0.9991111111111111, + "grad_norm": 1.2351717948913574, + "learning_rate": 1.6869324888524748e-05, + "loss": 0.3876, + "step": 1405 + }, + { + "epoch": 0.9998222222222222, + "grad_norm": 1.443494439125061, + "learning_rate": 1.686330814761289e-05, + "loss": 0.4145, + "step": 1406 + }, + { + "epoch": 1.0005333333333333, + "grad_norm": 1.6163692474365234, + "learning_rate": 1.685728670580321e-05, + "loss": 0.3663, + "step": 1407 + }, + { + "epoch": 1.0012444444444444, + "grad_norm": 1.582128882408142, + "learning_rate": 1.685126056721999e-05, + "loss": 0.365, + "step": 1408 + }, + { + "epoch": 1.0019555555555555, + "grad_norm": 1.407363772392273, + "learning_rate": 1.684522973599072e-05, + "loss": 0.4649, + "step": 1409 + }, + { + "epoch": 1.0026666666666666, + "grad_norm": 1.31049644947052, + "learning_rate": 1.683919421624611e-05, + "loss": 0.3957, + "step": 1410 + }, + { + "epoch": 1.0033777777777777, + "grad_norm": 1.235746145248413, + "learning_rate": 1.6833154012120077e-05, + "loss": 0.3771, + "step": 1411 + }, + { + "epoch": 1.0040888888888888, + "grad_norm": 2.105921983718872, + "learning_rate": 1.682710912774975e-05, + "loss": 0.3731, + "step": 1412 + }, + { + "epoch": 1.0048, + "grad_norm": 1.0362101793289185, + "learning_rate": 1.6821059567275463e-05, + "loss": 0.3884, + "step": 1413 + }, + { + "epoch": 1.005511111111111, + "grad_norm": 1.8378596305847168, + "learning_rate": 1.681500533484075e-05, + "loss": 0.4151, + "step": 1414 + }, + { + "epoch": 1.0062222222222221, + "grad_norm": 1.299248456954956, + "learning_rate": 1.6808946434592348e-05, + "loss": 0.3641, + "step": 1415 + }, + { + "epoch": 1.0069333333333332, + "grad_norm": 1.457021713256836, + "learning_rate": 1.680288287068019e-05, + "loss": 0.3281, + "step": 1416 + }, + { + "epoch": 1.0076444444444443, + "grad_norm": 1.7000689506530762, + "learning_rate": 1.679681464725741e-05, + "loss": 0.3642, + "step": 1417 + }, + { + "epoch": 1.0083555555555554, + "grad_norm": 1.4647246599197388, + "learning_rate": 1.6790741768480322e-05, + "loss": 0.3836, + "step": 1418 + }, + { + "epoch": 1.0090666666666666, + "grad_norm": 1.2759373188018799, + "learning_rate": 1.678466423850843e-05, + "loss": 0.3285, + "step": 1419 + }, + { + "epoch": 1.0097777777777779, + "grad_norm": 1.357131004333496, + "learning_rate": 1.677858206150443e-05, + "loss": 0.3672, + "step": 1420 + }, + { + "epoch": 1.010488888888889, + "grad_norm": 1.355940580368042, + "learning_rate": 1.6772495241634202e-05, + "loss": 0.3736, + "step": 1421 + }, + { + "epoch": 1.0112, + "grad_norm": 0.9268417358398438, + "learning_rate": 1.67664037830668e-05, + "loss": 0.3627, + "step": 1422 + }, + { + "epoch": 1.0119111111111112, + "grad_norm": 2.0063180923461914, + "learning_rate": 1.676030768997445e-05, + "loss": 0.3392, + "step": 1423 + }, + { + "epoch": 1.0126222222222223, + "grad_norm": 1.0902903079986572, + "learning_rate": 1.6754206966532575e-05, + "loss": 0.372, + "step": 1424 + }, + { + "epoch": 1.0133333333333334, + "grad_norm": 1.5008772611618042, + "learning_rate": 1.674810161691974e-05, + "loss": 0.3835, + "step": 1425 + }, + { + "epoch": 1.0140444444444445, + "grad_norm": 1.349814772605896, + "learning_rate": 1.6741991645317698e-05, + "loss": 0.3724, + "step": 1426 + }, + { + "epoch": 1.0147555555555556, + "grad_norm": 0.8739634156227112, + "learning_rate": 1.6735877055911368e-05, + "loss": 0.3497, + "step": 1427 + }, + { + "epoch": 1.0154666666666667, + "grad_norm": 1.1868383884429932, + "learning_rate": 1.672975785288882e-05, + "loss": 0.4187, + "step": 1428 + }, + { + "epoch": 1.0161777777777778, + "grad_norm": 1.3832119703292847, + "learning_rate": 1.6723634040441293e-05, + "loss": 0.387, + "step": 1429 + }, + { + "epoch": 1.016888888888889, + "grad_norm": 1.025839924812317, + "learning_rate": 1.6717505622763184e-05, + "loss": 0.3461, + "step": 1430 + }, + { + "epoch": 1.0176, + "grad_norm": 1.0739519596099854, + "learning_rate": 1.6711372604052033e-05, + "loss": 0.3907, + "step": 1431 + }, + { + "epoch": 1.0183111111111112, + "grad_norm": 0.9434837698936462, + "learning_rate": 1.6705234988508544e-05, + "loss": 0.3505, + "step": 1432 + }, + { + "epoch": 1.0190222222222223, + "grad_norm": 1.1662428379058838, + "learning_rate": 1.6699092780336566e-05, + "loss": 0.358, + "step": 1433 + }, + { + "epoch": 1.0197333333333334, + "grad_norm": 1.1528176069259644, + "learning_rate": 1.6692945983743093e-05, + "loss": 0.3241, + "step": 1434 + }, + { + "epoch": 1.0204444444444445, + "grad_norm": 1.7034268379211426, + "learning_rate": 1.6686794602938253e-05, + "loss": 0.3551, + "step": 1435 + }, + { + "epoch": 1.0211555555555556, + "grad_norm": 1.3942110538482666, + "learning_rate": 1.6680638642135337e-05, + "loss": 0.3972, + "step": 1436 + }, + { + "epoch": 1.0218666666666667, + "grad_norm": 1.7678548097610474, + "learning_rate": 1.6674478105550745e-05, + "loss": 0.4101, + "step": 1437 + }, + { + "epoch": 1.0225777777777778, + "grad_norm": 1.4009203910827637, + "learning_rate": 1.666831299740403e-05, + "loss": 0.3847, + "step": 1438 + }, + { + "epoch": 1.023288888888889, + "grad_norm": 1.2953382730484009, + "learning_rate": 1.6662143321917872e-05, + "loss": 0.3976, + "step": 1439 + }, + { + "epoch": 1.024, + "grad_norm": 1.4829137325286865, + "learning_rate": 1.665596908331807e-05, + "loss": 0.3653, + "step": 1440 + }, + { + "epoch": 1.0247111111111111, + "grad_norm": 1.1819709539413452, + "learning_rate": 1.6649790285833566e-05, + "loss": 0.4213, + "step": 1441 + }, + { + "epoch": 1.0254222222222222, + "grad_norm": 1.3849505186080933, + "learning_rate": 1.664360693369641e-05, + "loss": 0.3619, + "step": 1442 + }, + { + "epoch": 1.0261333333333333, + "grad_norm": 1.6494675874710083, + "learning_rate": 1.6637419031141776e-05, + "loss": 0.4537, + "step": 1443 + }, + { + "epoch": 1.0268444444444444, + "grad_norm": 1.4870827198028564, + "learning_rate": 1.6631226582407954e-05, + "loss": 0.4301, + "step": 1444 + }, + { + "epoch": 1.0275555555555556, + "grad_norm": 1.5517767667770386, + "learning_rate": 1.6625029591736353e-05, + "loss": 0.3649, + "step": 1445 + }, + { + "epoch": 1.0282666666666667, + "grad_norm": 1.1304274797439575, + "learning_rate": 1.6618828063371482e-05, + "loss": 0.3984, + "step": 1446 + }, + { + "epoch": 1.0289777777777778, + "grad_norm": 1.001749038696289, + "learning_rate": 1.6612622001560974e-05, + "loss": 0.4208, + "step": 1447 + }, + { + "epoch": 1.0296888888888889, + "grad_norm": 1.9795804023742676, + "learning_rate": 1.6606411410555552e-05, + "loss": 0.3373, + "step": 1448 + }, + { + "epoch": 1.0304, + "grad_norm": 0.9525133967399597, + "learning_rate": 1.6600196294609047e-05, + "loss": 0.3405, + "step": 1449 + }, + { + "epoch": 1.031111111111111, + "grad_norm": 2.3432202339172363, + "learning_rate": 1.659397665797839e-05, + "loss": 0.4142, + "step": 1450 + }, + { + "epoch": 1.0318222222222222, + "grad_norm": 1.3999598026275635, + "learning_rate": 1.658775250492361e-05, + "loss": 0.389, + "step": 1451 + }, + { + "epoch": 1.0325333333333333, + "grad_norm": 1.7075377702713013, + "learning_rate": 1.6581523839707823e-05, + "loss": 0.4068, + "step": 1452 + }, + { + "epoch": 1.0332444444444444, + "grad_norm": 1.4066457748413086, + "learning_rate": 1.6575290666597246e-05, + "loss": 0.3897, + "step": 1453 + }, + { + "epoch": 1.0339555555555555, + "grad_norm": 1.158903956413269, + "learning_rate": 1.656905298986117e-05, + "loss": 0.3576, + "step": 1454 + }, + { + "epoch": 1.0346666666666666, + "grad_norm": 1.3930326700210571, + "learning_rate": 1.6562810813771983e-05, + "loss": 0.3639, + "step": 1455 + }, + { + "epoch": 1.0353777777777777, + "grad_norm": 1.027299165725708, + "learning_rate": 1.655656414260515e-05, + "loss": 0.3742, + "step": 1456 + }, + { + "epoch": 1.0360888888888888, + "grad_norm": 1.3871808052062988, + "learning_rate": 1.655031298063921e-05, + "loss": 0.3631, + "step": 1457 + }, + { + "epoch": 1.0368, + "grad_norm": 1.2734557390213013, + "learning_rate": 1.6544057332155787e-05, + "loss": 0.3846, + "step": 1458 + }, + { + "epoch": 1.037511111111111, + "grad_norm": 1.0770387649536133, + "learning_rate": 1.653779720143957e-05, + "loss": 0.4124, + "step": 1459 + }, + { + "epoch": 1.0382222222222222, + "grad_norm": 1.2208340167999268, + "learning_rate": 1.653153259277833e-05, + "loss": 0.3593, + "step": 1460 + }, + { + "epoch": 1.0389333333333333, + "grad_norm": 1.353206753730774, + "learning_rate": 1.6525263510462882e-05, + "loss": 0.3632, + "step": 1461 + }, + { + "epoch": 1.0396444444444444, + "grad_norm": 1.14902663230896, + "learning_rate": 1.6518989958787126e-05, + "loss": 0.3906, + "step": 1462 + }, + { + "epoch": 1.0403555555555555, + "grad_norm": 2.0567171573638916, + "learning_rate": 1.6512711942048023e-05, + "loss": 0.3775, + "step": 1463 + }, + { + "epoch": 1.0410666666666666, + "grad_norm": 1.2517986297607422, + "learning_rate": 1.650642946454558e-05, + "loss": 0.3889, + "step": 1464 + }, + { + "epoch": 1.0417777777777777, + "grad_norm": 1.1523346900939941, + "learning_rate": 1.6500142530582864e-05, + "loss": 0.3721, + "step": 1465 + }, + { + "epoch": 1.0424888888888888, + "grad_norm": 1.5854793787002563, + "learning_rate": 1.6493851144466003e-05, + "loss": 0.3882, + "step": 1466 + }, + { + "epoch": 1.0432, + "grad_norm": 1.158118724822998, + "learning_rate": 1.6487555310504164e-05, + "loss": 0.3797, + "step": 1467 + }, + { + "epoch": 1.043911111111111, + "grad_norm": 1.648210048675537, + "learning_rate": 1.6481255033009556e-05, + "loss": 0.4202, + "step": 1468 + }, + { + "epoch": 1.0446222222222221, + "grad_norm": 1.2021180391311646, + "learning_rate": 1.6474950316297448e-05, + "loss": 0.3234, + "step": 1469 + }, + { + "epoch": 1.0453333333333332, + "grad_norm": 1.4078212976455688, + "learning_rate": 1.646864116468614e-05, + "loss": 0.3837, + "step": 1470 + }, + { + "epoch": 1.0460444444444446, + "grad_norm": 1.3696666955947876, + "learning_rate": 1.646232758249697e-05, + "loss": 0.3931, + "step": 1471 + }, + { + "epoch": 1.0467555555555557, + "grad_norm": 2.1197900772094727, + "learning_rate": 1.6456009574054307e-05, + "loss": 0.3194, + "step": 1472 + }, + { + "epoch": 1.0474666666666668, + "grad_norm": 1.828758716583252, + "learning_rate": 1.6449687143685564e-05, + "loss": 0.3702, + "step": 1473 + }, + { + "epoch": 1.0481777777777779, + "grad_norm": 1.4040167331695557, + "learning_rate": 1.6443360295721163e-05, + "loss": 0.3765, + "step": 1474 + }, + { + "epoch": 1.048888888888889, + "grad_norm": 1.1777747869491577, + "learning_rate": 1.6437029034494574e-05, + "loss": 0.3216, + "step": 1475 + }, + { + "epoch": 1.0496, + "grad_norm": 1.243025302886963, + "learning_rate": 1.643069336434227e-05, + "loss": 0.3893, + "step": 1476 + }, + { + "epoch": 1.0503111111111112, + "grad_norm": 1.8576377630233765, + "learning_rate": 1.6424353289603763e-05, + "loss": 0.3415, + "step": 1477 + }, + { + "epoch": 1.0510222222222223, + "grad_norm": 1.2895561456680298, + "learning_rate": 1.6418008814621563e-05, + "loss": 0.356, + "step": 1478 + }, + { + "epoch": 1.0517333333333334, + "grad_norm": 1.914525032043457, + "learning_rate": 1.6411659943741205e-05, + "loss": 0.3661, + "step": 1479 + }, + { + "epoch": 1.0524444444444445, + "grad_norm": 1.448792815208435, + "learning_rate": 1.640530668131123e-05, + "loss": 0.3544, + "step": 1480 + }, + { + "epoch": 1.0531555555555556, + "grad_norm": 2.221672773361206, + "learning_rate": 1.6398949031683197e-05, + "loss": 0.3573, + "step": 1481 + }, + { + "epoch": 1.0538666666666667, + "grad_norm": 1.833411455154419, + "learning_rate": 1.6392586999211654e-05, + "loss": 0.3999, + "step": 1482 + }, + { + "epoch": 1.0545777777777778, + "grad_norm": 1.7201879024505615, + "learning_rate": 1.6386220588254164e-05, + "loss": 0.3942, + "step": 1483 + }, + { + "epoch": 1.055288888888889, + "grad_norm": 2.046027660369873, + "learning_rate": 1.637984980317128e-05, + "loss": 0.3483, + "step": 1484 + }, + { + "epoch": 1.056, + "grad_norm": 1.6656941175460815, + "learning_rate": 1.637347464832656e-05, + "loss": 0.3597, + "step": 1485 + }, + { + "epoch": 1.0567111111111112, + "grad_norm": 1.869568943977356, + "learning_rate": 1.6367095128086544e-05, + "loss": 0.3838, + "step": 1486 + }, + { + "epoch": 1.0574222222222223, + "grad_norm": 2.173246145248413, + "learning_rate": 1.6360711246820777e-05, + "loss": 0.3119, + "step": 1487 + }, + { + "epoch": 1.0581333333333334, + "grad_norm": 1.326370358467102, + "learning_rate": 1.6354323008901774e-05, + "loss": 0.4022, + "step": 1488 + }, + { + "epoch": 1.0588444444444445, + "grad_norm": 2.4347167015075684, + "learning_rate": 1.6347930418705047e-05, + "loss": 0.4045, + "step": 1489 + }, + { + "epoch": 1.0595555555555556, + "grad_norm": 2.291299343109131, + "learning_rate": 1.6341533480609082e-05, + "loss": 0.3602, + "step": 1490 + }, + { + "epoch": 1.0602666666666667, + "grad_norm": 1.2356294393539429, + "learning_rate": 1.6335132198995344e-05, + "loss": 0.3437, + "step": 1491 + }, + { + "epoch": 1.0609777777777778, + "grad_norm": 1.675897240638733, + "learning_rate": 1.632872657824828e-05, + "loss": 0.3385, + "step": 1492 + }, + { + "epoch": 1.061688888888889, + "grad_norm": 1.345184087753296, + "learning_rate": 1.6322316622755294e-05, + "loss": 0.3707, + "step": 1493 + }, + { + "epoch": 1.0624, + "grad_norm": 1.1538026332855225, + "learning_rate": 1.6315902336906775e-05, + "loss": 0.3603, + "step": 1494 + }, + { + "epoch": 1.0631111111111111, + "grad_norm": 1.2303919792175293, + "learning_rate": 1.630948372509607e-05, + "loss": 0.3963, + "step": 1495 + }, + { + "epoch": 1.0638222222222222, + "grad_norm": 1.7455172538757324, + "learning_rate": 1.6303060791719487e-05, + "loss": 0.3609, + "step": 1496 + }, + { + "epoch": 1.0645333333333333, + "grad_norm": 1.4888566732406616, + "learning_rate": 1.62966335411763e-05, + "loss": 0.4198, + "step": 1497 + }, + { + "epoch": 1.0652444444444444, + "grad_norm": 1.459816575050354, + "learning_rate": 1.6290201977868738e-05, + "loss": 0.3598, + "step": 1498 + }, + { + "epoch": 1.0659555555555555, + "grad_norm": 1.0855525732040405, + "learning_rate": 1.628376610620198e-05, + "loss": 0.3835, + "step": 1499 + }, + { + "epoch": 1.0666666666666667, + "grad_norm": 1.6785037517547607, + "learning_rate": 1.627732593058416e-05, + "loss": 0.3637, + "step": 1500 + }, + { + "epoch": 1.0673777777777778, + "grad_norm": 1.3206483125686646, + "learning_rate": 1.6270881455426358e-05, + "loss": 0.409, + "step": 1501 + }, + { + "epoch": 1.0680888888888889, + "grad_norm": 1.4702718257904053, + "learning_rate": 1.6264432685142603e-05, + "loss": 0.3443, + "step": 1502 + }, + { + "epoch": 1.0688, + "grad_norm": 1.4314407110214233, + "learning_rate": 1.6257979624149863e-05, + "loss": 0.3809, + "step": 1503 + }, + { + "epoch": 1.069511111111111, + "grad_norm": 1.238577961921692, + "learning_rate": 1.6251522276868042e-05, + "loss": 0.3524, + "step": 1504 + }, + { + "epoch": 1.0702222222222222, + "grad_norm": 1.6288745403289795, + "learning_rate": 1.6245060647719985e-05, + "loss": 0.3723, + "step": 1505 + }, + { + "epoch": 1.0709333333333333, + "grad_norm": 1.3678317070007324, + "learning_rate": 1.6238594741131458e-05, + "loss": 0.3734, + "step": 1506 + }, + { + "epoch": 1.0716444444444444, + "grad_norm": 1.5155709981918335, + "learning_rate": 1.6232124561531177e-05, + "loss": 0.3997, + "step": 1507 + }, + { + "epoch": 1.0723555555555555, + "grad_norm": 1.604856252670288, + "learning_rate": 1.6225650113350772e-05, + "loss": 0.3487, + "step": 1508 + }, + { + "epoch": 1.0730666666666666, + "grad_norm": 1.582165241241455, + "learning_rate": 1.621917140102479e-05, + "loss": 0.3621, + "step": 1509 + }, + { + "epoch": 1.0737777777777777, + "grad_norm": 1.8324062824249268, + "learning_rate": 1.6212688428990717e-05, + "loss": 0.4144, + "step": 1510 + }, + { + "epoch": 1.0744888888888888, + "grad_norm": 1.464176058769226, + "learning_rate": 1.6206201201688943e-05, + "loss": 0.4307, + "step": 1511 + }, + { + "epoch": 1.0752, + "grad_norm": 1.433565616607666, + "learning_rate": 1.6199709723562772e-05, + "loss": 0.391, + "step": 1512 + }, + { + "epoch": 1.075911111111111, + "grad_norm": 1.0623693466186523, + "learning_rate": 1.6193213999058426e-05, + "loss": 0.3777, + "step": 1513 + }, + { + "epoch": 1.0766222222222221, + "grad_norm": 1.0461713075637817, + "learning_rate": 1.6186714032625036e-05, + "loss": 0.3817, + "step": 1514 + }, + { + "epoch": 1.0773333333333333, + "grad_norm": 1.0895780324935913, + "learning_rate": 1.6180209828714633e-05, + "loss": 0.3945, + "step": 1515 + }, + { + "epoch": 1.0780444444444444, + "grad_norm": 1.0859469175338745, + "learning_rate": 1.6173701391782154e-05, + "loss": 0.3384, + "step": 1516 + }, + { + "epoch": 1.0787555555555555, + "grad_norm": 1.1009870767593384, + "learning_rate": 1.6167188726285433e-05, + "loss": 0.3417, + "step": 1517 + }, + { + "epoch": 1.0794666666666666, + "grad_norm": 1.2068406343460083, + "learning_rate": 1.61606718366852e-05, + "loss": 0.3608, + "step": 1518 + }, + { + "epoch": 1.0801777777777777, + "grad_norm": 1.4417222738265991, + "learning_rate": 1.6154150727445084e-05, + "loss": 0.3725, + "step": 1519 + }, + { + "epoch": 1.0808888888888888, + "grad_norm": 1.209671974182129, + "learning_rate": 1.6147625403031598e-05, + "loss": 0.3603, + "step": 1520 + }, + { + "epoch": 1.0816, + "grad_norm": 1.3988327980041504, + "learning_rate": 1.6141095867914146e-05, + "loss": 0.3618, + "step": 1521 + }, + { + "epoch": 1.0823111111111112, + "grad_norm": 1.2390469312667847, + "learning_rate": 1.6134562126565007e-05, + "loss": 0.3492, + "step": 1522 + }, + { + "epoch": 1.083022222222222, + "grad_norm": 1.2718360424041748, + "learning_rate": 1.6128024183459355e-05, + "loss": 0.3576, + "step": 1523 + }, + { + "epoch": 1.0837333333333334, + "grad_norm": 1.0481328964233398, + "learning_rate": 1.6121482043075233e-05, + "loss": 0.3775, + "step": 1524 + }, + { + "epoch": 1.0844444444444445, + "grad_norm": 1.0894217491149902, + "learning_rate": 1.6114935709893566e-05, + "loss": 0.3885, + "step": 1525 + }, + { + "epoch": 1.0851555555555557, + "grad_norm": 1.2993861436843872, + "learning_rate": 1.610838518839813e-05, + "loss": 0.3524, + "step": 1526 + }, + { + "epoch": 1.0858666666666668, + "grad_norm": 1.0059324502944946, + "learning_rate": 1.6101830483075603e-05, + "loss": 0.3438, + "step": 1527 + }, + { + "epoch": 1.0865777777777779, + "grad_norm": 1.2559374570846558, + "learning_rate": 1.60952715984155e-05, + "loss": 0.355, + "step": 1528 + }, + { + "epoch": 1.087288888888889, + "grad_norm": 1.3722953796386719, + "learning_rate": 1.6088708538910217e-05, + "loss": 0.3571, + "step": 1529 + }, + { + "epoch": 1.088, + "grad_norm": 1.212699055671692, + "learning_rate": 1.6082141309054996e-05, + "loss": 0.4429, + "step": 1530 + }, + { + "epoch": 1.0887111111111112, + "grad_norm": 1.4237638711929321, + "learning_rate": 1.6075569913347947e-05, + "loss": 0.4182, + "step": 1531 + }, + { + "epoch": 1.0894222222222223, + "grad_norm": 0.9675988554954529, + "learning_rate": 1.6068994356290018e-05, + "loss": 0.3754, + "step": 1532 + }, + { + "epoch": 1.0901333333333334, + "grad_norm": 1.2506200075149536, + "learning_rate": 1.6062414642385026e-05, + "loss": 0.398, + "step": 1533 + }, + { + "epoch": 1.0908444444444445, + "grad_norm": 1.06365966796875, + "learning_rate": 1.6055830776139622e-05, + "loss": 0.3845, + "step": 1534 + }, + { + "epoch": 1.0915555555555556, + "grad_norm": 1.046570897102356, + "learning_rate": 1.604924276206331e-05, + "loss": 0.3318, + "step": 1535 + }, + { + "epoch": 1.0922666666666667, + "grad_norm": 1.1827338933944702, + "learning_rate": 1.6042650604668423e-05, + "loss": 0.398, + "step": 1536 + }, + { + "epoch": 1.0929777777777778, + "grad_norm": 1.2510976791381836, + "learning_rate": 1.6036054308470143e-05, + "loss": 0.392, + "step": 1537 + }, + { + "epoch": 1.093688888888889, + "grad_norm": 1.0307379961013794, + "learning_rate": 1.602945387798648e-05, + "loss": 0.3134, + "step": 1538 + }, + { + "epoch": 1.0944, + "grad_norm": 1.315824270248413, + "learning_rate": 1.6022849317738283e-05, + "loss": 0.3639, + "step": 1539 + }, + { + "epoch": 1.0951111111111111, + "grad_norm": 1.2362080812454224, + "learning_rate": 1.6016240632249224e-05, + "loss": 0.3279, + "step": 1540 + }, + { + "epoch": 1.0958222222222223, + "grad_norm": 1.497611403465271, + "learning_rate": 1.6009627826045802e-05, + "loss": 0.4048, + "step": 1541 + }, + { + "epoch": 1.0965333333333334, + "grad_norm": 1.0705037117004395, + "learning_rate": 1.6003010903657337e-05, + "loss": 0.384, + "step": 1542 + }, + { + "epoch": 1.0972444444444445, + "grad_norm": 1.1323872804641724, + "learning_rate": 1.599638986961597e-05, + "loss": 0.4, + "step": 1543 + }, + { + "epoch": 1.0979555555555556, + "grad_norm": 1.102585792541504, + "learning_rate": 1.5989764728456655e-05, + "loss": 0.4203, + "step": 1544 + }, + { + "epoch": 1.0986666666666667, + "grad_norm": 1.2941927909851074, + "learning_rate": 1.5983135484717164e-05, + "loss": 0.3716, + "step": 1545 + }, + { + "epoch": 1.0993777777777778, + "grad_norm": 1.0409690141677856, + "learning_rate": 1.5976502142938074e-05, + "loss": 0.3764, + "step": 1546 + }, + { + "epoch": 1.100088888888889, + "grad_norm": 1.2093676328659058, + "learning_rate": 1.596986470766278e-05, + "loss": 0.3549, + "step": 1547 + }, + { + "epoch": 1.1008, + "grad_norm": 1.1049468517303467, + "learning_rate": 1.596322318343746e-05, + "loss": 0.3842, + "step": 1548 + }, + { + "epoch": 1.101511111111111, + "grad_norm": 1.1940687894821167, + "learning_rate": 1.5956577574811113e-05, + "loss": 0.3322, + "step": 1549 + }, + { + "epoch": 1.1022222222222222, + "grad_norm": 0.9617937803268433, + "learning_rate": 1.5949927886335522e-05, + "loss": 0.3616, + "step": 1550 + }, + { + "epoch": 1.1029333333333333, + "grad_norm": 0.9707024693489075, + "learning_rate": 1.594327412256528e-05, + "loss": 0.396, + "step": 1551 + }, + { + "epoch": 1.1036444444444444, + "grad_norm": 0.9055770039558411, + "learning_rate": 1.593661628805775e-05, + "loss": 0.383, + "step": 1552 + }, + { + "epoch": 1.1043555555555555, + "grad_norm": 1.0313224792480469, + "learning_rate": 1.5929954387373103e-05, + "loss": 0.3637, + "step": 1553 + }, + { + "epoch": 1.1050666666666666, + "grad_norm": 1.3132818937301636, + "learning_rate": 1.592328842507428e-05, + "loss": 0.3856, + "step": 1554 + }, + { + "epoch": 1.1057777777777777, + "grad_norm": 1.8435251712799072, + "learning_rate": 1.5916618405727012e-05, + "loss": 0.3876, + "step": 1555 + }, + { + "epoch": 1.1064888888888889, + "grad_norm": 1.6566928625106812, + "learning_rate": 1.590994433389981e-05, + "loss": 0.35, + "step": 1556 + }, + { + "epoch": 1.1072, + "grad_norm": 1.6706584692001343, + "learning_rate": 1.5903266214163954e-05, + "loss": 0.3886, + "step": 1557 + }, + { + "epoch": 1.107911111111111, + "grad_norm": 1.7861742973327637, + "learning_rate": 1.5896584051093508e-05, + "loss": 0.3536, + "step": 1558 + }, + { + "epoch": 1.1086222222222222, + "grad_norm": 1.0645160675048828, + "learning_rate": 1.5889897849265286e-05, + "loss": 0.3952, + "step": 1559 + }, + { + "epoch": 1.1093333333333333, + "grad_norm": 1.3725948333740234, + "learning_rate": 1.5883207613258887e-05, + "loss": 0.3454, + "step": 1560 + }, + { + "epoch": 1.1100444444444444, + "grad_norm": 1.179975152015686, + "learning_rate": 1.587651334765667e-05, + "loss": 0.3826, + "step": 1561 + }, + { + "epoch": 1.1107555555555555, + "grad_norm": 0.9270128607749939, + "learning_rate": 1.586981505704374e-05, + "loss": 0.3486, + "step": 1562 + }, + { + "epoch": 1.1114666666666666, + "grad_norm": 1.0714941024780273, + "learning_rate": 1.586311274600798e-05, + "loss": 0.3856, + "step": 1563 + }, + { + "epoch": 1.1121777777777777, + "grad_norm": 1.0796102285385132, + "learning_rate": 1.5856406419140006e-05, + "loss": 0.3549, + "step": 1564 + }, + { + "epoch": 1.1128888888888888, + "grad_norm": 1.2076587677001953, + "learning_rate": 1.58496960810332e-05, + "loss": 0.418, + "step": 1565 + }, + { + "epoch": 1.1136, + "grad_norm": 1.6168028116226196, + "learning_rate": 1.5842981736283686e-05, + "loss": 0.3276, + "step": 1566 + }, + { + "epoch": 1.114311111111111, + "grad_norm": 1.179087519645691, + "learning_rate": 1.5836263389490328e-05, + "loss": 0.3428, + "step": 1567 + }, + { + "epoch": 1.1150222222222221, + "grad_norm": 1.59413480758667, + "learning_rate": 1.582954104525474e-05, + "loss": 0.4129, + "step": 1568 + }, + { + "epoch": 1.1157333333333332, + "grad_norm": 1.2078983783721924, + "learning_rate": 1.5822814708181268e-05, + "loss": 0.3698, + "step": 1569 + }, + { + "epoch": 1.1164444444444444, + "grad_norm": 0.9954230189323425, + "learning_rate": 1.5816084382876994e-05, + "loss": 0.3507, + "step": 1570 + }, + { + "epoch": 1.1171555555555555, + "grad_norm": 1.4919719696044922, + "learning_rate": 1.5809350073951727e-05, + "loss": 0.3684, + "step": 1571 + }, + { + "epoch": 1.1178666666666666, + "grad_norm": 1.4621227979660034, + "learning_rate": 1.5802611786018013e-05, + "loss": 0.3611, + "step": 1572 + }, + { + "epoch": 1.118577777777778, + "grad_norm": 1.0913331508636475, + "learning_rate": 1.579586952369112e-05, + "loss": 0.3355, + "step": 1573 + }, + { + "epoch": 1.1192888888888888, + "grad_norm": 1.2063369750976562, + "learning_rate": 1.578912329158904e-05, + "loss": 0.3759, + "step": 1574 + }, + { + "epoch": 1.12, + "grad_norm": 1.3107824325561523, + "learning_rate": 1.5782373094332476e-05, + "loss": 0.3671, + "step": 1575 + }, + { + "epoch": 1.1207111111111112, + "grad_norm": 1.3110278844833374, + "learning_rate": 1.5775618936544853e-05, + "loss": 0.3751, + "step": 1576 + }, + { + "epoch": 1.1214222222222223, + "grad_norm": 1.3041657209396362, + "learning_rate": 1.5768860822852316e-05, + "loss": 0.3474, + "step": 1577 + }, + { + "epoch": 1.1221333333333334, + "grad_norm": 1.1096371412277222, + "learning_rate": 1.5762098757883703e-05, + "loss": 0.3562, + "step": 1578 + }, + { + "epoch": 1.1228444444444445, + "grad_norm": 1.180467128753662, + "learning_rate": 1.5755332746270574e-05, + "loss": 0.3841, + "step": 1579 + }, + { + "epoch": 1.1235555555555556, + "grad_norm": 1.351670503616333, + "learning_rate": 1.5748562792647177e-05, + "loss": 0.3684, + "step": 1580 + }, + { + "epoch": 1.1242666666666667, + "grad_norm": 1.0823265314102173, + "learning_rate": 1.5741788901650475e-05, + "loss": 0.3816, + "step": 1581 + }, + { + "epoch": 1.1249777777777779, + "grad_norm": 1.6262462139129639, + "learning_rate": 1.5735011077920117e-05, + "loss": 0.3922, + "step": 1582 + }, + { + "epoch": 1.125688888888889, + "grad_norm": 1.271303653717041, + "learning_rate": 1.5728229326098458e-05, + "loss": 0.4229, + "step": 1583 + }, + { + "epoch": 1.1264, + "grad_norm": 1.076433539390564, + "learning_rate": 1.5721443650830527e-05, + "loss": 0.3575, + "step": 1584 + }, + { + "epoch": 1.1271111111111112, + "grad_norm": 1.4613815546035767, + "learning_rate": 1.571465405676405e-05, + "loss": 0.3781, + "step": 1585 + }, + { + "epoch": 1.1278222222222223, + "grad_norm": 1.44933021068573, + "learning_rate": 1.5707860548549433e-05, + "loss": 0.3989, + "step": 1586 + }, + { + "epoch": 1.1285333333333334, + "grad_norm": 1.2528963088989258, + "learning_rate": 1.5701063130839777e-05, + "loss": 0.3994, + "step": 1587 + }, + { + "epoch": 1.1292444444444445, + "grad_norm": 1.4565110206604004, + "learning_rate": 1.5694261808290836e-05, + "loss": 0.3263, + "step": 1588 + }, + { + "epoch": 1.1299555555555556, + "grad_norm": 1.5201177597045898, + "learning_rate": 1.568745658556106e-05, + "loss": 0.4055, + "step": 1589 + }, + { + "epoch": 1.1306666666666667, + "grad_norm": 1.5089398622512817, + "learning_rate": 1.568064746731156e-05, + "loss": 0.423, + "step": 1590 + }, + { + "epoch": 1.1313777777777778, + "grad_norm": 1.3794519901275635, + "learning_rate": 1.5673834458206117e-05, + "loss": 0.3765, + "step": 1591 + }, + { + "epoch": 1.132088888888889, + "grad_norm": 1.3257907629013062, + "learning_rate": 1.566701756291118e-05, + "loss": 0.3678, + "step": 1592 + }, + { + "epoch": 1.1328, + "grad_norm": 1.613646149635315, + "learning_rate": 1.566019678609585e-05, + "loss": 0.4031, + "step": 1593 + }, + { + "epoch": 1.1335111111111111, + "grad_norm": 0.9790866374969482, + "learning_rate": 1.5653372132431902e-05, + "loss": 0.3698, + "step": 1594 + }, + { + "epoch": 1.1342222222222222, + "grad_norm": 1.657321810722351, + "learning_rate": 1.5646543606593752e-05, + "loss": 0.3548, + "step": 1595 + }, + { + "epoch": 1.1349333333333333, + "grad_norm": 1.1848413944244385, + "learning_rate": 1.563971121325848e-05, + "loss": 0.3851, + "step": 1596 + }, + { + "epoch": 1.1356444444444445, + "grad_norm": 1.6236119270324707, + "learning_rate": 1.563287495710581e-05, + "loss": 0.4044, + "step": 1597 + }, + { + "epoch": 1.1363555555555556, + "grad_norm": 1.529735803604126, + "learning_rate": 1.56260348428181e-05, + "loss": 0.3892, + "step": 1598 + }, + { + "epoch": 1.1370666666666667, + "grad_norm": 1.0895881652832031, + "learning_rate": 1.5619190875080377e-05, + "loss": 0.3825, + "step": 1599 + }, + { + "epoch": 1.1377777777777778, + "grad_norm": 1.0523102283477783, + "learning_rate": 1.561234305858028e-05, + "loss": 0.3607, + "step": 1600 + }, + { + "epoch": 1.1384888888888889, + "grad_norm": 0.9675933718681335, + "learning_rate": 1.56054913980081e-05, + "loss": 0.3212, + "step": 1601 + }, + { + "epoch": 1.1392, + "grad_norm": 1.2120637893676758, + "learning_rate": 1.5598635898056754e-05, + "loss": 0.3623, + "step": 1602 + }, + { + "epoch": 1.139911111111111, + "grad_norm": 0.8857448697090149, + "learning_rate": 1.55917765634218e-05, + "loss": 0.3785, + "step": 1603 + }, + { + "epoch": 1.1406222222222222, + "grad_norm": 1.6310255527496338, + "learning_rate": 1.5584913398801407e-05, + "loss": 0.3752, + "step": 1604 + }, + { + "epoch": 1.1413333333333333, + "grad_norm": 1.1172126531600952, + "learning_rate": 1.5578046408896378e-05, + "loss": 0.336, + "step": 1605 + }, + { + "epoch": 1.1420444444444444, + "grad_norm": 2.297649383544922, + "learning_rate": 1.557117559841013e-05, + "loss": 0.3514, + "step": 1606 + }, + { + "epoch": 1.1427555555555555, + "grad_norm": 1.8724740743637085, + "learning_rate": 1.55643009720487e-05, + "loss": 0.4154, + "step": 1607 + }, + { + "epoch": 1.1434666666666666, + "grad_norm": 1.1874552965164185, + "learning_rate": 1.5557422534520742e-05, + "loss": 0.3908, + "step": 1608 + }, + { + "epoch": 1.1441777777777777, + "grad_norm": 1.6146670579910278, + "learning_rate": 1.555054029053751e-05, + "loss": 0.3435, + "step": 1609 + }, + { + "epoch": 1.1448888888888888, + "grad_norm": 1.56118905544281, + "learning_rate": 1.5543654244812873e-05, + "loss": 0.3759, + "step": 1610 + }, + { + "epoch": 1.1456, + "grad_norm": 1.448218584060669, + "learning_rate": 1.553676440206331e-05, + "loss": 0.3546, + "step": 1611 + }, + { + "epoch": 1.146311111111111, + "grad_norm": 1.3575453758239746, + "learning_rate": 1.5529870767007883e-05, + "loss": 0.3618, + "step": 1612 + }, + { + "epoch": 1.1470222222222222, + "grad_norm": 1.1245771646499634, + "learning_rate": 1.5522973344368265e-05, + "loss": 0.3555, + "step": 1613 + }, + { + "epoch": 1.1477333333333333, + "grad_norm": 1.3774206638336182, + "learning_rate": 1.5516072138868726e-05, + "loss": 0.3493, + "step": 1614 + }, + { + "epoch": 1.1484444444444444, + "grad_norm": 1.0657647848129272, + "learning_rate": 1.5509167155236115e-05, + "loss": 0.3591, + "step": 1615 + }, + { + "epoch": 1.1491555555555555, + "grad_norm": 1.276619553565979, + "learning_rate": 1.5502258398199873e-05, + "loss": 0.4003, + "step": 1616 + }, + { + "epoch": 1.1498666666666666, + "grad_norm": 1.4831396341323853, + "learning_rate": 1.549534587249203e-05, + "loss": 0.3763, + "step": 1617 + }, + { + "epoch": 1.1505777777777777, + "grad_norm": 1.2963954210281372, + "learning_rate": 1.5488429582847194e-05, + "loss": 0.3757, + "step": 1618 + }, + { + "epoch": 1.1512888888888888, + "grad_norm": 1.1077241897583008, + "learning_rate": 1.5481509534002555e-05, + "loss": 0.3641, + "step": 1619 + }, + { + "epoch": 1.152, + "grad_norm": 1.0784932374954224, + "learning_rate": 1.5474585730697866e-05, + "loss": 0.3567, + "step": 1620 + }, + { + "epoch": 1.152711111111111, + "grad_norm": 1.4119181632995605, + "learning_rate": 1.546765817767547e-05, + "loss": 0.3948, + "step": 1621 + }, + { + "epoch": 1.1534222222222221, + "grad_norm": 1.0807956457138062, + "learning_rate": 1.546072687968026e-05, + "loss": 0.3609, + "step": 1622 + }, + { + "epoch": 1.1541333333333332, + "grad_norm": 1.4896992444992065, + "learning_rate": 1.5453791841459707e-05, + "loss": 0.3568, + "step": 1623 + }, + { + "epoch": 1.1548444444444446, + "grad_norm": 1.1284968852996826, + "learning_rate": 1.5446853067763836e-05, + "loss": 0.3586, + "step": 1624 + }, + { + "epoch": 1.1555555555555554, + "grad_norm": 1.2527087926864624, + "learning_rate": 1.5439910563345237e-05, + "loss": 0.3623, + "step": 1625 + }, + { + "epoch": 1.1562666666666668, + "grad_norm": 1.1972874402999878, + "learning_rate": 1.5432964332959047e-05, + "loss": 0.3894, + "step": 1626 + }, + { + "epoch": 1.1569777777777777, + "grad_norm": 1.5017765760421753, + "learning_rate": 1.5426014381362967e-05, + "loss": 0.3539, + "step": 1627 + }, + { + "epoch": 1.157688888888889, + "grad_norm": 0.9492869973182678, + "learning_rate": 1.5419060713317233e-05, + "loss": 0.405, + "step": 1628 + }, + { + "epoch": 1.1584, + "grad_norm": 1.5174669027328491, + "learning_rate": 1.5412103333584637e-05, + "loss": 0.3624, + "step": 1629 + }, + { + "epoch": 1.1591111111111112, + "grad_norm": 1.0004023313522339, + "learning_rate": 1.5405142246930505e-05, + "loss": 0.345, + "step": 1630 + }, + { + "epoch": 1.1598222222222223, + "grad_norm": 1.5260214805603027, + "learning_rate": 1.539817745812271e-05, + "loss": 0.356, + "step": 1631 + }, + { + "epoch": 1.1605333333333334, + "grad_norm": 1.1079480648040771, + "learning_rate": 1.539120897193166e-05, + "loss": 0.3684, + "step": 1632 + }, + { + "epoch": 1.1612444444444445, + "grad_norm": 1.0019522905349731, + "learning_rate": 1.5384236793130287e-05, + "loss": 0.387, + "step": 1633 + }, + { + "epoch": 1.1619555555555556, + "grad_norm": 1.1614753007888794, + "learning_rate": 1.537726092649406e-05, + "loss": 0.3749, + "step": 1634 + }, + { + "epoch": 1.1626666666666667, + "grad_norm": 1.418670892715454, + "learning_rate": 1.537028137680097e-05, + "loss": 0.4066, + "step": 1635 + }, + { + "epoch": 1.1633777777777778, + "grad_norm": 1.1970601081848145, + "learning_rate": 1.536329814883154e-05, + "loss": 0.39, + "step": 1636 + }, + { + "epoch": 1.164088888888889, + "grad_norm": 1.850089192390442, + "learning_rate": 1.5356311247368794e-05, + "loss": 0.3642, + "step": 1637 + }, + { + "epoch": 1.1648, + "grad_norm": 1.219102144241333, + "learning_rate": 1.5349320677198288e-05, + "loss": 0.3663, + "step": 1638 + }, + { + "epoch": 1.1655111111111112, + "grad_norm": 0.9276999235153198, + "learning_rate": 1.5342326443108086e-05, + "loss": 0.3625, + "step": 1639 + }, + { + "epoch": 1.1662222222222223, + "grad_norm": 1.6728863716125488, + "learning_rate": 1.533532854988876e-05, + "loss": 0.3754, + "step": 1640 + }, + { + "epoch": 1.1669333333333334, + "grad_norm": 1.1448969841003418, + "learning_rate": 1.532832700233339e-05, + "loss": 0.3632, + "step": 1641 + }, + { + "epoch": 1.1676444444444445, + "grad_norm": 1.656470537185669, + "learning_rate": 1.532132180523756e-05, + "loss": 0.3549, + "step": 1642 + }, + { + "epoch": 1.1683555555555556, + "grad_norm": 1.289372205734253, + "learning_rate": 1.5314312963399346e-05, + "loss": 0.3752, + "step": 1643 + }, + { + "epoch": 1.1690666666666667, + "grad_norm": 1.3323861360549927, + "learning_rate": 1.5307300481619334e-05, + "loss": 0.3508, + "step": 1644 + }, + { + "epoch": 1.1697777777777778, + "grad_norm": 1.1845046281814575, + "learning_rate": 1.530028436470059e-05, + "loss": 0.4274, + "step": 1645 + }, + { + "epoch": 1.170488888888889, + "grad_norm": 1.401002287864685, + "learning_rate": 1.529326461744868e-05, + "loss": 0.3383, + "step": 1646 + }, + { + "epoch": 1.1712, + "grad_norm": 1.5275874137878418, + "learning_rate": 1.528624124467165e-05, + "loss": 0.3788, + "step": 1647 + }, + { + "epoch": 1.1719111111111111, + "grad_norm": 1.3050048351287842, + "learning_rate": 1.5279214251180034e-05, + "loss": 0.3446, + "step": 1648 + }, + { + "epoch": 1.1726222222222222, + "grad_norm": 1.4267185926437378, + "learning_rate": 1.5272183641786835e-05, + "loss": 0.323, + "step": 1649 + }, + { + "epoch": 1.1733333333333333, + "grad_norm": 1.2118345499038696, + "learning_rate": 1.5265149421307545e-05, + "loss": 0.358, + "step": 1650 + }, + { + "epoch": 1.1740444444444444, + "grad_norm": 1.8252267837524414, + "learning_rate": 1.5258111594560126e-05, + "loss": 0.3568, + "step": 1651 + }, + { + "epoch": 1.1747555555555556, + "grad_norm": 1.2652820348739624, + "learning_rate": 1.525107016636501e-05, + "loss": 0.3515, + "step": 1652 + }, + { + "epoch": 1.1754666666666667, + "grad_norm": 1.8256908655166626, + "learning_rate": 1.5244025141545096e-05, + "loss": 0.3803, + "step": 1653 + }, + { + "epoch": 1.1761777777777778, + "grad_norm": 1.9689395427703857, + "learning_rate": 1.5236976524925743e-05, + "loss": 0.3669, + "step": 1654 + }, + { + "epoch": 1.1768888888888889, + "grad_norm": 0.902789831161499, + "learning_rate": 1.5229924321334768e-05, + "loss": 0.2946, + "step": 1655 + }, + { + "epoch": 1.1776, + "grad_norm": 1.6305521726608276, + "learning_rate": 1.5222868535602455e-05, + "loss": 0.3689, + "step": 1656 + }, + { + "epoch": 1.178311111111111, + "grad_norm": 0.9687508344650269, + "learning_rate": 1.521580917256154e-05, + "loss": 0.3331, + "step": 1657 + }, + { + "epoch": 1.1790222222222222, + "grad_norm": 1.2149345874786377, + "learning_rate": 1.5208746237047196e-05, + "loss": 0.3698, + "step": 1658 + }, + { + "epoch": 1.1797333333333333, + "grad_norm": 1.08888578414917, + "learning_rate": 1.5201679733897057e-05, + "loss": 0.3644, + "step": 1659 + }, + { + "epoch": 1.1804444444444444, + "grad_norm": 1.360252857208252, + "learning_rate": 1.5194609667951192e-05, + "loss": 0.341, + "step": 1660 + }, + { + "epoch": 1.1811555555555555, + "grad_norm": 1.522853136062622, + "learning_rate": 1.5187536044052119e-05, + "loss": 0.3633, + "step": 1661 + }, + { + "epoch": 1.1818666666666666, + "grad_norm": 1.2648450136184692, + "learning_rate": 1.5180458867044783e-05, + "loss": 0.354, + "step": 1662 + }, + { + "epoch": 1.1825777777777777, + "grad_norm": 1.154701828956604, + "learning_rate": 1.5173378141776569e-05, + "loss": 0.3707, + "step": 1663 + }, + { + "epoch": 1.1832888888888888, + "grad_norm": 1.0972743034362793, + "learning_rate": 1.5166293873097293e-05, + "loss": 0.3682, + "step": 1664 + }, + { + "epoch": 1.184, + "grad_norm": 1.8519688844680786, + "learning_rate": 1.5159206065859194e-05, + "loss": 0.4123, + "step": 1665 + }, + { + "epoch": 1.184711111111111, + "grad_norm": 1.3008527755737305, + "learning_rate": 1.5152114724916932e-05, + "loss": 0.4012, + "step": 1666 + }, + { + "epoch": 1.1854222222222222, + "grad_norm": 1.7172551155090332, + "learning_rate": 1.51450198551276e-05, + "loss": 0.3609, + "step": 1667 + }, + { + "epoch": 1.1861333333333333, + "grad_norm": 1.752185344696045, + "learning_rate": 1.5137921461350693e-05, + "loss": 0.3665, + "step": 1668 + }, + { + "epoch": 1.1868444444444444, + "grad_norm": 1.123744010925293, + "learning_rate": 1.5130819548448132e-05, + "loss": 0.3475, + "step": 1669 + }, + { + "epoch": 1.1875555555555555, + "grad_norm": 1.258014440536499, + "learning_rate": 1.512371412128424e-05, + "loss": 0.3767, + "step": 1670 + }, + { + "epoch": 1.1882666666666666, + "grad_norm": 1.3789246082305908, + "learning_rate": 1.5116605184725745e-05, + "loss": 0.35, + "step": 1671 + }, + { + "epoch": 1.1889777777777777, + "grad_norm": 1.198419451713562, + "learning_rate": 1.5109492743641791e-05, + "loss": 0.3838, + "step": 1672 + }, + { + "epoch": 1.1896888888888888, + "grad_norm": 1.126793622970581, + "learning_rate": 1.510237680290391e-05, + "loss": 0.3384, + "step": 1673 + }, + { + "epoch": 1.1904, + "grad_norm": 1.1961809396743774, + "learning_rate": 1.509525736738604e-05, + "loss": 0.3912, + "step": 1674 + }, + { + "epoch": 1.1911111111111112, + "grad_norm": 1.0691102743148804, + "learning_rate": 1.5088134441964507e-05, + "loss": 0.3663, + "step": 1675 + }, + { + "epoch": 1.1918222222222221, + "grad_norm": 1.2645443677902222, + "learning_rate": 1.5081008031518027e-05, + "loss": 0.3695, + "step": 1676 + }, + { + "epoch": 1.1925333333333334, + "grad_norm": 1.0092182159423828, + "learning_rate": 1.5073878140927707e-05, + "loss": 0.3731, + "step": 1677 + }, + { + "epoch": 1.1932444444444443, + "grad_norm": 1.940197467803955, + "learning_rate": 1.5066744775077036e-05, + "loss": 0.3048, + "step": 1678 + }, + { + "epoch": 1.1939555555555557, + "grad_norm": 1.039131999015808, + "learning_rate": 1.5059607938851883e-05, + "loss": 0.3761, + "step": 1679 + }, + { + "epoch": 1.1946666666666665, + "grad_norm": 1.2816296815872192, + "learning_rate": 1.505246763714049e-05, + "loss": 0.3495, + "step": 1680 + }, + { + "epoch": 1.1953777777777779, + "grad_norm": 1.3773049116134644, + "learning_rate": 1.504532387483348e-05, + "loss": 0.3462, + "step": 1681 + }, + { + "epoch": 1.196088888888889, + "grad_norm": 1.1692376136779785, + "learning_rate": 1.5038176656823844e-05, + "loss": 0.4073, + "step": 1682 + }, + { + "epoch": 1.1968, + "grad_norm": 1.2812310457229614, + "learning_rate": 1.5031025988006935e-05, + "loss": 0.398, + "step": 1683 + }, + { + "epoch": 1.1975111111111112, + "grad_norm": 0.9984177350997925, + "learning_rate": 1.5023871873280475e-05, + "loss": 0.3926, + "step": 1684 + }, + { + "epoch": 1.1982222222222223, + "grad_norm": 1.0882482528686523, + "learning_rate": 1.5016714317544546e-05, + "loss": 0.3474, + "step": 1685 + }, + { + "epoch": 1.1989333333333334, + "grad_norm": 1.0194960832595825, + "learning_rate": 1.5009553325701582e-05, + "loss": 0.3813, + "step": 1686 + }, + { + "epoch": 1.1996444444444445, + "grad_norm": 1.299595832824707, + "learning_rate": 1.5002388902656377e-05, + "loss": 0.3342, + "step": 1687 + }, + { + "epoch": 1.2003555555555556, + "grad_norm": 1.2409868240356445, + "learning_rate": 1.499522105331607e-05, + "loss": 0.3615, + "step": 1688 + }, + { + "epoch": 1.2010666666666667, + "grad_norm": 1.240588903427124, + "learning_rate": 1.498804978259015e-05, + "loss": 0.3619, + "step": 1689 + }, + { + "epoch": 1.2017777777777778, + "grad_norm": 1.175532341003418, + "learning_rate": 1.4980875095390447e-05, + "loss": 0.3826, + "step": 1690 + }, + { + "epoch": 1.202488888888889, + "grad_norm": 1.0013070106506348, + "learning_rate": 1.4973696996631135e-05, + "loss": 0.3926, + "step": 1691 + }, + { + "epoch": 1.2032, + "grad_norm": 1.0396312475204468, + "learning_rate": 1.4966515491228717e-05, + "loss": 0.3881, + "step": 1692 + }, + { + "epoch": 1.2039111111111112, + "grad_norm": 1.2598756551742554, + "learning_rate": 1.4959330584102038e-05, + "loss": 0.3682, + "step": 1693 + }, + { + "epoch": 1.2046222222222223, + "grad_norm": 1.022870421409607, + "learning_rate": 1.495214228017227e-05, + "loss": 0.3335, + "step": 1694 + }, + { + "epoch": 1.2053333333333334, + "grad_norm": 1.4439841508865356, + "learning_rate": 1.4944950584362907e-05, + "loss": 0.3583, + "step": 1695 + }, + { + "epoch": 1.2060444444444445, + "grad_norm": 1.301297903060913, + "learning_rate": 1.4937755501599774e-05, + "loss": 0.3895, + "step": 1696 + }, + { + "epoch": 1.2067555555555556, + "grad_norm": 1.3895888328552246, + "learning_rate": 1.4930557036811008e-05, + "loss": 0.3286, + "step": 1697 + }, + { + "epoch": 1.2074666666666667, + "grad_norm": 1.1497952938079834, + "learning_rate": 1.4923355194927069e-05, + "loss": 0.3215, + "step": 1698 + }, + { + "epoch": 1.2081777777777778, + "grad_norm": 1.7077027559280396, + "learning_rate": 1.4916149980880726e-05, + "loss": 0.406, + "step": 1699 + }, + { + "epoch": 1.208888888888889, + "grad_norm": 1.233298897743225, + "learning_rate": 1.490894139960706e-05, + "loss": 0.3537, + "step": 1700 + }, + { + "epoch": 1.2096, + "grad_norm": 1.5423519611358643, + "learning_rate": 1.4901729456043456e-05, + "loss": 0.3374, + "step": 1701 + }, + { + "epoch": 1.2103111111111111, + "grad_norm": 1.669114351272583, + "learning_rate": 1.4894514155129605e-05, + "loss": 0.3466, + "step": 1702 + }, + { + "epoch": 1.2110222222222222, + "grad_norm": 1.5848363637924194, + "learning_rate": 1.4887295501807492e-05, + "loss": 0.3321, + "step": 1703 + }, + { + "epoch": 1.2117333333333333, + "grad_norm": 1.5823243856430054, + "learning_rate": 1.4880073501021407e-05, + "loss": 0.3836, + "step": 1704 + }, + { + "epoch": 1.2124444444444444, + "grad_norm": 1.3323930501937866, + "learning_rate": 1.4872848157717925e-05, + "loss": 0.3065, + "step": 1705 + }, + { + "epoch": 1.2131555555555555, + "grad_norm": 1.1247029304504395, + "learning_rate": 1.4865619476845914e-05, + "loss": 0.3318, + "step": 1706 + }, + { + "epoch": 1.2138666666666666, + "grad_norm": 1.0606509447097778, + "learning_rate": 1.4858387463356527e-05, + "loss": 0.3315, + "step": 1707 + }, + { + "epoch": 1.2145777777777778, + "grad_norm": 1.3859400749206543, + "learning_rate": 1.4851152122203198e-05, + "loss": 0.3343, + "step": 1708 + }, + { + "epoch": 1.2152888888888889, + "grad_norm": 0.9979626536369324, + "learning_rate": 1.4843913458341646e-05, + "loss": 0.328, + "step": 1709 + }, + { + "epoch": 1.216, + "grad_norm": 1.1690951585769653, + "learning_rate": 1.4836671476729862e-05, + "loss": 0.3937, + "step": 1710 + }, + { + "epoch": 1.216711111111111, + "grad_norm": 1.3967643976211548, + "learning_rate": 1.4829426182328106e-05, + "loss": 0.3875, + "step": 1711 + }, + { + "epoch": 1.2174222222222222, + "grad_norm": 1.0909804105758667, + "learning_rate": 1.482217758009891e-05, + "loss": 0.3765, + "step": 1712 + }, + { + "epoch": 1.2181333333333333, + "grad_norm": 1.380018711090088, + "learning_rate": 1.4814925675007074e-05, + "loss": 0.3418, + "step": 1713 + }, + { + "epoch": 1.2188444444444444, + "grad_norm": 1.0474892854690552, + "learning_rate": 1.4807670472019657e-05, + "loss": 0.298, + "step": 1714 + }, + { + "epoch": 1.2195555555555555, + "grad_norm": 1.54561185836792, + "learning_rate": 1.4800411976105982e-05, + "loss": 0.3788, + "step": 1715 + }, + { + "epoch": 1.2202666666666666, + "grad_norm": 1.2497735023498535, + "learning_rate": 1.4793150192237615e-05, + "loss": 0.4142, + "step": 1716 + }, + { + "epoch": 1.2209777777777777, + "grad_norm": 1.933948040008545, + "learning_rate": 1.4785885125388392e-05, + "loss": 0.3232, + "step": 1717 + }, + { + "epoch": 1.2216888888888888, + "grad_norm": 1.6546741724014282, + "learning_rate": 1.4778616780534383e-05, + "loss": 0.401, + "step": 1718 + }, + { + "epoch": 1.2224, + "grad_norm": 1.078659176826477, + "learning_rate": 1.4771345162653907e-05, + "loss": 0.3662, + "step": 1719 + }, + { + "epoch": 1.223111111111111, + "grad_norm": 1.76822829246521, + "learning_rate": 1.4764070276727528e-05, + "loss": 0.3767, + "step": 1720 + }, + { + "epoch": 1.2238222222222221, + "grad_norm": 1.647196888923645, + "learning_rate": 1.4756792127738053e-05, + "loss": 0.3644, + "step": 1721 + }, + { + "epoch": 1.2245333333333333, + "grad_norm": 1.1572775840759277, + "learning_rate": 1.4749510720670506e-05, + "loss": 0.4006, + "step": 1722 + }, + { + "epoch": 1.2252444444444444, + "grad_norm": 1.1662623882293701, + "learning_rate": 1.474222606051216e-05, + "loss": 0.3353, + "step": 1723 + }, + { + "epoch": 1.2259555555555555, + "grad_norm": 1.2381447553634644, + "learning_rate": 1.473493815225251e-05, + "loss": 0.3595, + "step": 1724 + }, + { + "epoch": 1.2266666666666666, + "grad_norm": 1.0921741724014282, + "learning_rate": 1.4727647000883272e-05, + "loss": 0.348, + "step": 1725 + }, + { + "epoch": 1.227377777777778, + "grad_norm": 1.2850148677825928, + "learning_rate": 1.4720352611398394e-05, + "loss": 0.3504, + "step": 1726 + }, + { + "epoch": 1.2280888888888888, + "grad_norm": 1.227962613105774, + "learning_rate": 1.471305498879403e-05, + "loss": 0.3403, + "step": 1727 + }, + { + "epoch": 1.2288000000000001, + "grad_norm": 1.2953152656555176, + "learning_rate": 1.4705754138068557e-05, + "loss": 0.3245, + "step": 1728 + }, + { + "epoch": 1.229511111111111, + "grad_norm": 1.13469660282135, + "learning_rate": 1.469845006422256e-05, + "loss": 0.3372, + "step": 1729 + }, + { + "epoch": 1.2302222222222223, + "grad_norm": 1.041793704032898, + "learning_rate": 1.4691142772258826e-05, + "loss": 0.3807, + "step": 1730 + }, + { + "epoch": 1.2309333333333332, + "grad_norm": 1.0709428787231445, + "learning_rate": 1.4683832267182358e-05, + "loss": 0.3694, + "step": 1731 + }, + { + "epoch": 1.2316444444444445, + "grad_norm": 1.4014722108840942, + "learning_rate": 1.467651855400035e-05, + "loss": 0.3897, + "step": 1732 + }, + { + "epoch": 1.2323555555555556, + "grad_norm": 1.2737044095993042, + "learning_rate": 1.4669201637722195e-05, + "loss": 0.3615, + "step": 1733 + }, + { + "epoch": 1.2330666666666668, + "grad_norm": 1.9407553672790527, + "learning_rate": 1.466188152335948e-05, + "loss": 0.3885, + "step": 1734 + }, + { + "epoch": 1.2337777777777779, + "grad_norm": 2.0936059951782227, + "learning_rate": 1.465455821592599e-05, + "loss": 0.3878, + "step": 1735 + }, + { + "epoch": 1.234488888888889, + "grad_norm": 1.1932826042175293, + "learning_rate": 1.4647231720437687e-05, + "loss": 0.3825, + "step": 1736 + }, + { + "epoch": 1.2352, + "grad_norm": 1.7954846620559692, + "learning_rate": 1.463990204191272e-05, + "loss": 0.3619, + "step": 1737 + }, + { + "epoch": 1.2359111111111112, + "grad_norm": 1.5129833221435547, + "learning_rate": 1.4632569185371424e-05, + "loss": 0.3668, + "step": 1738 + }, + { + "epoch": 1.2366222222222223, + "grad_norm": 1.2177913188934326, + "learning_rate": 1.4625233155836299e-05, + "loss": 0.3747, + "step": 1739 + }, + { + "epoch": 1.2373333333333334, + "grad_norm": 1.253519058227539, + "learning_rate": 1.4617893958332025e-05, + "loss": 0.3252, + "step": 1740 + }, + { + "epoch": 1.2380444444444445, + "grad_norm": 1.208216667175293, + "learning_rate": 1.4610551597885456e-05, + "loss": 0.3039, + "step": 1741 + }, + { + "epoch": 1.2387555555555556, + "grad_norm": 1.629568338394165, + "learning_rate": 1.4603206079525607e-05, + "loss": 0.3272, + "step": 1742 + }, + { + "epoch": 1.2394666666666667, + "grad_norm": 1.1290948390960693, + "learning_rate": 1.4595857408283653e-05, + "loss": 0.3641, + "step": 1743 + }, + { + "epoch": 1.2401777777777778, + "grad_norm": 1.8320355415344238, + "learning_rate": 1.4588505589192936e-05, + "loss": 0.3617, + "step": 1744 + }, + { + "epoch": 1.240888888888889, + "grad_norm": 1.4470397233963013, + "learning_rate": 1.4581150627288952e-05, + "loss": 0.3086, + "step": 1745 + }, + { + "epoch": 1.2416, + "grad_norm": 1.294467568397522, + "learning_rate": 1.4573792527609343e-05, + "loss": 0.4057, + "step": 1746 + }, + { + "epoch": 1.2423111111111111, + "grad_norm": 1.4583157300949097, + "learning_rate": 1.4566431295193908e-05, + "loss": 0.3461, + "step": 1747 + }, + { + "epoch": 1.2430222222222223, + "grad_norm": 1.076961874961853, + "learning_rate": 1.455906693508459e-05, + "loss": 0.4184, + "step": 1748 + }, + { + "epoch": 1.2437333333333334, + "grad_norm": 1.0808669328689575, + "learning_rate": 1.4551699452325469e-05, + "loss": 0.3868, + "step": 1749 + }, + { + "epoch": 1.2444444444444445, + "grad_norm": 1.125744104385376, + "learning_rate": 1.4544328851962774e-05, + "loss": 0.3427, + "step": 1750 + }, + { + "epoch": 1.2451555555555556, + "grad_norm": 1.045751690864563, + "learning_rate": 1.453695513904486e-05, + "loss": 0.3731, + "step": 1751 + }, + { + "epoch": 1.2458666666666667, + "grad_norm": 1.438534140586853, + "learning_rate": 1.452957831862222e-05, + "loss": 0.3105, + "step": 1752 + }, + { + "epoch": 1.2465777777777778, + "grad_norm": 1.1718136072158813, + "learning_rate": 1.4522198395747468e-05, + "loss": 0.3629, + "step": 1753 + }, + { + "epoch": 1.247288888888889, + "grad_norm": 0.9107833504676819, + "learning_rate": 1.4514815375475351e-05, + "loss": 0.3668, + "step": 1754 + }, + { + "epoch": 1.248, + "grad_norm": 1.287714958190918, + "learning_rate": 1.4507429262862732e-05, + "loss": 0.3355, + "step": 1755 + }, + { + "epoch": 1.248711111111111, + "grad_norm": 1.1244010925292969, + "learning_rate": 1.4500040062968597e-05, + "loss": 0.3259, + "step": 1756 + }, + { + "epoch": 1.2494222222222222, + "grad_norm": 0.935107946395874, + "learning_rate": 1.449264778085404e-05, + "loss": 0.3534, + "step": 1757 + }, + { + "epoch": 1.2501333333333333, + "grad_norm": 1.1748002767562866, + "learning_rate": 1.4485252421582274e-05, + "loss": 0.3414, + "step": 1758 + }, + { + "epoch": 1.2508444444444444, + "grad_norm": 1.6436052322387695, + "learning_rate": 1.4477853990218613e-05, + "loss": 0.3328, + "step": 1759 + }, + { + "epoch": 1.2515555555555555, + "grad_norm": 1.2531428337097168, + "learning_rate": 1.4470452491830477e-05, + "loss": 0.3533, + "step": 1760 + }, + { + "epoch": 1.2522666666666666, + "grad_norm": 1.2014554738998413, + "learning_rate": 1.4463047931487391e-05, + "loss": 0.384, + "step": 1761 + }, + { + "epoch": 1.2529777777777777, + "grad_norm": 1.1578319072723389, + "learning_rate": 1.4455640314260972e-05, + "loss": 0.3409, + "step": 1762 + }, + { + "epoch": 1.2536888888888889, + "grad_norm": 0.931397020816803, + "learning_rate": 1.4448229645224931e-05, + "loss": 0.3569, + "step": 1763 + }, + { + "epoch": 1.2544, + "grad_norm": 1.0760736465454102, + "learning_rate": 1.4440815929455072e-05, + "loss": 0.4031, + "step": 1764 + }, + { + "epoch": 1.255111111111111, + "grad_norm": 0.8935790657997131, + "learning_rate": 1.4433399172029283e-05, + "loss": 0.3389, + "step": 1765 + }, + { + "epoch": 1.2558222222222222, + "grad_norm": 0.8354713916778564, + "learning_rate": 1.442597937802754e-05, + "loss": 0.3175, + "step": 1766 + }, + { + "epoch": 1.2565333333333333, + "grad_norm": 1.5856411457061768, + "learning_rate": 1.4418556552531891e-05, + "loss": 0.4174, + "step": 1767 + }, + { + "epoch": 1.2572444444444444, + "grad_norm": 1.0641839504241943, + "learning_rate": 1.4411130700626467e-05, + "loss": 0.3343, + "step": 1768 + }, + { + "epoch": 1.2579555555555555, + "grad_norm": 0.9821728467941284, + "learning_rate": 1.4403701827397472e-05, + "loss": 0.3461, + "step": 1769 + }, + { + "epoch": 1.2586666666666666, + "grad_norm": 0.9135785102844238, + "learning_rate": 1.4396269937933176e-05, + "loss": 0.3428, + "step": 1770 + }, + { + "epoch": 1.2593777777777777, + "grad_norm": 1.2681455612182617, + "learning_rate": 1.4388835037323917e-05, + "loss": 0.3785, + "step": 1771 + }, + { + "epoch": 1.2600888888888888, + "grad_norm": 0.9902653098106384, + "learning_rate": 1.4381397130662092e-05, + "loss": 0.3519, + "step": 1772 + }, + { + "epoch": 1.2608, + "grad_norm": 0.9300793409347534, + "learning_rate": 1.4373956223042162e-05, + "loss": 0.3518, + "step": 1773 + }, + { + "epoch": 1.261511111111111, + "grad_norm": 0.9744993448257446, + "learning_rate": 1.4366512319560642e-05, + "loss": 0.3375, + "step": 1774 + }, + { + "epoch": 1.2622222222222224, + "grad_norm": 1.3036099672317505, + "learning_rate": 1.4359065425316094e-05, + "loss": 0.3721, + "step": 1775 + }, + { + "epoch": 1.2629333333333332, + "grad_norm": 0.935070812702179, + "learning_rate": 1.4351615545409137e-05, + "loss": 0.3508, + "step": 1776 + }, + { + "epoch": 1.2636444444444446, + "grad_norm": 1.222903847694397, + "learning_rate": 1.4344162684942422e-05, + "loss": 0.366, + "step": 1777 + }, + { + "epoch": 1.2643555555555555, + "grad_norm": 0.9188815355300903, + "learning_rate": 1.433670684902066e-05, + "loss": 0.3493, + "step": 1778 + }, + { + "epoch": 1.2650666666666668, + "grad_norm": 1.379719853401184, + "learning_rate": 1.4329248042750588e-05, + "loss": 0.3023, + "step": 1779 + }, + { + "epoch": 1.2657777777777777, + "grad_norm": 1.1020480394363403, + "learning_rate": 1.4321786271240975e-05, + "loss": 0.3799, + "step": 1780 + }, + { + "epoch": 1.266488888888889, + "grad_norm": 0.9944208860397339, + "learning_rate": 1.431432153960263e-05, + "loss": 0.3032, + "step": 1781 + }, + { + "epoch": 1.2671999999999999, + "grad_norm": 1.1045870780944824, + "learning_rate": 1.4306853852948383e-05, + "loss": 0.3581, + "step": 1782 + }, + { + "epoch": 1.2679111111111112, + "grad_norm": 1.0580039024353027, + "learning_rate": 1.429938321639309e-05, + "loss": 0.378, + "step": 1783 + }, + { + "epoch": 1.268622222222222, + "grad_norm": 1.0314234495162964, + "learning_rate": 1.4291909635053627e-05, + "loss": 0.3471, + "step": 1784 + }, + { + "epoch": 1.2693333333333334, + "grad_norm": 1.0104880332946777, + "learning_rate": 1.4284433114048885e-05, + "loss": 0.3601, + "step": 1785 + }, + { + "epoch": 1.2700444444444443, + "grad_norm": 1.0712839365005493, + "learning_rate": 1.4276953658499775e-05, + "loss": 0.3228, + "step": 1786 + }, + { + "epoch": 1.2707555555555556, + "grad_norm": 1.1971325874328613, + "learning_rate": 1.4269471273529212e-05, + "loss": 0.3397, + "step": 1787 + }, + { + "epoch": 1.2714666666666667, + "grad_norm": 1.334942102432251, + "learning_rate": 1.4261985964262118e-05, + "loss": 0.3508, + "step": 1788 + }, + { + "epoch": 1.2721777777777779, + "grad_norm": 1.1479504108428955, + "learning_rate": 1.425449773582542e-05, + "loss": 0.337, + "step": 1789 + }, + { + "epoch": 1.272888888888889, + "grad_norm": 1.4597622156143188, + "learning_rate": 1.4247006593348042e-05, + "loss": 0.3488, + "step": 1790 + }, + { + "epoch": 1.2736, + "grad_norm": 1.0295283794403076, + "learning_rate": 1.4239512541960912e-05, + "loss": 0.3689, + "step": 1791 + }, + { + "epoch": 1.2743111111111112, + "grad_norm": 0.928257405757904, + "learning_rate": 1.4232015586796938e-05, + "loss": 0.3384, + "step": 1792 + }, + { + "epoch": 1.2750222222222223, + "grad_norm": 1.0047242641448975, + "learning_rate": 1.422451573299102e-05, + "loss": 0.3539, + "step": 1793 + }, + { + "epoch": 1.2757333333333334, + "grad_norm": 0.9337893128395081, + "learning_rate": 1.4217012985680054e-05, + "loss": 0.3423, + "step": 1794 + }, + { + "epoch": 1.2764444444444445, + "grad_norm": 1.2879778146743774, + "learning_rate": 1.4209507350002905e-05, + "loss": 0.3176, + "step": 1795 + }, + { + "epoch": 1.2771555555555556, + "grad_norm": 0.9787472486495972, + "learning_rate": 1.4201998831100424e-05, + "loss": 0.3442, + "step": 1796 + }, + { + "epoch": 1.2778666666666667, + "grad_norm": 1.5146849155426025, + "learning_rate": 1.419448743411543e-05, + "loss": 0.3867, + "step": 1797 + }, + { + "epoch": 1.2785777777777778, + "grad_norm": 1.2285405397415161, + "learning_rate": 1.4186973164192722e-05, + "loss": 0.3433, + "step": 1798 + }, + { + "epoch": 1.279288888888889, + "grad_norm": 1.1675403118133545, + "learning_rate": 1.417945602647906e-05, + "loss": 0.3712, + "step": 1799 + }, + { + "epoch": 1.28, + "grad_norm": 1.1022363901138306, + "learning_rate": 1.417193602612317e-05, + "loss": 0.3175, + "step": 1800 + }, + { + "epoch": 1.2807111111111111, + "grad_norm": 1.2638027667999268, + "learning_rate": 1.416441316827574e-05, + "loss": 0.3527, + "step": 1801 + }, + { + "epoch": 1.2814222222222222, + "grad_norm": 1.222473382949829, + "learning_rate": 1.4156887458089414e-05, + "loss": 0.3214, + "step": 1802 + }, + { + "epoch": 1.2821333333333333, + "grad_norm": 1.2296959161758423, + "learning_rate": 1.4149358900718792e-05, + "loss": 0.3411, + "step": 1803 + }, + { + "epoch": 1.2828444444444445, + "grad_norm": 1.342448115348816, + "learning_rate": 1.4141827501320422e-05, + "loss": 0.3326, + "step": 1804 + }, + { + "epoch": 1.2835555555555556, + "grad_norm": 1.5369163751602173, + "learning_rate": 1.4134293265052796e-05, + "loss": 0.3715, + "step": 1805 + }, + { + "epoch": 1.2842666666666667, + "grad_norm": 1.6523274183273315, + "learning_rate": 1.4126756197076352e-05, + "loss": 0.3667, + "step": 1806 + }, + { + "epoch": 1.2849777777777778, + "grad_norm": 0.9668262600898743, + "learning_rate": 1.411921630255347e-05, + "loss": 0.3475, + "step": 1807 + }, + { + "epoch": 1.2856888888888889, + "grad_norm": 1.5367858409881592, + "learning_rate": 1.411167358664846e-05, + "loss": 0.3709, + "step": 1808 + }, + { + "epoch": 1.2864, + "grad_norm": 0.9682096242904663, + "learning_rate": 1.410412805452757e-05, + "loss": 0.3338, + "step": 1809 + }, + { + "epoch": 1.287111111111111, + "grad_norm": 1.115648627281189, + "learning_rate": 1.4096579711358975e-05, + "loss": 0.3633, + "step": 1810 + }, + { + "epoch": 1.2878222222222222, + "grad_norm": 1.2411116361618042, + "learning_rate": 1.4089028562312771e-05, + "loss": 0.3801, + "step": 1811 + }, + { + "epoch": 1.2885333333333333, + "grad_norm": 1.0629624128341675, + "learning_rate": 1.4081474612560986e-05, + "loss": 0.3944, + "step": 1812 + }, + { + "epoch": 1.2892444444444444, + "grad_norm": 1.1398797035217285, + "learning_rate": 1.4073917867277557e-05, + "loss": 0.316, + "step": 1813 + }, + { + "epoch": 1.2899555555555555, + "grad_norm": 1.0691968202590942, + "learning_rate": 1.4066358331638344e-05, + "loss": 0.3569, + "step": 1814 + }, + { + "epoch": 1.2906666666666666, + "grad_norm": 0.8812613487243652, + "learning_rate": 1.4058796010821102e-05, + "loss": 0.3111, + "step": 1815 + }, + { + "epoch": 1.2913777777777777, + "grad_norm": 1.209945559501648, + "learning_rate": 1.4051230910005516e-05, + "loss": 0.3942, + "step": 1816 + }, + { + "epoch": 1.2920888888888888, + "grad_norm": 1.1320005655288696, + "learning_rate": 1.4043663034373156e-05, + "loss": 0.3444, + "step": 1817 + }, + { + "epoch": 1.2928, + "grad_norm": 1.831848382949829, + "learning_rate": 1.4036092389107502e-05, + "loss": 0.3584, + "step": 1818 + }, + { + "epoch": 1.293511111111111, + "grad_norm": 1.6399831771850586, + "learning_rate": 1.4028518979393936e-05, + "loss": 0.3431, + "step": 1819 + }, + { + "epoch": 1.2942222222222222, + "grad_norm": 1.0750391483306885, + "learning_rate": 1.402094281041972e-05, + "loss": 0.333, + "step": 1820 + }, + { + "epoch": 1.2949333333333333, + "grad_norm": 1.6502825021743774, + "learning_rate": 1.4013363887374017e-05, + "loss": 0.3529, + "step": 1821 + }, + { + "epoch": 1.2956444444444444, + "grad_norm": 1.026528000831604, + "learning_rate": 1.4005782215447865e-05, + "loss": 0.3384, + "step": 1822 + }, + { + "epoch": 1.2963555555555555, + "grad_norm": 1.3232946395874023, + "learning_rate": 1.39981977998342e-05, + "loss": 0.3399, + "step": 1823 + }, + { + "epoch": 1.2970666666666666, + "grad_norm": 1.0985386371612549, + "learning_rate": 1.3990610645727829e-05, + "loss": 0.3394, + "step": 1824 + }, + { + "epoch": 1.2977777777777777, + "grad_norm": 1.2091968059539795, + "learning_rate": 1.3983020758325429e-05, + "loss": 0.3653, + "step": 1825 + }, + { + "epoch": 1.298488888888889, + "grad_norm": 1.2108651399612427, + "learning_rate": 1.3975428142825562e-05, + "loss": 0.3418, + "step": 1826 + }, + { + "epoch": 1.2992, + "grad_norm": 1.003811001777649, + "learning_rate": 1.3967832804428645e-05, + "loss": 0.3431, + "step": 1827 + }, + { + "epoch": 1.2999111111111112, + "grad_norm": 1.1857658624649048, + "learning_rate": 1.396023474833697e-05, + "loss": 0.3562, + "step": 1828 + }, + { + "epoch": 1.3006222222222221, + "grad_norm": 1.1599621772766113, + "learning_rate": 1.3952633979754687e-05, + "loss": 0.373, + "step": 1829 + }, + { + "epoch": 1.3013333333333335, + "grad_norm": 1.1729315519332886, + "learning_rate": 1.3945030503887801e-05, + "loss": 0.3536, + "step": 1830 + }, + { + "epoch": 1.3020444444444443, + "grad_norm": 1.196595549583435, + "learning_rate": 1.3937424325944177e-05, + "loss": 0.3417, + "step": 1831 + }, + { + "epoch": 1.3027555555555557, + "grad_norm": 1.2489627599716187, + "learning_rate": 1.392981545113353e-05, + "loss": 0.3118, + "step": 1832 + }, + { + "epoch": 1.3034666666666666, + "grad_norm": 1.1984188556671143, + "learning_rate": 1.3922203884667417e-05, + "loss": 0.367, + "step": 1833 + }, + { + "epoch": 1.3041777777777779, + "grad_norm": 1.0921893119812012, + "learning_rate": 1.3914589631759245e-05, + "loss": 0.375, + "step": 1834 + }, + { + "epoch": 1.3048888888888888, + "grad_norm": 1.0042160749435425, + "learning_rate": 1.3906972697624254e-05, + "loss": 0.3416, + "step": 1835 + }, + { + "epoch": 1.3056, + "grad_norm": 0.9939702749252319, + "learning_rate": 1.3899353087479526e-05, + "loss": 0.3642, + "step": 1836 + }, + { + "epoch": 1.306311111111111, + "grad_norm": 1.0965861082077026, + "learning_rate": 1.3891730806543971e-05, + "loss": 0.3436, + "step": 1837 + }, + { + "epoch": 1.3070222222222223, + "grad_norm": 0.9993685483932495, + "learning_rate": 1.3884105860038335e-05, + "loss": 0.3664, + "step": 1838 + }, + { + "epoch": 1.3077333333333334, + "grad_norm": 1.2422786951065063, + "learning_rate": 1.3876478253185183e-05, + "loss": 0.3693, + "step": 1839 + }, + { + "epoch": 1.3084444444444445, + "grad_norm": 1.0209753513336182, + "learning_rate": 1.386884799120891e-05, + "loss": 0.3669, + "step": 1840 + }, + { + "epoch": 1.3091555555555556, + "grad_norm": 1.2958180904388428, + "learning_rate": 1.3861215079335723e-05, + "loss": 0.3475, + "step": 1841 + }, + { + "epoch": 1.3098666666666667, + "grad_norm": 0.9971857070922852, + "learning_rate": 1.3853579522793643e-05, + "loss": 0.3467, + "step": 1842 + }, + { + "epoch": 1.3105777777777778, + "grad_norm": 0.9054272770881653, + "learning_rate": 1.3845941326812509e-05, + "loss": 0.3776, + "step": 1843 + }, + { + "epoch": 1.311288888888889, + "grad_norm": 1.8911468982696533, + "learning_rate": 1.383830049662396e-05, + "loss": 0.3675, + "step": 1844 + }, + { + "epoch": 1.312, + "grad_norm": 1.5305122137069702, + "learning_rate": 1.3830657037461449e-05, + "loss": 0.3375, + "step": 1845 + }, + { + "epoch": 1.3127111111111112, + "grad_norm": 1.3429510593414307, + "learning_rate": 1.382301095456022e-05, + "loss": 0.3511, + "step": 1846 + }, + { + "epoch": 1.3134222222222223, + "grad_norm": 1.557291030883789, + "learning_rate": 1.3815362253157317e-05, + "loss": 0.3428, + "step": 1847 + }, + { + "epoch": 1.3141333333333334, + "grad_norm": 1.0459033250808716, + "learning_rate": 1.3807710938491581e-05, + "loss": 0.3636, + "step": 1848 + }, + { + "epoch": 1.3148444444444445, + "grad_norm": 1.4477393627166748, + "learning_rate": 1.380005701580364e-05, + "loss": 0.33, + "step": 1849 + }, + { + "epoch": 1.3155555555555556, + "grad_norm": 1.2664580345153809, + "learning_rate": 1.3792400490335911e-05, + "loss": 0.3473, + "step": 1850 + }, + { + "epoch": 1.3162666666666667, + "grad_norm": 1.548133373260498, + "learning_rate": 1.3784741367332588e-05, + "loss": 0.3334, + "step": 1851 + }, + { + "epoch": 1.3169777777777778, + "grad_norm": 1.8642499446868896, + "learning_rate": 1.3777079652039649e-05, + "loss": 0.3873, + "step": 1852 + }, + { + "epoch": 1.317688888888889, + "grad_norm": 0.9011029005050659, + "learning_rate": 1.3769415349704849e-05, + "loss": 0.3526, + "step": 1853 + }, + { + "epoch": 1.3184, + "grad_norm": 1.599564552307129, + "learning_rate": 1.3761748465577706e-05, + "loss": 0.2958, + "step": 1854 + }, + { + "epoch": 1.3191111111111111, + "grad_norm": 0.9831275343894958, + "learning_rate": 1.3754079004909517e-05, + "loss": 0.3819, + "step": 1855 + }, + { + "epoch": 1.3198222222222222, + "grad_norm": 1.0613319873809814, + "learning_rate": 1.374640697295334e-05, + "loss": 0.3177, + "step": 1856 + }, + { + "epoch": 1.3205333333333333, + "grad_norm": 1.019305944442749, + "learning_rate": 1.3738732374963995e-05, + "loss": 0.3327, + "step": 1857 + }, + { + "epoch": 1.3212444444444444, + "grad_norm": 1.2607380151748657, + "learning_rate": 1.373105521619805e-05, + "loss": 0.3472, + "step": 1858 + }, + { + "epoch": 1.3219555555555556, + "grad_norm": 1.0395318269729614, + "learning_rate": 1.3723375501913843e-05, + "loss": 0.3474, + "step": 1859 + }, + { + "epoch": 1.3226666666666667, + "grad_norm": 0.9738747477531433, + "learning_rate": 1.3715693237371448e-05, + "loss": 0.33, + "step": 1860 + }, + { + "epoch": 1.3233777777777778, + "grad_norm": 1.2247892618179321, + "learning_rate": 1.3708008427832698e-05, + "loss": 0.3522, + "step": 1861 + }, + { + "epoch": 1.3240888888888889, + "grad_norm": 1.0103676319122314, + "learning_rate": 1.3700321078561158e-05, + "loss": 0.3287, + "step": 1862 + }, + { + "epoch": 1.3248, + "grad_norm": 1.0274372100830078, + "learning_rate": 1.3692631194822141e-05, + "loss": 0.3845, + "step": 1863 + }, + { + "epoch": 1.325511111111111, + "grad_norm": 0.9531586766242981, + "learning_rate": 1.3684938781882692e-05, + "loss": 0.3125, + "step": 1864 + }, + { + "epoch": 1.3262222222222222, + "grad_norm": 0.9188787937164307, + "learning_rate": 1.3677243845011589e-05, + "loss": 0.3654, + "step": 1865 + }, + { + "epoch": 1.3269333333333333, + "grad_norm": 0.9649814963340759, + "learning_rate": 1.3669546389479342e-05, + "loss": 0.3654, + "step": 1866 + }, + { + "epoch": 1.3276444444444444, + "grad_norm": 0.9829484224319458, + "learning_rate": 1.3661846420558178e-05, + "loss": 0.3356, + "step": 1867 + }, + { + "epoch": 1.3283555555555555, + "grad_norm": 1.5391261577606201, + "learning_rate": 1.3654143943522051e-05, + "loss": 0.3592, + "step": 1868 + }, + { + "epoch": 1.3290666666666666, + "grad_norm": 0.9508391618728638, + "learning_rate": 1.3646438963646635e-05, + "loss": 0.3355, + "step": 1869 + }, + { + "epoch": 1.3297777777777777, + "grad_norm": 1.2651641368865967, + "learning_rate": 1.363873148620931e-05, + "loss": 0.3982, + "step": 1870 + }, + { + "epoch": 1.3304888888888888, + "grad_norm": 1.1310056447982788, + "learning_rate": 1.363102151648918e-05, + "loss": 0.3451, + "step": 1871 + }, + { + "epoch": 1.3312, + "grad_norm": 1.0021058320999146, + "learning_rate": 1.3623309059767043e-05, + "loss": 0.3364, + "step": 1872 + }, + { + "epoch": 1.331911111111111, + "grad_norm": 1.187443494796753, + "learning_rate": 1.3615594121325405e-05, + "loss": 0.354, + "step": 1873 + }, + { + "epoch": 1.3326222222222222, + "grad_norm": 1.0819636583328247, + "learning_rate": 1.3607876706448477e-05, + "loss": 0.3469, + "step": 1874 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 1.311607003211975, + "learning_rate": 1.3600156820422159e-05, + "loss": 0.3495, + "step": 1875 + }, + { + "epoch": 1.3340444444444444, + "grad_norm": 1.3692563772201538, + "learning_rate": 1.3592434468534046e-05, + "loss": 0.3158, + "step": 1876 + }, + { + "epoch": 1.3347555555555555, + "grad_norm": 1.329331874847412, + "learning_rate": 1.358470965607342e-05, + "loss": 0.3689, + "step": 1877 + }, + { + "epoch": 1.3354666666666666, + "grad_norm": 1.3141770362854004, + "learning_rate": 1.3576982388331258e-05, + "loss": 0.3624, + "step": 1878 + }, + { + "epoch": 1.336177777777778, + "grad_norm": 1.0761195421218872, + "learning_rate": 1.3569252670600204e-05, + "loss": 0.3219, + "step": 1879 + }, + { + "epoch": 1.3368888888888888, + "grad_norm": 0.8989361524581909, + "learning_rate": 1.3561520508174586e-05, + "loss": 0.3261, + "step": 1880 + }, + { + "epoch": 1.3376000000000001, + "grad_norm": 1.4322372674942017, + "learning_rate": 1.3553785906350412e-05, + "loss": 0.3586, + "step": 1881 + }, + { + "epoch": 1.338311111111111, + "grad_norm": 0.8652627468109131, + "learning_rate": 1.3546048870425356e-05, + "loss": 0.3616, + "step": 1882 + }, + { + "epoch": 1.3390222222222223, + "grad_norm": 1.5348631143569946, + "learning_rate": 1.3538309405698761e-05, + "loss": 0.3396, + "step": 1883 + }, + { + "epoch": 1.3397333333333332, + "grad_norm": 0.9790137410163879, + "learning_rate": 1.3530567517471632e-05, + "loss": 0.3517, + "step": 1884 + }, + { + "epoch": 1.3404444444444445, + "grad_norm": 1.4620872735977173, + "learning_rate": 1.3522823211046629e-05, + "loss": 0.2955, + "step": 1885 + }, + { + "epoch": 1.3411555555555554, + "grad_norm": 1.397040605545044, + "learning_rate": 1.3515076491728079e-05, + "loss": 0.3591, + "step": 1886 + }, + { + "epoch": 1.3418666666666668, + "grad_norm": 1.191253900527954, + "learning_rate": 1.3507327364821957e-05, + "loss": 0.3366, + "step": 1887 + }, + { + "epoch": 1.3425777777777776, + "grad_norm": 1.2473969459533691, + "learning_rate": 1.3499575835635884e-05, + "loss": 0.3621, + "step": 1888 + }, + { + "epoch": 1.343288888888889, + "grad_norm": 0.8970474004745483, + "learning_rate": 1.3491821909479126e-05, + "loss": 0.3819, + "step": 1889 + }, + { + "epoch": 1.3439999999999999, + "grad_norm": 1.320258378982544, + "learning_rate": 1.3484065591662596e-05, + "loss": 0.3457, + "step": 1890 + }, + { + "epoch": 1.3447111111111112, + "grad_norm": 1.0356687307357788, + "learning_rate": 1.3476306887498842e-05, + "loss": 0.338, + "step": 1891 + }, + { + "epoch": 1.3454222222222223, + "grad_norm": 1.0101186037063599, + "learning_rate": 1.3468545802302048e-05, + "loss": 0.3507, + "step": 1892 + }, + { + "epoch": 1.3461333333333334, + "grad_norm": 1.0057048797607422, + "learning_rate": 1.3460782341388028e-05, + "loss": 0.4175, + "step": 1893 + }, + { + "epoch": 1.3468444444444445, + "grad_norm": 0.954946756362915, + "learning_rate": 1.3453016510074222e-05, + "loss": 0.3557, + "step": 1894 + }, + { + "epoch": 1.3475555555555556, + "grad_norm": 0.9988337755203247, + "learning_rate": 1.3445248313679695e-05, + "loss": 0.3703, + "step": 1895 + }, + { + "epoch": 1.3482666666666667, + "grad_norm": 0.8933445811271667, + "learning_rate": 1.3437477757525131e-05, + "loss": 0.3448, + "step": 1896 + }, + { + "epoch": 1.3489777777777778, + "grad_norm": 1.1586555242538452, + "learning_rate": 1.3429704846932832e-05, + "loss": 0.3089, + "step": 1897 + }, + { + "epoch": 1.349688888888889, + "grad_norm": 1.100070595741272, + "learning_rate": 1.342192958722671e-05, + "loss": 0.3769, + "step": 1898 + }, + { + "epoch": 1.3504, + "grad_norm": 0.9014139175415039, + "learning_rate": 1.3414151983732286e-05, + "loss": 0.3084, + "step": 1899 + }, + { + "epoch": 1.3511111111111112, + "grad_norm": 1.0079975128173828, + "learning_rate": 1.3406372041776694e-05, + "loss": 0.33, + "step": 1900 + }, + { + "epoch": 1.3518222222222223, + "grad_norm": 0.8590186238288879, + "learning_rate": 1.3398589766688656e-05, + "loss": 0.3017, + "step": 1901 + }, + { + "epoch": 1.3525333333333334, + "grad_norm": 0.9999695420265198, + "learning_rate": 1.3390805163798506e-05, + "loss": 0.3485, + "step": 1902 + }, + { + "epoch": 1.3532444444444445, + "grad_norm": 1.0796483755111694, + "learning_rate": 1.338301823843816e-05, + "loss": 0.3667, + "step": 1903 + }, + { + "epoch": 1.3539555555555556, + "grad_norm": 0.966543436050415, + "learning_rate": 1.3375228995941135e-05, + "loss": 0.3287, + "step": 1904 + }, + { + "epoch": 1.3546666666666667, + "grad_norm": 1.597611665725708, + "learning_rate": 1.3367437441642528e-05, + "loss": 0.3915, + "step": 1905 + }, + { + "epoch": 1.3553777777777778, + "grad_norm": 1.2659244537353516, + "learning_rate": 1.3359643580879023e-05, + "loss": 0.3508, + "step": 1906 + }, + { + "epoch": 1.356088888888889, + "grad_norm": 1.235440731048584, + "learning_rate": 1.3351847418988881e-05, + "loss": 0.3519, + "step": 1907 + }, + { + "epoch": 1.3568, + "grad_norm": 1.6578164100646973, + "learning_rate": 1.3344048961311947e-05, + "loss": 0.3526, + "step": 1908 + }, + { + "epoch": 1.3575111111111111, + "grad_norm": 1.2138242721557617, + "learning_rate": 1.3336248213189627e-05, + "loss": 0.3071, + "step": 1909 + }, + { + "epoch": 1.3582222222222222, + "grad_norm": 0.9337212443351746, + "learning_rate": 1.3328445179964902e-05, + "loss": 0.3437, + "step": 1910 + }, + { + "epoch": 1.3589333333333333, + "grad_norm": 1.0759578943252563, + "learning_rate": 1.3320639866982317e-05, + "loss": 0.3416, + "step": 1911 + }, + { + "epoch": 1.3596444444444444, + "grad_norm": 0.9825204014778137, + "learning_rate": 1.3312832279587981e-05, + "loss": 0.3352, + "step": 1912 + }, + { + "epoch": 1.3603555555555555, + "grad_norm": 1.0256832838058472, + "learning_rate": 1.3305022423129556e-05, + "loss": 0.3721, + "step": 1913 + }, + { + "epoch": 1.3610666666666666, + "grad_norm": 0.9726967811584473, + "learning_rate": 1.3297210302956263e-05, + "loss": 0.3183, + "step": 1914 + }, + { + "epoch": 1.3617777777777778, + "grad_norm": 1.1379772424697876, + "learning_rate": 1.3289395924418868e-05, + "loss": 0.321, + "step": 1915 + }, + { + "epoch": 1.3624888888888889, + "grad_norm": 1.0053682327270508, + "learning_rate": 1.3281579292869693e-05, + "loss": 0.3269, + "step": 1916 + }, + { + "epoch": 1.3632, + "grad_norm": 1.4353225231170654, + "learning_rate": 1.3273760413662596e-05, + "loss": 0.311, + "step": 1917 + }, + { + "epoch": 1.363911111111111, + "grad_norm": 1.4805223941802979, + "learning_rate": 1.3265939292152971e-05, + "loss": 0.3426, + "step": 1918 + }, + { + "epoch": 1.3646222222222222, + "grad_norm": 1.3222260475158691, + "learning_rate": 1.3258115933697757e-05, + "loss": 0.3233, + "step": 1919 + }, + { + "epoch": 1.3653333333333333, + "grad_norm": 1.7215754985809326, + "learning_rate": 1.3250290343655419e-05, + "loss": 0.3103, + "step": 1920 + }, + { + "epoch": 1.3660444444444444, + "grad_norm": 1.3102890253067017, + "learning_rate": 1.324246252738595e-05, + "loss": 0.3823, + "step": 1921 + }, + { + "epoch": 1.3667555555555555, + "grad_norm": 1.0736052989959717, + "learning_rate": 1.3234632490250875e-05, + "loss": 0.3721, + "step": 1922 + }, + { + "epoch": 1.3674666666666666, + "grad_norm": 1.100261926651001, + "learning_rate": 1.3226800237613233e-05, + "loss": 0.3037, + "step": 1923 + }, + { + "epoch": 1.3681777777777777, + "grad_norm": 1.0285751819610596, + "learning_rate": 1.3218965774837584e-05, + "loss": 0.3382, + "step": 1924 + }, + { + "epoch": 1.3688888888888888, + "grad_norm": 1.033090353012085, + "learning_rate": 1.3211129107289996e-05, + "loss": 0.3287, + "step": 1925 + }, + { + "epoch": 1.3696, + "grad_norm": 1.3884388208389282, + "learning_rate": 1.3203290240338056e-05, + "loss": 0.3843, + "step": 1926 + }, + { + "epoch": 1.370311111111111, + "grad_norm": 0.9861066341400146, + "learning_rate": 1.3195449179350853e-05, + "loss": 0.3275, + "step": 1927 + }, + { + "epoch": 1.3710222222222221, + "grad_norm": 1.1580296754837036, + "learning_rate": 1.3187605929698975e-05, + "loss": 0.3413, + "step": 1928 + }, + { + "epoch": 1.3717333333333332, + "grad_norm": 1.1952943801879883, + "learning_rate": 1.3179760496754516e-05, + "loss": 0.4256, + "step": 1929 + }, + { + "epoch": 1.3724444444444446, + "grad_norm": 1.1256976127624512, + "learning_rate": 1.3171912885891063e-05, + "loss": 0.4169, + "step": 1930 + }, + { + "epoch": 1.3731555555555555, + "grad_norm": 1.2216428518295288, + "learning_rate": 1.316406310248369e-05, + "loss": 0.3209, + "step": 1931 + }, + { + "epoch": 1.3738666666666668, + "grad_norm": 1.4927558898925781, + "learning_rate": 1.3156211151908967e-05, + "loss": 0.3412, + "step": 1932 + }, + { + "epoch": 1.3745777777777777, + "grad_norm": 0.9821125864982605, + "learning_rate": 1.3148357039544943e-05, + "loss": 0.3684, + "step": 1933 + }, + { + "epoch": 1.375288888888889, + "grad_norm": 1.5303454399108887, + "learning_rate": 1.314050077077115e-05, + "loss": 0.3551, + "step": 1934 + }, + { + "epoch": 1.376, + "grad_norm": 1.5260412693023682, + "learning_rate": 1.3132642350968597e-05, + "loss": 0.3122, + "step": 1935 + }, + { + "epoch": 1.3767111111111112, + "grad_norm": 1.2463611364364624, + "learning_rate": 1.312478178551976e-05, + "loss": 0.3327, + "step": 1936 + }, + { + "epoch": 1.377422222222222, + "grad_norm": 2.0501275062561035, + "learning_rate": 1.3116919079808596e-05, + "loss": 0.3514, + "step": 1937 + }, + { + "epoch": 1.3781333333333334, + "grad_norm": 0.9716201424598694, + "learning_rate": 1.310905423922052e-05, + "loss": 0.3225, + "step": 1938 + }, + { + "epoch": 1.3788444444444443, + "grad_norm": 1.2475647926330566, + "learning_rate": 1.310118726914241e-05, + "loss": 0.3521, + "step": 1939 + }, + { + "epoch": 1.3795555555555556, + "grad_norm": 1.7950023412704468, + "learning_rate": 1.3093318174962609e-05, + "loss": 0.3684, + "step": 1940 + }, + { + "epoch": 1.3802666666666665, + "grad_norm": 1.155335545539856, + "learning_rate": 1.3085446962070902e-05, + "loss": 0.3161, + "step": 1941 + }, + { + "epoch": 1.3809777777777779, + "grad_norm": 1.974995732307434, + "learning_rate": 1.3077573635858536e-05, + "loss": 0.3323, + "step": 1942 + }, + { + "epoch": 1.381688888888889, + "grad_norm": 1.7780001163482666, + "learning_rate": 1.3069698201718202e-05, + "loss": 0.3475, + "step": 1943 + }, + { + "epoch": 1.3824, + "grad_norm": 1.2933237552642822, + "learning_rate": 1.3061820665044036e-05, + "loss": 0.2899, + "step": 1944 + }, + { + "epoch": 1.3831111111111112, + "grad_norm": 1.3762352466583252, + "learning_rate": 1.3053941031231613e-05, + "loss": 0.3304, + "step": 1945 + }, + { + "epoch": 1.3838222222222223, + "grad_norm": 0.9893377423286438, + "learning_rate": 1.3046059305677944e-05, + "loss": 0.3773, + "step": 1946 + }, + { + "epoch": 1.3845333333333334, + "grad_norm": 1.3018232583999634, + "learning_rate": 1.3038175493781473e-05, + "loss": 0.3242, + "step": 1947 + }, + { + "epoch": 1.3852444444444445, + "grad_norm": 0.8674962520599365, + "learning_rate": 1.3030289600942074e-05, + "loss": 0.3215, + "step": 1948 + }, + { + "epoch": 1.3859555555555556, + "grad_norm": 1.2026973962783813, + "learning_rate": 1.3022401632561046e-05, + "loss": 0.3386, + "step": 1949 + }, + { + "epoch": 1.3866666666666667, + "grad_norm": 0.9031578898429871, + "learning_rate": 1.301451159404111e-05, + "loss": 0.3297, + "step": 1950 + }, + { + "epoch": 1.3873777777777778, + "grad_norm": 1.0883301496505737, + "learning_rate": 1.3006619490786401e-05, + "loss": 0.3349, + "step": 1951 + }, + { + "epoch": 1.388088888888889, + "grad_norm": 1.1375614404678345, + "learning_rate": 1.2998725328202473e-05, + "loss": 0.3152, + "step": 1952 + }, + { + "epoch": 1.3888, + "grad_norm": 1.4187626838684082, + "learning_rate": 1.2990829111696289e-05, + "loss": 0.3524, + "step": 1953 + }, + { + "epoch": 1.3895111111111111, + "grad_norm": 0.9588616490364075, + "learning_rate": 1.2982930846676215e-05, + "loss": 0.3563, + "step": 1954 + }, + { + "epoch": 1.3902222222222222, + "grad_norm": 1.1677207946777344, + "learning_rate": 1.297503053855203e-05, + "loss": 0.3386, + "step": 1955 + }, + { + "epoch": 1.3909333333333334, + "grad_norm": 0.9868923425674438, + "learning_rate": 1.2967128192734903e-05, + "loss": 0.3414, + "step": 1956 + }, + { + "epoch": 1.3916444444444445, + "grad_norm": 1.069117546081543, + "learning_rate": 1.2959223814637402e-05, + "loss": 0.2953, + "step": 1957 + }, + { + "epoch": 1.3923555555555556, + "grad_norm": 1.4793705940246582, + "learning_rate": 1.2951317409673484e-05, + "loss": 0.3579, + "step": 1958 + }, + { + "epoch": 1.3930666666666667, + "grad_norm": 1.0319242477416992, + "learning_rate": 1.2943408983258502e-05, + "loss": 0.3939, + "step": 1959 + }, + { + "epoch": 1.3937777777777778, + "grad_norm": 1.1300039291381836, + "learning_rate": 1.2935498540809186e-05, + "loss": 0.3029, + "step": 1960 + }, + { + "epoch": 1.3944888888888889, + "grad_norm": 1.0453277826309204, + "learning_rate": 1.2927586087743651e-05, + "loss": 0.3717, + "step": 1961 + }, + { + "epoch": 1.3952, + "grad_norm": 0.8745691180229187, + "learning_rate": 1.2919671629481383e-05, + "loss": 0.2831, + "step": 1962 + }, + { + "epoch": 1.395911111111111, + "grad_norm": 0.9575726985931396, + "learning_rate": 1.2911755171443247e-05, + "loss": 0.3376, + "step": 1963 + }, + { + "epoch": 1.3966222222222222, + "grad_norm": 1.0643141269683838, + "learning_rate": 1.2903836719051482e-05, + "loss": 0.3405, + "step": 1964 + }, + { + "epoch": 1.3973333333333333, + "grad_norm": 0.874663233757019, + "learning_rate": 1.2895916277729681e-05, + "loss": 0.343, + "step": 1965 + }, + { + "epoch": 1.3980444444444444, + "grad_norm": 0.9421390295028687, + "learning_rate": 1.2887993852902811e-05, + "loss": 0.3716, + "step": 1966 + }, + { + "epoch": 1.3987555555555555, + "grad_norm": 0.9721556901931763, + "learning_rate": 1.2880069449997186e-05, + "loss": 0.3444, + "step": 1967 + }, + { + "epoch": 1.3994666666666666, + "grad_norm": 0.8497186899185181, + "learning_rate": 1.287214307444049e-05, + "loss": 0.3231, + "step": 1968 + }, + { + "epoch": 1.4001777777777777, + "grad_norm": 1.267404556274414, + "learning_rate": 1.2864214731661743e-05, + "loss": 0.3766, + "step": 1969 + }, + { + "epoch": 1.4008888888888889, + "grad_norm": 1.0990440845489502, + "learning_rate": 1.2856284427091324e-05, + "loss": 0.333, + "step": 1970 + }, + { + "epoch": 1.4016, + "grad_norm": 0.8809444308280945, + "learning_rate": 1.2848352166160946e-05, + "loss": 0.3244, + "step": 1971 + }, + { + "epoch": 1.402311111111111, + "grad_norm": 0.9222981929779053, + "learning_rate": 1.284041795430367e-05, + "loss": 0.3588, + "step": 1972 + }, + { + "epoch": 1.4030222222222222, + "grad_norm": 1.3981788158416748, + "learning_rate": 1.2832481796953887e-05, + "loss": 0.3093, + "step": 1973 + }, + { + "epoch": 1.4037333333333333, + "grad_norm": 0.9204205274581909, + "learning_rate": 1.2824543699547323e-05, + "loss": 0.364, + "step": 1974 + }, + { + "epoch": 1.4044444444444444, + "grad_norm": 1.1781657934188843, + "learning_rate": 1.2816603667521035e-05, + "loss": 0.3504, + "step": 1975 + }, + { + "epoch": 1.4051555555555555, + "grad_norm": 1.2354865074157715, + "learning_rate": 1.2808661706313402e-05, + "loss": 0.3534, + "step": 1976 + }, + { + "epoch": 1.4058666666666666, + "grad_norm": 0.9524812698364258, + "learning_rate": 1.2800717821364126e-05, + "loss": 0.3182, + "step": 1977 + }, + { + "epoch": 1.4065777777777777, + "grad_norm": 1.2486318349838257, + "learning_rate": 1.2792772018114227e-05, + "loss": 0.3404, + "step": 1978 + }, + { + "epoch": 1.4072888888888888, + "grad_norm": 0.9903961420059204, + "learning_rate": 1.2784824302006039e-05, + "loss": 0.3424, + "step": 1979 + }, + { + "epoch": 1.408, + "grad_norm": 0.9187667369842529, + "learning_rate": 1.2776874678483201e-05, + "loss": 0.3242, + "step": 1980 + }, + { + "epoch": 1.4087111111111112, + "grad_norm": 1.2320094108581543, + "learning_rate": 1.276892315299067e-05, + "loss": 0.3422, + "step": 1981 + }, + { + "epoch": 1.4094222222222221, + "grad_norm": 1.0689445734024048, + "learning_rate": 1.2760969730974692e-05, + "loss": 0.2992, + "step": 1982 + }, + { + "epoch": 1.4101333333333335, + "grad_norm": 1.8239772319793701, + "learning_rate": 1.2753014417882818e-05, + "loss": 0.3319, + "step": 1983 + }, + { + "epoch": 1.4108444444444443, + "grad_norm": 1.6998757123947144, + "learning_rate": 1.2745057219163898e-05, + "loss": 0.3329, + "step": 1984 + }, + { + "epoch": 1.4115555555555557, + "grad_norm": 0.9019652009010315, + "learning_rate": 1.2737098140268072e-05, + "loss": 0.3474, + "step": 1985 + }, + { + "epoch": 1.4122666666666666, + "grad_norm": 1.5732698440551758, + "learning_rate": 1.272913718664676e-05, + "loss": 0.3487, + "step": 1986 + }, + { + "epoch": 1.4129777777777779, + "grad_norm": 1.3415766954421997, + "learning_rate": 1.2721174363752678e-05, + "loss": 0.2878, + "step": 1987 + }, + { + "epoch": 1.4136888888888888, + "grad_norm": 1.0732128620147705, + "learning_rate": 1.2713209677039813e-05, + "loss": 0.3319, + "step": 1988 + }, + { + "epoch": 1.4144, + "grad_norm": 1.2274067401885986, + "learning_rate": 1.2705243131963431e-05, + "loss": 0.3296, + "step": 1989 + }, + { + "epoch": 1.415111111111111, + "grad_norm": 1.1528514623641968, + "learning_rate": 1.2697274733980077e-05, + "loss": 0.3479, + "step": 1990 + }, + { + "epoch": 1.4158222222222223, + "grad_norm": 0.994275689125061, + "learning_rate": 1.2689304488547553e-05, + "loss": 0.3294, + "step": 1991 + }, + { + "epoch": 1.4165333333333332, + "grad_norm": 1.4808032512664795, + "learning_rate": 1.2681332401124943e-05, + "loss": 0.316, + "step": 1992 + }, + { + "epoch": 1.4172444444444445, + "grad_norm": 1.3210557699203491, + "learning_rate": 1.2673358477172575e-05, + "loss": 0.3801, + "step": 1993 + }, + { + "epoch": 1.4179555555555556, + "grad_norm": 1.3047425746917725, + "learning_rate": 1.2665382722152042e-05, + "loss": 0.3191, + "step": 1994 + }, + { + "epoch": 1.4186666666666667, + "grad_norm": 1.3349416255950928, + "learning_rate": 1.2657405141526196e-05, + "loss": 0.3524, + "step": 1995 + }, + { + "epoch": 1.4193777777777778, + "grad_norm": 1.1185462474822998, + "learning_rate": 1.2649425740759133e-05, + "loss": 0.3378, + "step": 1996 + }, + { + "epoch": 1.420088888888889, + "grad_norm": 1.0580371618270874, + "learning_rate": 1.2641444525316203e-05, + "loss": 0.3253, + "step": 1997 + }, + { + "epoch": 1.4208, + "grad_norm": 1.4426844120025635, + "learning_rate": 1.2633461500663989e-05, + "loss": 0.3656, + "step": 1998 + }, + { + "epoch": 1.4215111111111112, + "grad_norm": 1.3692562580108643, + "learning_rate": 1.2625476672270318e-05, + "loss": 0.3234, + "step": 1999 + }, + { + "epoch": 1.4222222222222223, + "grad_norm": 1.3330110311508179, + "learning_rate": 1.2617490045604256e-05, + "loss": 0.3487, + "step": 2000 + }, + { + "epoch": 1.4229333333333334, + "grad_norm": 1.3293020725250244, + "learning_rate": 1.2609501626136095e-05, + "loss": 0.3104, + "step": 2001 + }, + { + "epoch": 1.4236444444444445, + "grad_norm": 1.045553207397461, + "learning_rate": 1.2601511419337358e-05, + "loss": 0.309, + "step": 2002 + }, + { + "epoch": 1.4243555555555556, + "grad_norm": 1.3281766176223755, + "learning_rate": 1.2593519430680794e-05, + "loss": 0.2895, + "step": 2003 + }, + { + "epoch": 1.4250666666666667, + "grad_norm": 0.9849463105201721, + "learning_rate": 1.2585525665640364e-05, + "loss": 0.3162, + "step": 2004 + }, + { + "epoch": 1.4257777777777778, + "grad_norm": 1.022937297821045, + "learning_rate": 1.2577530129691257e-05, + "loss": 0.3139, + "step": 2005 + }, + { + "epoch": 1.426488888888889, + "grad_norm": 1.2001746892929077, + "learning_rate": 1.2569532828309866e-05, + "loss": 0.374, + "step": 2006 + }, + { + "epoch": 1.4272, + "grad_norm": 0.848394513130188, + "learning_rate": 1.2561533766973804e-05, + "loss": 0.2856, + "step": 2007 + }, + { + "epoch": 1.4279111111111111, + "grad_norm": 1.080496907234192, + "learning_rate": 1.255353295116187e-05, + "loss": 0.325, + "step": 2008 + }, + { + "epoch": 1.4286222222222222, + "grad_norm": 1.1606289148330688, + "learning_rate": 1.2545530386354088e-05, + "loss": 0.3288, + "step": 2009 + }, + { + "epoch": 1.4293333333333333, + "grad_norm": 1.1202162504196167, + "learning_rate": 1.2537526078031666e-05, + "loss": 0.3438, + "step": 2010 + }, + { + "epoch": 1.4300444444444445, + "grad_norm": 1.1058259010314941, + "learning_rate": 1.2529520031677005e-05, + "loss": 0.3544, + "step": 2011 + }, + { + "epoch": 1.4307555555555556, + "grad_norm": 0.9518548250198364, + "learning_rate": 1.2521512252773705e-05, + "loss": 0.3319, + "step": 2012 + }, + { + "epoch": 1.4314666666666667, + "grad_norm": 1.1492412090301514, + "learning_rate": 1.2513502746806547e-05, + "loss": 0.3646, + "step": 2013 + }, + { + "epoch": 1.4321777777777778, + "grad_norm": 1.057096004486084, + "learning_rate": 1.2505491519261495e-05, + "loss": 0.3419, + "step": 2014 + }, + { + "epoch": 1.4328888888888889, + "grad_norm": 1.4600645303726196, + "learning_rate": 1.2497478575625691e-05, + "loss": 0.3238, + "step": 2015 + }, + { + "epoch": 1.4336, + "grad_norm": 1.3451528549194336, + "learning_rate": 1.2489463921387461e-05, + "loss": 0.2975, + "step": 2016 + }, + { + "epoch": 1.434311111111111, + "grad_norm": 1.0557570457458496, + "learning_rate": 1.2481447562036294e-05, + "loss": 0.3595, + "step": 2017 + }, + { + "epoch": 1.4350222222222222, + "grad_norm": 1.3412071466445923, + "learning_rate": 1.2473429503062846e-05, + "loss": 0.329, + "step": 2018 + }, + { + "epoch": 1.4357333333333333, + "grad_norm": 1.003546118736267, + "learning_rate": 1.246540974995894e-05, + "loss": 0.3731, + "step": 2019 + }, + { + "epoch": 1.4364444444444444, + "grad_norm": 1.1344603300094604, + "learning_rate": 1.2457388308217565e-05, + "loss": 0.3251, + "step": 2020 + }, + { + "epoch": 1.4371555555555555, + "grad_norm": 1.13152015209198, + "learning_rate": 1.2449365183332862e-05, + "loss": 0.3641, + "step": 2021 + }, + { + "epoch": 1.4378666666666666, + "grad_norm": 0.9238525629043579, + "learning_rate": 1.2441340380800118e-05, + "loss": 0.3275, + "step": 2022 + }, + { + "epoch": 1.4385777777777777, + "grad_norm": 1.375988483428955, + "learning_rate": 1.243331390611578e-05, + "loss": 0.3239, + "step": 2023 + }, + { + "epoch": 1.4392888888888888, + "grad_norm": 1.125479817390442, + "learning_rate": 1.242528576477743e-05, + "loss": 0.3311, + "step": 2024 + }, + { + "epoch": 1.44, + "grad_norm": 1.0239261388778687, + "learning_rate": 1.2417255962283803e-05, + "loss": 0.3328, + "step": 2025 + }, + { + "epoch": 1.440711111111111, + "grad_norm": 1.2067495584487915, + "learning_rate": 1.2409224504134763e-05, + "loss": 0.3268, + "step": 2026 + }, + { + "epoch": 1.4414222222222222, + "grad_norm": 1.1152054071426392, + "learning_rate": 1.2401191395831311e-05, + "loss": 0.322, + "step": 2027 + }, + { + "epoch": 1.4421333333333333, + "grad_norm": 0.9502072334289551, + "learning_rate": 1.2393156642875579e-05, + "loss": 0.345, + "step": 2028 + }, + { + "epoch": 1.4428444444444444, + "grad_norm": 1.2309901714324951, + "learning_rate": 1.2385120250770823e-05, + "loss": 0.3427, + "step": 2029 + }, + { + "epoch": 1.4435555555555555, + "grad_norm": 1.1993813514709473, + "learning_rate": 1.2377082225021426e-05, + "loss": 0.3402, + "step": 2030 + }, + { + "epoch": 1.4442666666666666, + "grad_norm": 1.2093757390975952, + "learning_rate": 1.2369042571132886e-05, + "loss": 0.3393, + "step": 2031 + }, + { + "epoch": 1.444977777777778, + "grad_norm": 1.1410267353057861, + "learning_rate": 1.2361001294611813e-05, + "loss": 0.3566, + "step": 2032 + }, + { + "epoch": 1.4456888888888888, + "grad_norm": 1.1099385023117065, + "learning_rate": 1.2352958400965943e-05, + "loss": 0.3647, + "step": 2033 + }, + { + "epoch": 1.4464000000000001, + "grad_norm": 0.907038688659668, + "learning_rate": 1.2344913895704099e-05, + "loss": 0.3564, + "step": 2034 + }, + { + "epoch": 1.447111111111111, + "grad_norm": 1.0992623567581177, + "learning_rate": 1.2336867784336226e-05, + "loss": 0.3543, + "step": 2035 + }, + { + "epoch": 1.4478222222222223, + "grad_norm": 1.083967924118042, + "learning_rate": 1.2328820072373354e-05, + "loss": 0.3216, + "step": 2036 + }, + { + "epoch": 1.4485333333333332, + "grad_norm": 0.9987021088600159, + "learning_rate": 1.232077076532762e-05, + "loss": 0.3038, + "step": 2037 + }, + { + "epoch": 1.4492444444444446, + "grad_norm": 1.084218144416809, + "learning_rate": 1.2312719868712251e-05, + "loss": 0.3431, + "step": 2038 + }, + { + "epoch": 1.4499555555555554, + "grad_norm": 1.227560043334961, + "learning_rate": 1.2304667388041562e-05, + "loss": 0.3139, + "step": 2039 + }, + { + "epoch": 1.4506666666666668, + "grad_norm": 1.1929010152816772, + "learning_rate": 1.2296613328830952e-05, + "loss": 0.2887, + "step": 2040 + }, + { + "epoch": 1.4513777777777777, + "grad_norm": 1.3664538860321045, + "learning_rate": 1.2288557696596906e-05, + "loss": 0.3061, + "step": 2041 + }, + { + "epoch": 1.452088888888889, + "grad_norm": 1.5216444730758667, + "learning_rate": 1.2280500496856976e-05, + "loss": 0.3197, + "step": 2042 + }, + { + "epoch": 1.4527999999999999, + "grad_norm": 1.1681137084960938, + "learning_rate": 1.2272441735129797e-05, + "loss": 0.3226, + "step": 2043 + }, + { + "epoch": 1.4535111111111112, + "grad_norm": 1.147595763206482, + "learning_rate": 1.2264381416935072e-05, + "loss": 0.3555, + "step": 2044 + }, + { + "epoch": 1.4542222222222223, + "grad_norm": 1.0411474704742432, + "learning_rate": 1.2256319547793566e-05, + "loss": 0.3138, + "step": 2045 + }, + { + "epoch": 1.4549333333333334, + "grad_norm": 0.9863610863685608, + "learning_rate": 1.2248256133227113e-05, + "loss": 0.2762, + "step": 2046 + }, + { + "epoch": 1.4556444444444445, + "grad_norm": 1.0725001096725464, + "learning_rate": 1.2240191178758598e-05, + "loss": 0.3392, + "step": 2047 + }, + { + "epoch": 1.4563555555555556, + "grad_norm": 1.3548716306686401, + "learning_rate": 1.2232124689911968e-05, + "loss": 0.3205, + "step": 2048 + }, + { + "epoch": 1.4570666666666667, + "grad_norm": 0.9779976606369019, + "learning_rate": 1.2224056672212216e-05, + "loss": 0.3313, + "step": 2049 + }, + { + "epoch": 1.4577777777777778, + "grad_norm": 1.1804918050765991, + "learning_rate": 1.2215987131185385e-05, + "loss": 0.369, + "step": 2050 + }, + { + "epoch": 1.458488888888889, + "grad_norm": 1.3375389575958252, + "learning_rate": 1.2207916072358556e-05, + "loss": 0.3553, + "step": 2051 + }, + { + "epoch": 1.4592, + "grad_norm": 1.2272851467132568, + "learning_rate": 1.219984350125986e-05, + "loss": 0.3132, + "step": 2052 + }, + { + "epoch": 1.4599111111111112, + "grad_norm": 0.9497923851013184, + "learning_rate": 1.2191769423418449e-05, + "loss": 0.3117, + "step": 2053 + }, + { + "epoch": 1.4606222222222223, + "grad_norm": 1.2994396686553955, + "learning_rate": 1.2183693844364527e-05, + "loss": 0.3317, + "step": 2054 + }, + { + "epoch": 1.4613333333333334, + "grad_norm": 1.093792200088501, + "learning_rate": 1.2175616769629304e-05, + "loss": 0.3193, + "step": 2055 + }, + { + "epoch": 1.4620444444444445, + "grad_norm": 1.3622666597366333, + "learning_rate": 1.2167538204745032e-05, + "loss": 0.3763, + "step": 2056 + }, + { + "epoch": 1.4627555555555556, + "grad_norm": 1.1779392957687378, + "learning_rate": 1.215945815524497e-05, + "loss": 0.3237, + "step": 2057 + }, + { + "epoch": 1.4634666666666667, + "grad_norm": 1.0432101488113403, + "learning_rate": 1.2151376626663407e-05, + "loss": 0.3168, + "step": 2058 + }, + { + "epoch": 1.4641777777777778, + "grad_norm": 1.1285439729690552, + "learning_rate": 1.2143293624535634e-05, + "loss": 0.3212, + "step": 2059 + }, + { + "epoch": 1.464888888888889, + "grad_norm": 1.050945520401001, + "learning_rate": 1.2135209154397962e-05, + "loss": 0.3254, + "step": 2060 + }, + { + "epoch": 1.4656, + "grad_norm": 1.431994080543518, + "learning_rate": 1.2127123221787695e-05, + "loss": 0.3773, + "step": 2061 + }, + { + "epoch": 1.4663111111111111, + "grad_norm": 1.0531187057495117, + "learning_rate": 1.2119035832243148e-05, + "loss": 0.3433, + "step": 2062 + }, + { + "epoch": 1.4670222222222222, + "grad_norm": 1.333201289176941, + "learning_rate": 1.2110946991303634e-05, + "loss": 0.3324, + "step": 2063 + }, + { + "epoch": 1.4677333333333333, + "grad_norm": 1.1536962985992432, + "learning_rate": 1.210285670450945e-05, + "loss": 0.281, + "step": 2064 + }, + { + "epoch": 1.4684444444444444, + "grad_norm": 1.0811021327972412, + "learning_rate": 1.20947649774019e-05, + "loss": 0.3143, + "step": 2065 + }, + { + "epoch": 1.4691555555555555, + "grad_norm": 1.1931098699569702, + "learning_rate": 1.2086671815523257e-05, + "loss": 0.3569, + "step": 2066 + }, + { + "epoch": 1.4698666666666667, + "grad_norm": 1.067240834236145, + "learning_rate": 1.2078577224416785e-05, + "loss": 0.3254, + "step": 2067 + }, + { + "epoch": 1.4705777777777778, + "grad_norm": 1.02130126953125, + "learning_rate": 1.2070481209626731e-05, + "loss": 0.3226, + "step": 2068 + }, + { + "epoch": 1.4712888888888889, + "grad_norm": 1.330866813659668, + "learning_rate": 1.206238377669831e-05, + "loss": 0.3868, + "step": 2069 + }, + { + "epoch": 1.472, + "grad_norm": 1.1567554473876953, + "learning_rate": 1.205428493117771e-05, + "loss": 0.371, + "step": 2070 + }, + { + "epoch": 1.472711111111111, + "grad_norm": 1.3533635139465332, + "learning_rate": 1.2046184678612092e-05, + "loss": 0.3625, + "step": 2071 + }, + { + "epoch": 1.4734222222222222, + "grad_norm": 0.9670425057411194, + "learning_rate": 1.2038083024549574e-05, + "loss": 0.3568, + "step": 2072 + }, + { + "epoch": 1.4741333333333333, + "grad_norm": 1.2578643560409546, + "learning_rate": 1.2029979974539233e-05, + "loss": 0.317, + "step": 2073 + }, + { + "epoch": 1.4748444444444444, + "grad_norm": 1.0469448566436768, + "learning_rate": 1.2021875534131114e-05, + "loss": 0.3374, + "step": 2074 + }, + { + "epoch": 1.4755555555555555, + "grad_norm": 0.8537464141845703, + "learning_rate": 1.2013769708876199e-05, + "loss": 0.291, + "step": 2075 + }, + { + "epoch": 1.4762666666666666, + "grad_norm": 1.1719653606414795, + "learning_rate": 1.2005662504326426e-05, + "loss": 0.3332, + "step": 2076 + }, + { + "epoch": 1.4769777777777777, + "grad_norm": 1.1523289680480957, + "learning_rate": 1.1997553926034675e-05, + "loss": 0.2903, + "step": 2077 + }, + { + "epoch": 1.4776888888888888, + "grad_norm": 0.9634256362915039, + "learning_rate": 1.1989443979554774e-05, + "loss": 0.2965, + "step": 2078 + }, + { + "epoch": 1.4784, + "grad_norm": 1.1809056997299194, + "learning_rate": 1.1981332670441476e-05, + "loss": 0.306, + "step": 2079 + }, + { + "epoch": 1.479111111111111, + "grad_norm": 1.343466877937317, + "learning_rate": 1.1973220004250483e-05, + "loss": 0.3447, + "step": 2080 + }, + { + "epoch": 1.4798222222222222, + "grad_norm": 1.1433184146881104, + "learning_rate": 1.1965105986538412e-05, + "loss": 0.3563, + "step": 2081 + }, + { + "epoch": 1.4805333333333333, + "grad_norm": 1.2959072589874268, + "learning_rate": 1.1956990622862813e-05, + "loss": 0.3051, + "step": 2082 + }, + { + "epoch": 1.4812444444444444, + "grad_norm": 1.3650143146514893, + "learning_rate": 1.194887391878216e-05, + "loss": 0.2775, + "step": 2083 + }, + { + "epoch": 1.4819555555555555, + "grad_norm": 1.013236403465271, + "learning_rate": 1.1940755879855834e-05, + "loss": 0.3204, + "step": 2084 + }, + { + "epoch": 1.4826666666666668, + "grad_norm": 1.3554506301879883, + "learning_rate": 1.1932636511644145e-05, + "loss": 0.3361, + "step": 2085 + }, + { + "epoch": 1.4833777777777777, + "grad_norm": 1.0100345611572266, + "learning_rate": 1.19245158197083e-05, + "loss": 0.3078, + "step": 2086 + }, + { + "epoch": 1.484088888888889, + "grad_norm": 0.9266943335533142, + "learning_rate": 1.1916393809610428e-05, + "loss": 0.3105, + "step": 2087 + }, + { + "epoch": 1.4848, + "grad_norm": 1.030460238456726, + "learning_rate": 1.1908270486913538e-05, + "loss": 0.3044, + "step": 2088 + }, + { + "epoch": 1.4855111111111112, + "grad_norm": 1.5367727279663086, + "learning_rate": 1.1900145857181562e-05, + "loss": 0.3242, + "step": 2089 + }, + { + "epoch": 1.4862222222222221, + "grad_norm": 1.0293164253234863, + "learning_rate": 1.1892019925979317e-05, + "loss": 0.3065, + "step": 2090 + }, + { + "epoch": 1.4869333333333334, + "grad_norm": 1.1390002965927124, + "learning_rate": 1.1883892698872509e-05, + "loss": 0.3491, + "step": 2091 + }, + { + "epoch": 1.4876444444444443, + "grad_norm": 1.0618352890014648, + "learning_rate": 1.187576418142773e-05, + "loss": 0.3217, + "step": 2092 + }, + { + "epoch": 1.4883555555555557, + "grad_norm": 1.078045129776001, + "learning_rate": 1.1867634379212469e-05, + "loss": 0.3279, + "step": 2093 + }, + { + "epoch": 1.4890666666666665, + "grad_norm": 0.841476321220398, + "learning_rate": 1.185950329779508e-05, + "loss": 0.3133, + "step": 2094 + }, + { + "epoch": 1.4897777777777779, + "grad_norm": 0.9689232707023621, + "learning_rate": 1.18513709427448e-05, + "loss": 0.3587, + "step": 2095 + }, + { + "epoch": 1.4904888888888888, + "grad_norm": 1.126975655555725, + "learning_rate": 1.1843237319631737e-05, + "loss": 0.3199, + "step": 2096 + }, + { + "epoch": 1.4912, + "grad_norm": 0.8914799690246582, + "learning_rate": 1.183510243402687e-05, + "loss": 0.3218, + "step": 2097 + }, + { + "epoch": 1.4919111111111112, + "grad_norm": 1.2953540086746216, + "learning_rate": 1.1826966291502036e-05, + "loss": 0.3432, + "step": 2098 + }, + { + "epoch": 1.4926222222222223, + "grad_norm": 0.8045092225074768, + "learning_rate": 1.1818828897629941e-05, + "loss": 0.3032, + "step": 2099 + }, + { + "epoch": 1.4933333333333334, + "grad_norm": 1.0627284049987793, + "learning_rate": 1.1810690257984145e-05, + "loss": 0.3489, + "step": 2100 + }, + { + "epoch": 1.4940444444444445, + "grad_norm": 1.1338776350021362, + "learning_rate": 1.180255037813906e-05, + "loss": 0.3591, + "step": 2101 + }, + { + "epoch": 1.4947555555555556, + "grad_norm": 0.9438974261283875, + "learning_rate": 1.1794409263669948e-05, + "loss": 0.3327, + "step": 2102 + }, + { + "epoch": 1.4954666666666667, + "grad_norm": 0.8158422112464905, + "learning_rate": 1.1786266920152915e-05, + "loss": 0.2777, + "step": 2103 + }, + { + "epoch": 1.4961777777777778, + "grad_norm": 1.1022312641143799, + "learning_rate": 1.1778123353164917e-05, + "loss": 0.3408, + "step": 2104 + }, + { + "epoch": 1.496888888888889, + "grad_norm": 0.922261118888855, + "learning_rate": 1.1769978568283732e-05, + "loss": 0.3475, + "step": 2105 + }, + { + "epoch": 1.4976, + "grad_norm": 0.9737580418586731, + "learning_rate": 1.1761832571087994e-05, + "loss": 0.3334, + "step": 2106 + }, + { + "epoch": 1.4983111111111111, + "grad_norm": 1.0673904418945312, + "learning_rate": 1.1753685367157146e-05, + "loss": 0.3312, + "step": 2107 + }, + { + "epoch": 1.4990222222222223, + "grad_norm": 1.0072141885757446, + "learning_rate": 1.1745536962071471e-05, + "loss": 0.3267, + "step": 2108 + }, + { + "epoch": 1.4997333333333334, + "grad_norm": 1.921704649925232, + "learning_rate": 1.1737387361412067e-05, + "loss": 0.3351, + "step": 2109 + }, + { + "epoch": 1.5004444444444445, + "grad_norm": 1.440348505973816, + "learning_rate": 1.172923657076086e-05, + "loss": 0.3533, + "step": 2110 + }, + { + "epoch": 1.5011555555555556, + "grad_norm": 1.2929506301879883, + "learning_rate": 1.172108459570058e-05, + "loss": 0.3215, + "step": 2111 + }, + { + "epoch": 1.5018666666666667, + "grad_norm": 1.8461576700210571, + "learning_rate": 1.1712931441814776e-05, + "loss": 0.3354, + "step": 2112 + }, + { + "epoch": 1.5025777777777778, + "grad_norm": 1.444932222366333, + "learning_rate": 1.1704777114687804e-05, + "loss": 0.3309, + "step": 2113 + }, + { + "epoch": 1.503288888888889, + "grad_norm": 1.637275218963623, + "learning_rate": 1.169662161990482e-05, + "loss": 0.3337, + "step": 2114 + }, + { + "epoch": 1.504, + "grad_norm": 2.187089204788208, + "learning_rate": 1.168846496305178e-05, + "loss": 0.307, + "step": 2115 + }, + { + "epoch": 1.504711111111111, + "grad_norm": 1.4246854782104492, + "learning_rate": 1.168030714971544e-05, + "loss": 0.3102, + "step": 2116 + }, + { + "epoch": 1.5054222222222222, + "grad_norm": 1.6544229984283447, + "learning_rate": 1.1672148185483343e-05, + "loss": 0.3385, + "step": 2117 + }, + { + "epoch": 1.5061333333333333, + "grad_norm": 2.455613136291504, + "learning_rate": 1.1663988075943828e-05, + "loss": 0.323, + "step": 2118 + }, + { + "epoch": 1.5068444444444444, + "grad_norm": 1.5486048460006714, + "learning_rate": 1.1655826826686005e-05, + "loss": 0.3377, + "step": 2119 + }, + { + "epoch": 1.5075555555555555, + "grad_norm": 1.5563117265701294, + "learning_rate": 1.164766444329978e-05, + "loss": 0.3609, + "step": 2120 + }, + { + "epoch": 1.5082666666666666, + "grad_norm": 2.5424158573150635, + "learning_rate": 1.1639500931375827e-05, + "loss": 0.3097, + "step": 2121 + }, + { + "epoch": 1.5089777777777778, + "grad_norm": 1.6545073986053467, + "learning_rate": 1.163133629650559e-05, + "loss": 0.354, + "step": 2122 + }, + { + "epoch": 1.5096888888888889, + "grad_norm": 1.5489046573638916, + "learning_rate": 1.1623170544281292e-05, + "loss": 0.3225, + "step": 2123 + }, + { + "epoch": 1.5104, + "grad_norm": 1.9389344453811646, + "learning_rate": 1.1615003680295917e-05, + "loss": 0.3749, + "step": 2124 + }, + { + "epoch": 1.511111111111111, + "grad_norm": 1.3022468090057373, + "learning_rate": 1.1606835710143207e-05, + "loss": 0.3522, + "step": 2125 + }, + { + "epoch": 1.5118222222222222, + "grad_norm": 1.2788872718811035, + "learning_rate": 1.1598666639417664e-05, + "loss": 0.2824, + "step": 2126 + }, + { + "epoch": 1.5125333333333333, + "grad_norm": 1.960778832435608, + "learning_rate": 1.1590496473714547e-05, + "loss": 0.3159, + "step": 2127 + }, + { + "epoch": 1.5132444444444444, + "grad_norm": 1.2102631330490112, + "learning_rate": 1.158232521862986e-05, + "loss": 0.2855, + "step": 2128 + }, + { + "epoch": 1.5139555555555555, + "grad_norm": 1.5570838451385498, + "learning_rate": 1.1574152879760354e-05, + "loss": 0.3524, + "step": 2129 + }, + { + "epoch": 1.5146666666666668, + "grad_norm": 1.7978627681732178, + "learning_rate": 1.1565979462703525e-05, + "loss": 0.3494, + "step": 2130 + }, + { + "epoch": 1.5153777777777777, + "grad_norm": 1.1260087490081787, + "learning_rate": 1.1557804973057605e-05, + "loss": 0.3334, + "step": 2131 + }, + { + "epoch": 1.516088888888889, + "grad_norm": 2.082268714904785, + "learning_rate": 1.1549629416421561e-05, + "loss": 0.3137, + "step": 2132 + }, + { + "epoch": 1.5168, + "grad_norm": 2.1752126216888428, + "learning_rate": 1.1541452798395096e-05, + "loss": 0.3074, + "step": 2133 + }, + { + "epoch": 1.5175111111111113, + "grad_norm": 1.2961194515228271, + "learning_rate": 1.1533275124578631e-05, + "loss": 0.3087, + "step": 2134 + }, + { + "epoch": 1.5182222222222221, + "grad_norm": 1.761418104171753, + "learning_rate": 1.1525096400573317e-05, + "loss": 0.3348, + "step": 2135 + }, + { + "epoch": 1.5189333333333335, + "grad_norm": 2.528090000152588, + "learning_rate": 1.1516916631981021e-05, + "loss": 0.3308, + "step": 2136 + }, + { + "epoch": 1.5196444444444444, + "grad_norm": 1.5905219316482544, + "learning_rate": 1.1508735824404326e-05, + "loss": 0.31, + "step": 2137 + }, + { + "epoch": 1.5203555555555557, + "grad_norm": 2.0823566913604736, + "learning_rate": 1.1500553983446527e-05, + "loss": 0.3257, + "step": 2138 + }, + { + "epoch": 1.5210666666666666, + "grad_norm": 1.776404857635498, + "learning_rate": 1.1492371114711628e-05, + "loss": 0.3255, + "step": 2139 + }, + { + "epoch": 1.521777777777778, + "grad_norm": 0.9690275192260742, + "learning_rate": 1.1484187223804337e-05, + "loss": 0.2898, + "step": 2140 + }, + { + "epoch": 1.5224888888888888, + "grad_norm": 1.3777306079864502, + "learning_rate": 1.1476002316330057e-05, + "loss": 0.3312, + "step": 2141 + }, + { + "epoch": 1.5232, + "grad_norm": 1.8055188655853271, + "learning_rate": 1.1467816397894893e-05, + "loss": 0.3186, + "step": 2142 + }, + { + "epoch": 1.523911111111111, + "grad_norm": 0.926442563533783, + "learning_rate": 1.1459629474105642e-05, + "loss": 0.3557, + "step": 2143 + }, + { + "epoch": 1.5246222222222223, + "grad_norm": 1.8001729249954224, + "learning_rate": 1.1451441550569787e-05, + "loss": 0.3128, + "step": 2144 + }, + { + "epoch": 1.5253333333333332, + "grad_norm": 1.4792191982269287, + "learning_rate": 1.1443252632895496e-05, + "loss": 0.3171, + "step": 2145 + }, + { + "epoch": 1.5260444444444445, + "grad_norm": 0.9275685548782349, + "learning_rate": 1.1435062726691622e-05, + "loss": 0.3564, + "step": 2146 + }, + { + "epoch": 1.5267555555555554, + "grad_norm": 1.9356348514556885, + "learning_rate": 1.1426871837567687e-05, + "loss": 0.2806, + "step": 2147 + }, + { + "epoch": 1.5274666666666668, + "grad_norm": 1.3051375150680542, + "learning_rate": 1.14186799711339e-05, + "loss": 0.3018, + "step": 2148 + }, + { + "epoch": 1.5281777777777776, + "grad_norm": 1.3035756349563599, + "learning_rate": 1.141048713300112e-05, + "loss": 0.3347, + "step": 2149 + }, + { + "epoch": 1.528888888888889, + "grad_norm": 1.6808441877365112, + "learning_rate": 1.1402293328780887e-05, + "loss": 0.3095, + "step": 2150 + }, + { + "epoch": 1.5295999999999998, + "grad_norm": 1.142425537109375, + "learning_rate": 1.1394098564085399e-05, + "loss": 0.3096, + "step": 2151 + }, + { + "epoch": 1.5303111111111112, + "grad_norm": 1.447026014328003, + "learning_rate": 1.1385902844527508e-05, + "loss": 0.3516, + "step": 2152 + }, + { + "epoch": 1.531022222222222, + "grad_norm": 1.7127861976623535, + "learning_rate": 1.1377706175720725e-05, + "loss": 0.3404, + "step": 2153 + }, + { + "epoch": 1.5317333333333334, + "grad_norm": 1.179124355316162, + "learning_rate": 1.1369508563279207e-05, + "loss": 0.312, + "step": 2154 + }, + { + "epoch": 1.5324444444444445, + "grad_norm": 1.5463944673538208, + "learning_rate": 1.1361310012817763e-05, + "loss": 0.2816, + "step": 2155 + }, + { + "epoch": 1.5331555555555556, + "grad_norm": 1.673920750617981, + "learning_rate": 1.1353110529951836e-05, + "loss": 0.3624, + "step": 2156 + }, + { + "epoch": 1.5338666666666667, + "grad_norm": 1.4524831771850586, + "learning_rate": 1.1344910120297514e-05, + "loss": 0.3089, + "step": 2157 + }, + { + "epoch": 1.5345777777777778, + "grad_norm": 1.5326095819473267, + "learning_rate": 1.1336708789471522e-05, + "loss": 0.3112, + "step": 2158 + }, + { + "epoch": 1.535288888888889, + "grad_norm": 2.347959280014038, + "learning_rate": 1.1328506543091207e-05, + "loss": 0.3425, + "step": 2159 + }, + { + "epoch": 1.536, + "grad_norm": 1.537996530532837, + "learning_rate": 1.1320303386774546e-05, + "loss": 0.3081, + "step": 2160 + }, + { + "epoch": 1.5367111111111111, + "grad_norm": 1.2789113521575928, + "learning_rate": 1.1312099326140148e-05, + "loss": 0.3446, + "step": 2161 + }, + { + "epoch": 1.5374222222222222, + "grad_norm": 2.293658494949341, + "learning_rate": 1.1303894366807234e-05, + "loss": 0.3471, + "step": 2162 + }, + { + "epoch": 1.5381333333333334, + "grad_norm": 1.562273383140564, + "learning_rate": 1.1295688514395638e-05, + "loss": 0.3215, + "step": 2163 + }, + { + "epoch": 1.5388444444444445, + "grad_norm": 1.4473588466644287, + "learning_rate": 1.128748177452581e-05, + "loss": 0.3438, + "step": 2164 + }, + { + "epoch": 1.5395555555555556, + "grad_norm": 1.7279552221298218, + "learning_rate": 1.1279274152818807e-05, + "loss": 0.2837, + "step": 2165 + }, + { + "epoch": 1.5402666666666667, + "grad_norm": 1.0090383291244507, + "learning_rate": 1.127106565489629e-05, + "loss": 0.2845, + "step": 2166 + }, + { + "epoch": 1.5409777777777778, + "grad_norm": 1.041144609451294, + "learning_rate": 1.126285628638052e-05, + "loss": 0.281, + "step": 2167 + }, + { + "epoch": 1.5416888888888889, + "grad_norm": 1.6673614978790283, + "learning_rate": 1.1254646052894353e-05, + "loss": 0.2866, + "step": 2168 + }, + { + "epoch": 1.5424, + "grad_norm": 0.9958292841911316, + "learning_rate": 1.1246434960061243e-05, + "loss": 0.3224, + "step": 2169 + }, + { + "epoch": 1.543111111111111, + "grad_norm": 1.2301416397094727, + "learning_rate": 1.1238223013505227e-05, + "loss": 0.303, + "step": 2170 + }, + { + "epoch": 1.5438222222222222, + "grad_norm": 1.238680362701416, + "learning_rate": 1.1230010218850925e-05, + "loss": 0.3534, + "step": 2171 + }, + { + "epoch": 1.5445333333333333, + "grad_norm": 0.8352081775665283, + "learning_rate": 1.1221796581723543e-05, + "loss": 0.3093, + "step": 2172 + }, + { + "epoch": 1.5452444444444444, + "grad_norm": 0.9778311848640442, + "learning_rate": 1.1213582107748859e-05, + "loss": 0.2986, + "step": 2173 + }, + { + "epoch": 1.5459555555555555, + "grad_norm": 0.9337235689163208, + "learning_rate": 1.1205366802553231e-05, + "loss": 0.3512, + "step": 2174 + }, + { + "epoch": 1.5466666666666666, + "grad_norm": 0.9245274662971497, + "learning_rate": 1.1197150671763584e-05, + "loss": 0.2989, + "step": 2175 + }, + { + "epoch": 1.5473777777777777, + "grad_norm": 0.994178056716919, + "learning_rate": 1.1188933721007402e-05, + "loss": 0.3038, + "step": 2176 + }, + { + "epoch": 1.5480888888888888, + "grad_norm": 1.1163045167922974, + "learning_rate": 1.118071595591274e-05, + "loss": 0.3549, + "step": 2177 + }, + { + "epoch": 1.5488, + "grad_norm": 0.9283407330513, + "learning_rate": 1.1172497382108203e-05, + "loss": 0.3145, + "step": 2178 + }, + { + "epoch": 1.549511111111111, + "grad_norm": 1.2556705474853516, + "learning_rate": 1.116427800522296e-05, + "loss": 0.3081, + "step": 2179 + }, + { + "epoch": 1.5502222222222222, + "grad_norm": 1.1361008882522583, + "learning_rate": 1.1156057830886713e-05, + "loss": 0.3283, + "step": 2180 + }, + { + "epoch": 1.5509333333333335, + "grad_norm": 1.1516867876052856, + "learning_rate": 1.114783686472973e-05, + "loss": 0.3318, + "step": 2181 + }, + { + "epoch": 1.5516444444444444, + "grad_norm": 1.6101040840148926, + "learning_rate": 1.1139615112382804e-05, + "loss": 0.2766, + "step": 2182 + }, + { + "epoch": 1.5523555555555557, + "grad_norm": 1.0736229419708252, + "learning_rate": 1.113139257947728e-05, + "loss": 0.3292, + "step": 2183 + }, + { + "epoch": 1.5530666666666666, + "grad_norm": 1.1984262466430664, + "learning_rate": 1.112316927164503e-05, + "loss": 0.2878, + "step": 2184 + }, + { + "epoch": 1.553777777777778, + "grad_norm": 0.8324856758117676, + "learning_rate": 1.111494519451846e-05, + "loss": 0.2985, + "step": 2185 + }, + { + "epoch": 1.5544888888888888, + "grad_norm": 1.4276353120803833, + "learning_rate": 1.11067203537305e-05, + "loss": 0.3045, + "step": 2186 + }, + { + "epoch": 1.5552000000000001, + "grad_norm": 0.9040723443031311, + "learning_rate": 1.1098494754914607e-05, + "loss": 0.3203, + "step": 2187 + }, + { + "epoch": 1.555911111111111, + "grad_norm": 1.0469070672988892, + "learning_rate": 1.1090268403704751e-05, + "loss": 0.3003, + "step": 2188 + }, + { + "epoch": 1.5566222222222224, + "grad_norm": 1.214608907699585, + "learning_rate": 1.1082041305735426e-05, + "loss": 0.3127, + "step": 2189 + }, + { + "epoch": 1.5573333333333332, + "grad_norm": 0.8778345584869385, + "learning_rate": 1.1073813466641633e-05, + "loss": 0.2808, + "step": 2190 + }, + { + "epoch": 1.5580444444444446, + "grad_norm": 1.0719598531723022, + "learning_rate": 1.1065584892058873e-05, + "loss": 0.3176, + "step": 2191 + }, + { + "epoch": 1.5587555555555554, + "grad_norm": 0.9143074750900269, + "learning_rate": 1.1057355587623168e-05, + "loss": 0.3003, + "step": 2192 + }, + { + "epoch": 1.5594666666666668, + "grad_norm": 1.087792992591858, + "learning_rate": 1.1049125558971024e-05, + "loss": 0.3049, + "step": 2193 + }, + { + "epoch": 1.5601777777777777, + "grad_norm": 1.0526504516601562, + "learning_rate": 1.1040894811739449e-05, + "loss": 0.3207, + "step": 2194 + }, + { + "epoch": 1.560888888888889, + "grad_norm": 0.9043090343475342, + "learning_rate": 1.1032663351565947e-05, + "loss": 0.3473, + "step": 2195 + }, + { + "epoch": 1.5615999999999999, + "grad_norm": 1.0102941989898682, + "learning_rate": 1.1024431184088505e-05, + "loss": 0.3336, + "step": 2196 + }, + { + "epoch": 1.5623111111111112, + "grad_norm": 0.9415282011032104, + "learning_rate": 1.1016198314945594e-05, + "loss": 0.2779, + "step": 2197 + }, + { + "epoch": 1.563022222222222, + "grad_norm": 0.9936779737472534, + "learning_rate": 1.1007964749776167e-05, + "loss": 0.3484, + "step": 2198 + }, + { + "epoch": 1.5637333333333334, + "grad_norm": 1.0516575574874878, + "learning_rate": 1.099973049421966e-05, + "loss": 0.2976, + "step": 2199 + }, + { + "epoch": 1.5644444444444443, + "grad_norm": 1.0810068845748901, + "learning_rate": 1.0991495553915974e-05, + "loss": 0.3534, + "step": 2200 + }, + { + "epoch": 1.5651555555555556, + "grad_norm": 1.052649736404419, + "learning_rate": 1.0983259934505479e-05, + "loss": 0.3639, + "step": 2201 + }, + { + "epoch": 1.5658666666666665, + "grad_norm": 0.9147171378135681, + "learning_rate": 1.0975023641629013e-05, + "loss": 0.3153, + "step": 2202 + }, + { + "epoch": 1.5665777777777778, + "grad_norm": 1.1616356372833252, + "learning_rate": 1.0966786680927875e-05, + "loss": 0.3348, + "step": 2203 + }, + { + "epoch": 1.5672888888888887, + "grad_norm": 1.3010486364364624, + "learning_rate": 1.0958549058043821e-05, + "loss": 0.3687, + "step": 2204 + }, + { + "epoch": 1.568, + "grad_norm": 1.0312206745147705, + "learning_rate": 1.0950310778619062e-05, + "loss": 0.341, + "step": 2205 + }, + { + "epoch": 1.568711111111111, + "grad_norm": 1.5693984031677246, + "learning_rate": 1.0942071848296257e-05, + "loss": 0.3165, + "step": 2206 + }, + { + "epoch": 1.5694222222222223, + "grad_norm": 1.4958378076553345, + "learning_rate": 1.0933832272718508e-05, + "loss": 0.3725, + "step": 2207 + }, + { + "epoch": 1.5701333333333334, + "grad_norm": 0.9749859571456909, + "learning_rate": 1.0925592057529364e-05, + "loss": 0.3197, + "step": 2208 + }, + { + "epoch": 1.5708444444444445, + "grad_norm": 1.7440361976623535, + "learning_rate": 1.0917351208372813e-05, + "loss": 0.3404, + "step": 2209 + }, + { + "epoch": 1.5715555555555556, + "grad_norm": 1.524501919746399, + "learning_rate": 1.0909109730893273e-05, + "loss": 0.3474, + "step": 2210 + }, + { + "epoch": 1.5722666666666667, + "grad_norm": 0.9472880959510803, + "learning_rate": 1.0900867630735592e-05, + "loss": 0.2671, + "step": 2211 + }, + { + "epoch": 1.5729777777777778, + "grad_norm": 1.152477741241455, + "learning_rate": 1.0892624913545046e-05, + "loss": 0.3672, + "step": 2212 + }, + { + "epoch": 1.573688888888889, + "grad_norm": 1.2934194803237915, + "learning_rate": 1.0884381584967337e-05, + "loss": 0.3122, + "step": 2213 + }, + { + "epoch": 1.5744, + "grad_norm": 1.6033523082733154, + "learning_rate": 1.0876137650648579e-05, + "loss": 0.342, + "step": 2214 + }, + { + "epoch": 1.5751111111111111, + "grad_norm": 1.3994086980819702, + "learning_rate": 1.0867893116235308e-05, + "loss": 0.3239, + "step": 2215 + }, + { + "epoch": 1.5758222222222222, + "grad_norm": 1.0733106136322021, + "learning_rate": 1.0859647987374467e-05, + "loss": 0.3156, + "step": 2216 + }, + { + "epoch": 1.5765333333333333, + "grad_norm": 1.266145944595337, + "learning_rate": 1.0851402269713403e-05, + "loss": 0.3116, + "step": 2217 + }, + { + "epoch": 1.5772444444444444, + "grad_norm": 0.846613883972168, + "learning_rate": 1.0843155968899875e-05, + "loss": 0.3162, + "step": 2218 + }, + { + "epoch": 1.5779555555555556, + "grad_norm": 0.9710591435432434, + "learning_rate": 1.0834909090582034e-05, + "loss": 0.2791, + "step": 2219 + }, + { + "epoch": 1.5786666666666667, + "grad_norm": 0.8732789158821106, + "learning_rate": 1.0826661640408427e-05, + "loss": 0.3311, + "step": 2220 + }, + { + "epoch": 1.5793777777777778, + "grad_norm": 0.8803874254226685, + "learning_rate": 1.0818413624027997e-05, + "loss": 0.314, + "step": 2221 + }, + { + "epoch": 1.5800888888888889, + "grad_norm": 1.0695397853851318, + "learning_rate": 1.0810165047090076e-05, + "loss": 0.3401, + "step": 2222 + }, + { + "epoch": 1.5808, + "grad_norm": 1.2143467664718628, + "learning_rate": 1.080191591524437e-05, + "loss": 0.3374, + "step": 2223 + }, + { + "epoch": 1.581511111111111, + "grad_norm": 0.9951425194740295, + "learning_rate": 1.0793666234140974e-05, + "loss": 0.3154, + "step": 2224 + }, + { + "epoch": 1.5822222222222222, + "grad_norm": 1.4517940282821655, + "learning_rate": 1.0785416009430355e-05, + "loss": 0.3189, + "step": 2225 + }, + { + "epoch": 1.5829333333333333, + "grad_norm": 0.8870776295661926, + "learning_rate": 1.0777165246763357e-05, + "loss": 0.3434, + "step": 2226 + }, + { + "epoch": 1.5836444444444444, + "grad_norm": 1.166549801826477, + "learning_rate": 1.076891395179119e-05, + "loss": 0.2616, + "step": 2227 + }, + { + "epoch": 1.5843555555555555, + "grad_norm": 1.0979877710342407, + "learning_rate": 1.0760662130165426e-05, + "loss": 0.3379, + "step": 2228 + }, + { + "epoch": 1.5850666666666666, + "grad_norm": 0.8411021828651428, + "learning_rate": 1.0752409787538e-05, + "loss": 0.3564, + "step": 2229 + }, + { + "epoch": 1.5857777777777777, + "grad_norm": 0.8912238478660583, + "learning_rate": 1.0744156929561206e-05, + "loss": 0.3524, + "step": 2230 + }, + { + "epoch": 1.5864888888888888, + "grad_norm": 0.9832900166511536, + "learning_rate": 1.0735903561887686e-05, + "loss": 0.31, + "step": 2231 + }, + { + "epoch": 1.5872000000000002, + "grad_norm": 0.8305566906929016, + "learning_rate": 1.0727649690170434e-05, + "loss": 0.2917, + "step": 2232 + }, + { + "epoch": 1.587911111111111, + "grad_norm": 1.0517387390136719, + "learning_rate": 1.071939532006279e-05, + "loss": 0.3026, + "step": 2233 + }, + { + "epoch": 1.5886222222222224, + "grad_norm": 0.9163334369659424, + "learning_rate": 1.0711140457218435e-05, + "loss": 0.2981, + "step": 2234 + }, + { + "epoch": 1.5893333333333333, + "grad_norm": 1.200811505317688, + "learning_rate": 1.0702885107291383e-05, + "loss": 0.3418, + "step": 2235 + }, + { + "epoch": 1.5900444444444446, + "grad_norm": 0.9205977916717529, + "learning_rate": 1.0694629275935989e-05, + "loss": 0.2831, + "step": 2236 + }, + { + "epoch": 1.5907555555555555, + "grad_norm": 1.2511857748031616, + "learning_rate": 1.0686372968806934e-05, + "loss": 0.3457, + "step": 2237 + }, + { + "epoch": 1.5914666666666668, + "grad_norm": 0.9048228859901428, + "learning_rate": 1.0678116191559222e-05, + "loss": 0.2972, + "step": 2238 + }, + { + "epoch": 1.5921777777777777, + "grad_norm": 1.3304879665374756, + "learning_rate": 1.0669858949848186e-05, + "loss": 0.3492, + "step": 2239 + }, + { + "epoch": 1.592888888888889, + "grad_norm": 0.8324533104896545, + "learning_rate": 1.0661601249329472e-05, + "loss": 0.2906, + "step": 2240 + }, + { + "epoch": 1.5936, + "grad_norm": 0.9805739521980286, + "learning_rate": 1.0653343095659042e-05, + "loss": 0.3117, + "step": 2241 + }, + { + "epoch": 1.5943111111111112, + "grad_norm": 1.039871096611023, + "learning_rate": 1.0645084494493166e-05, + "loss": 0.3143, + "step": 2242 + }, + { + "epoch": 1.5950222222222221, + "grad_norm": 0.8251035809516907, + "learning_rate": 1.0636825451488422e-05, + "loss": 0.3008, + "step": 2243 + }, + { + "epoch": 1.5957333333333334, + "grad_norm": 1.297406554222107, + "learning_rate": 1.0628565972301694e-05, + "loss": 0.3401, + "step": 2244 + }, + { + "epoch": 1.5964444444444443, + "grad_norm": 0.9985034465789795, + "learning_rate": 1.0620306062590158e-05, + "loss": 0.3179, + "step": 2245 + }, + { + "epoch": 1.5971555555555557, + "grad_norm": 1.061869502067566, + "learning_rate": 1.0612045728011294e-05, + "loss": 0.3047, + "step": 2246 + }, + { + "epoch": 1.5978666666666665, + "grad_norm": 0.9505857229232788, + "learning_rate": 1.0603784974222862e-05, + "loss": 0.335, + "step": 2247 + }, + { + "epoch": 1.5985777777777779, + "grad_norm": 0.9365562796592712, + "learning_rate": 1.0595523806882916e-05, + "loss": 0.3473, + "step": 2248 + }, + { + "epoch": 1.5992888888888888, + "grad_norm": 0.9698362946510315, + "learning_rate": 1.0587262231649795e-05, + "loss": 0.3695, + "step": 2249 + }, + { + "epoch": 1.6, + "grad_norm": 0.795159101486206, + "learning_rate": 1.0579000254182112e-05, + "loss": 0.3151, + "step": 2250 + }, + { + "epoch": 1.600711111111111, + "grad_norm": 1.0474199056625366, + "learning_rate": 1.0570737880138757e-05, + "loss": 0.3859, + "step": 2251 + }, + { + "epoch": 1.6014222222222223, + "grad_norm": 1.15534508228302, + "learning_rate": 1.0562475115178896e-05, + "loss": 0.3225, + "step": 2252 + }, + { + "epoch": 1.6021333333333332, + "grad_norm": 0.8872203826904297, + "learning_rate": 1.0554211964961958e-05, + "loss": 0.3402, + "step": 2253 + }, + { + "epoch": 1.6028444444444445, + "grad_norm": 0.9757724404335022, + "learning_rate": 1.0545948435147633e-05, + "loss": 0.3419, + "step": 2254 + }, + { + "epoch": 1.6035555555555554, + "grad_norm": 0.9866970777511597, + "learning_rate": 1.053768453139588e-05, + "loss": 0.3092, + "step": 2255 + }, + { + "epoch": 1.6042666666666667, + "grad_norm": 0.9997326731681824, + "learning_rate": 1.0529420259366907e-05, + "loss": 0.2971, + "step": 2256 + }, + { + "epoch": 1.6049777777777776, + "grad_norm": 1.0893367528915405, + "learning_rate": 1.0521155624721175e-05, + "loss": 0.3408, + "step": 2257 + }, + { + "epoch": 1.605688888888889, + "grad_norm": 0.9673916697502136, + "learning_rate": 1.05128906331194e-05, + "loss": 0.331, + "step": 2258 + }, + { + "epoch": 1.6064, + "grad_norm": 0.911054253578186, + "learning_rate": 1.0504625290222532e-05, + "loss": 0.308, + "step": 2259 + }, + { + "epoch": 1.6071111111111112, + "grad_norm": 0.8583958148956299, + "learning_rate": 1.0496359601691768e-05, + "loss": 0.2836, + "step": 2260 + }, + { + "epoch": 1.6078222222222223, + "grad_norm": 0.9622536301612854, + "learning_rate": 1.0488093573188542e-05, + "loss": 0.3613, + "step": 2261 + }, + { + "epoch": 1.6085333333333334, + "grad_norm": 1.0959736108779907, + "learning_rate": 1.0479827210374525e-05, + "loss": 0.3512, + "step": 2262 + }, + { + "epoch": 1.6092444444444445, + "grad_norm": 0.8651695847511292, + "learning_rate": 1.0471560518911601e-05, + "loss": 0.2916, + "step": 2263 + }, + { + "epoch": 1.6099555555555556, + "grad_norm": 0.9027907848358154, + "learning_rate": 1.0463293504461898e-05, + "loss": 0.3008, + "step": 2264 + }, + { + "epoch": 1.6106666666666667, + "grad_norm": 0.8730674982070923, + "learning_rate": 1.045502617268775e-05, + "loss": 0.3328, + "step": 2265 + }, + { + "epoch": 1.6113777777777778, + "grad_norm": 1.0336840152740479, + "learning_rate": 1.044675852925172e-05, + "loss": 0.3596, + "step": 2266 + }, + { + "epoch": 1.612088888888889, + "grad_norm": 0.9181562662124634, + "learning_rate": 1.0438490579816578e-05, + "loss": 0.3184, + "step": 2267 + }, + { + "epoch": 1.6128, + "grad_norm": 0.7724072933197021, + "learning_rate": 1.0430222330045306e-05, + "loss": 0.2918, + "step": 2268 + }, + { + "epoch": 1.6135111111111111, + "grad_norm": 0.9560757875442505, + "learning_rate": 1.0421953785601091e-05, + "loss": 0.3069, + "step": 2269 + }, + { + "epoch": 1.6142222222222222, + "grad_norm": 0.8410465717315674, + "learning_rate": 1.041368495214732e-05, + "loss": 0.3286, + "step": 2270 + }, + { + "epoch": 1.6149333333333333, + "grad_norm": 1.268462061882019, + "learning_rate": 1.0405415835347584e-05, + "loss": 0.3092, + "step": 2271 + }, + { + "epoch": 1.6156444444444444, + "grad_norm": 0.8140009641647339, + "learning_rate": 1.0397146440865658e-05, + "loss": 0.3162, + "step": 2272 + }, + { + "epoch": 1.6163555555555555, + "grad_norm": 0.9721074104309082, + "learning_rate": 1.0388876774365517e-05, + "loss": 0.3214, + "step": 2273 + }, + { + "epoch": 1.6170666666666667, + "grad_norm": 1.108055830001831, + "learning_rate": 1.038060684151132e-05, + "loss": 0.2887, + "step": 2274 + }, + { + "epoch": 1.6177777777777778, + "grad_norm": 0.9702160358428955, + "learning_rate": 1.0372336647967403e-05, + "loss": 0.3169, + "step": 2275 + }, + { + "epoch": 1.6184888888888889, + "grad_norm": 1.1693552732467651, + "learning_rate": 1.0364066199398285e-05, + "loss": 0.3048, + "step": 2276 + }, + { + "epoch": 1.6192, + "grad_norm": 1.2937742471694946, + "learning_rate": 1.0355795501468655e-05, + "loss": 0.3186, + "step": 2277 + }, + { + "epoch": 1.619911111111111, + "grad_norm": 1.1894135475158691, + "learning_rate": 1.0347524559843385e-05, + "loss": 0.3016, + "step": 2278 + }, + { + "epoch": 1.6206222222222222, + "grad_norm": 1.1828880310058594, + "learning_rate": 1.0339253380187502e-05, + "loss": 0.2769, + "step": 2279 + }, + { + "epoch": 1.6213333333333333, + "grad_norm": 0.9668726325035095, + "learning_rate": 1.03309819681662e-05, + "loss": 0.2981, + "step": 2280 + }, + { + "epoch": 1.6220444444444444, + "grad_norm": 1.6937973499298096, + "learning_rate": 1.0322710329444831e-05, + "loss": 0.3089, + "step": 2281 + }, + { + "epoch": 1.6227555555555555, + "grad_norm": 1.2278549671173096, + "learning_rate": 1.0314438469688905e-05, + "loss": 0.2871, + "step": 2282 + }, + { + "epoch": 1.6234666666666666, + "grad_norm": 0.9259233474731445, + "learning_rate": 1.030616639456408e-05, + "loss": 0.2928, + "step": 2283 + }, + { + "epoch": 1.6241777777777777, + "grad_norm": 1.6678519248962402, + "learning_rate": 1.0297894109736165e-05, + "loss": 0.3122, + "step": 2284 + }, + { + "epoch": 1.624888888888889, + "grad_norm": 1.3482153415679932, + "learning_rate": 1.0289621620871106e-05, + "loss": 0.2936, + "step": 2285 + }, + { + "epoch": 1.6256, + "grad_norm": 1.1843116283416748, + "learning_rate": 1.0281348933634997e-05, + "loss": 0.2764, + "step": 2286 + }, + { + "epoch": 1.6263111111111113, + "grad_norm": 1.2801930904388428, + "learning_rate": 1.0273076053694067e-05, + "loss": 0.2603, + "step": 2287 + }, + { + "epoch": 1.6270222222222221, + "grad_norm": 0.9063803553581238, + "learning_rate": 1.026480298671467e-05, + "loss": 0.3486, + "step": 2288 + }, + { + "epoch": 1.6277333333333335, + "grad_norm": 1.102111577987671, + "learning_rate": 1.0256529738363292e-05, + "loss": 0.3456, + "step": 2289 + }, + { + "epoch": 1.6284444444444444, + "grad_norm": 1.2714719772338867, + "learning_rate": 1.0248256314306547e-05, + "loss": 0.3496, + "step": 2290 + }, + { + "epoch": 1.6291555555555557, + "grad_norm": 1.430924892425537, + "learning_rate": 1.0239982720211164e-05, + "loss": 0.3006, + "step": 2291 + }, + { + "epoch": 1.6298666666666666, + "grad_norm": 1.3930686712265015, + "learning_rate": 1.0231708961743991e-05, + "loss": 0.3475, + "step": 2292 + }, + { + "epoch": 1.630577777777778, + "grad_norm": 1.3275595903396606, + "learning_rate": 1.0223435044571987e-05, + "loss": 0.3026, + "step": 2293 + }, + { + "epoch": 1.6312888888888888, + "grad_norm": 1.0124094486236572, + "learning_rate": 1.0215160974362224e-05, + "loss": 0.3223, + "step": 2294 + }, + { + "epoch": 1.6320000000000001, + "grad_norm": 1.304327368736267, + "learning_rate": 1.020688675678187e-05, + "loss": 0.2718, + "step": 2295 + }, + { + "epoch": 1.632711111111111, + "grad_norm": 0.8258498907089233, + "learning_rate": 1.0198612397498207e-05, + "loss": 0.3415, + "step": 2296 + }, + { + "epoch": 1.6334222222222223, + "grad_norm": 1.0637232065200806, + "learning_rate": 1.01903379021786e-05, + "loss": 0.2998, + "step": 2297 + }, + { + "epoch": 1.6341333333333332, + "grad_norm": 0.9016327261924744, + "learning_rate": 1.0182063276490515e-05, + "loss": 0.3036, + "step": 2298 + }, + { + "epoch": 1.6348444444444445, + "grad_norm": 1.0266599655151367, + "learning_rate": 1.017378852610151e-05, + "loss": 0.3144, + "step": 2299 + }, + { + "epoch": 1.6355555555555554, + "grad_norm": 0.9603660106658936, + "learning_rate": 1.016551365667922e-05, + "loss": 0.2985, + "step": 2300 + }, + { + "epoch": 1.6362666666666668, + "grad_norm": 0.8281471133232117, + "learning_rate": 1.0157238673891368e-05, + "loss": 0.2963, + "step": 2301 + }, + { + "epoch": 1.6369777777777776, + "grad_norm": 0.9733481407165527, + "learning_rate": 1.0148963583405751e-05, + "loss": 0.3255, + "step": 2302 + }, + { + "epoch": 1.637688888888889, + "grad_norm": 1.0669150352478027, + "learning_rate": 1.0140688390890243e-05, + "loss": 0.3114, + "step": 2303 + }, + { + "epoch": 1.6383999999999999, + "grad_norm": 1.0364471673965454, + "learning_rate": 1.0132413102012788e-05, + "loss": 0.2781, + "step": 2304 + }, + { + "epoch": 1.6391111111111112, + "grad_norm": 0.9933536052703857, + "learning_rate": 1.0124137722441394e-05, + "loss": 0.3531, + "step": 2305 + }, + { + "epoch": 1.639822222222222, + "grad_norm": 1.3185728788375854, + "learning_rate": 1.0115862257844124e-05, + "loss": 0.293, + "step": 2306 + }, + { + "epoch": 1.6405333333333334, + "grad_norm": 1.0906741619110107, + "learning_rate": 1.0107586713889117e-05, + "loss": 0.3357, + "step": 2307 + }, + { + "epoch": 1.6412444444444443, + "grad_norm": 1.3471808433532715, + "learning_rate": 1.0099311096244549e-05, + "loss": 0.2866, + "step": 2308 + }, + { + "epoch": 1.6419555555555556, + "grad_norm": 1.5939133167266846, + "learning_rate": 1.0091035410578656e-05, + "loss": 0.283, + "step": 2309 + }, + { + "epoch": 1.6426666666666667, + "grad_norm": 1.2914799451828003, + "learning_rate": 1.0082759662559718e-05, + "loss": 0.3282, + "step": 2310 + }, + { + "epoch": 1.6433777777777778, + "grad_norm": 1.470038652420044, + "learning_rate": 1.0074483857856056e-05, + "loss": 0.2923, + "step": 2311 + }, + { + "epoch": 1.644088888888889, + "grad_norm": 0.8832451701164246, + "learning_rate": 1.0066208002136033e-05, + "loss": 0.3362, + "step": 2312 + }, + { + "epoch": 1.6448, + "grad_norm": 1.3132169246673584, + "learning_rate": 1.0057932101068046e-05, + "loss": 0.3316, + "step": 2313 + }, + { + "epoch": 1.6455111111111111, + "grad_norm": 1.2329777479171753, + "learning_rate": 1.0049656160320524e-05, + "loss": 0.3065, + "step": 2314 + }, + { + "epoch": 1.6462222222222223, + "grad_norm": 1.157369613647461, + "learning_rate": 1.0041380185561922e-05, + "loss": 0.3631, + "step": 2315 + }, + { + "epoch": 1.6469333333333334, + "grad_norm": 1.0468432903289795, + "learning_rate": 1.0033104182460714e-05, + "loss": 0.3224, + "step": 2316 + }, + { + "epoch": 1.6476444444444445, + "grad_norm": 0.9328768253326416, + "learning_rate": 1.00248281566854e-05, + "loss": 0.3279, + "step": 2317 + }, + { + "epoch": 1.6483555555555556, + "grad_norm": 1.2093514204025269, + "learning_rate": 1.0016552113904492e-05, + "loss": 0.2957, + "step": 2318 + }, + { + "epoch": 1.6490666666666667, + "grad_norm": 0.9207165241241455, + "learning_rate": 1.0008276059786514e-05, + "loss": 0.276, + "step": 2319 + }, + { + "epoch": 1.6497777777777778, + "grad_norm": 1.011078953742981, + "learning_rate": 1e-05, + "loss": 0.2795, + "step": 2320 + }, + { + "epoch": 1.650488888888889, + "grad_norm": 1.1982479095458984, + "learning_rate": 9.991723940213486e-06, + "loss": 0.3128, + "step": 2321 + }, + { + "epoch": 1.6512, + "grad_norm": 1.044519305229187, + "learning_rate": 9.983447886095512e-06, + "loss": 0.3165, + "step": 2322 + }, + { + "epoch": 1.651911111111111, + "grad_norm": 0.9618070125579834, + "learning_rate": 9.975171843314601e-06, + "loss": 0.3466, + "step": 2323 + }, + { + "epoch": 1.6526222222222222, + "grad_norm": 0.9991148114204407, + "learning_rate": 9.966895817539288e-06, + "loss": 0.2888, + "step": 2324 + }, + { + "epoch": 1.6533333333333333, + "grad_norm": 0.895054280757904, + "learning_rate": 9.958619814438081e-06, + "loss": 0.3013, + "step": 2325 + }, + { + "epoch": 1.6540444444444444, + "grad_norm": 1.07194185256958, + "learning_rate": 9.950343839679478e-06, + "loss": 0.3082, + "step": 2326 + }, + { + "epoch": 1.6547555555555555, + "grad_norm": 1.0105996131896973, + "learning_rate": 9.942067898931953e-06, + "loss": 0.2913, + "step": 2327 + }, + { + "epoch": 1.6554666666666666, + "grad_norm": 0.792972981929779, + "learning_rate": 9.93379199786397e-06, + "loss": 0.2914, + "step": 2328 + }, + { + "epoch": 1.6561777777777777, + "grad_norm": 1.0036022663116455, + "learning_rate": 9.925516142143947e-06, + "loss": 0.2723, + "step": 2329 + }, + { + "epoch": 1.6568888888888889, + "grad_norm": 0.9416024088859558, + "learning_rate": 9.917240337440288e-06, + "loss": 0.2867, + "step": 2330 + }, + { + "epoch": 1.6576, + "grad_norm": 1.236242651939392, + "learning_rate": 9.908964589421347e-06, + "loss": 0.3164, + "step": 2331 + }, + { + "epoch": 1.658311111111111, + "grad_norm": 1.2025145292282104, + "learning_rate": 9.900688903755456e-06, + "loss": 0.3438, + "step": 2332 + }, + { + "epoch": 1.6590222222222222, + "grad_norm": 1.3977911472320557, + "learning_rate": 9.892413286110886e-06, + "loss": 0.3044, + "step": 2333 + }, + { + "epoch": 1.6597333333333333, + "grad_norm": 1.4708276987075806, + "learning_rate": 9.88413774215588e-06, + "loss": 0.3453, + "step": 2334 + }, + { + "epoch": 1.6604444444444444, + "grad_norm": 0.9992352724075317, + "learning_rate": 9.87586227755861e-06, + "loss": 0.3412, + "step": 2335 + }, + { + "epoch": 1.6611555555555557, + "grad_norm": 1.403929352760315, + "learning_rate": 9.867586897987214e-06, + "loss": 0.3282, + "step": 2336 + }, + { + "epoch": 1.6618666666666666, + "grad_norm": 1.2323157787322998, + "learning_rate": 9.859311609109758e-06, + "loss": 0.3112, + "step": 2337 + }, + { + "epoch": 1.662577777777778, + "grad_norm": 1.3156667947769165, + "learning_rate": 9.851036416594249e-06, + "loss": 0.279, + "step": 2338 + }, + { + "epoch": 1.6632888888888888, + "grad_norm": 1.5239284038543701, + "learning_rate": 9.842761326108637e-06, + "loss": 0.3016, + "step": 2339 + }, + { + "epoch": 1.6640000000000001, + "grad_norm": 1.1974745988845825, + "learning_rate": 9.834486343320782e-06, + "loss": 0.3146, + "step": 2340 + }, + { + "epoch": 1.664711111111111, + "grad_norm": 0.9425774216651917, + "learning_rate": 9.826211473898494e-06, + "loss": 0.3096, + "step": 2341 + }, + { + "epoch": 1.6654222222222224, + "grad_norm": 1.533220887184143, + "learning_rate": 9.817936723509485e-06, + "loss": 0.2956, + "step": 2342 + }, + { + "epoch": 1.6661333333333332, + "grad_norm": 1.4072816371917725, + "learning_rate": 9.809662097821404e-06, + "loss": 0.3277, + "step": 2343 + }, + { + "epoch": 1.6668444444444446, + "grad_norm": 1.198259711265564, + "learning_rate": 9.801387602501795e-06, + "loss": 0.3148, + "step": 2344 + }, + { + "epoch": 1.6675555555555555, + "grad_norm": 1.3175500631332397, + "learning_rate": 9.793113243218131e-06, + "loss": 0.3445, + "step": 2345 + }, + { + "epoch": 1.6682666666666668, + "grad_norm": 1.2379541397094727, + "learning_rate": 9.78483902563778e-06, + "loss": 0.3546, + "step": 2346 + }, + { + "epoch": 1.6689777777777777, + "grad_norm": 0.8862751126289368, + "learning_rate": 9.776564955428015e-06, + "loss": 0.2844, + "step": 2347 + }, + { + "epoch": 1.669688888888889, + "grad_norm": 0.9087072014808655, + "learning_rate": 9.76829103825601e-06, + "loss": 0.2929, + "step": 2348 + }, + { + "epoch": 1.6703999999999999, + "grad_norm": 0.9455989599227905, + "learning_rate": 9.760017279788841e-06, + "loss": 0.2935, + "step": 2349 + }, + { + "epoch": 1.6711111111111112, + "grad_norm": 0.8876254558563232, + "learning_rate": 9.751743685693455e-06, + "loss": 0.3188, + "step": 2350 + }, + { + "epoch": 1.671822222222222, + "grad_norm": 1.394474983215332, + "learning_rate": 9.743470261636713e-06, + "loss": 0.3117, + "step": 2351 + }, + { + "epoch": 1.6725333333333334, + "grad_norm": 1.0598961114883423, + "learning_rate": 9.735197013285334e-06, + "loss": 0.3064, + "step": 2352 + }, + { + "epoch": 1.6732444444444443, + "grad_norm": 1.4159256219863892, + "learning_rate": 9.726923946305938e-06, + "loss": 0.3289, + "step": 2353 + }, + { + "epoch": 1.6739555555555556, + "grad_norm": 1.00727379322052, + "learning_rate": 9.718651066365004e-06, + "loss": 0.3105, + "step": 2354 + }, + { + "epoch": 1.6746666666666665, + "grad_norm": 0.9932565689086914, + "learning_rate": 9.710378379128897e-06, + "loss": 0.2836, + "step": 2355 + }, + { + "epoch": 1.6753777777777779, + "grad_norm": 1.0650088787078857, + "learning_rate": 9.702105890263839e-06, + "loss": 0.3153, + "step": 2356 + }, + { + "epoch": 1.6760888888888887, + "grad_norm": 0.9466317296028137, + "learning_rate": 9.693833605435922e-06, + "loss": 0.3534, + "step": 2357 + }, + { + "epoch": 1.6768, + "grad_norm": 1.1230615377426147, + "learning_rate": 9.685561530311098e-06, + "loss": 0.3003, + "step": 2358 + }, + { + "epoch": 1.677511111111111, + "grad_norm": 1.115964412689209, + "learning_rate": 9.67728967055517e-06, + "loss": 0.2816, + "step": 2359 + }, + { + "epoch": 1.6782222222222223, + "grad_norm": 0.9345912337303162, + "learning_rate": 9.669018031833803e-06, + "loss": 0.2859, + "step": 2360 + }, + { + "epoch": 1.6789333333333334, + "grad_norm": 1.5221635103225708, + "learning_rate": 9.660746619812498e-06, + "loss": 0.3277, + "step": 2361 + }, + { + "epoch": 1.6796444444444445, + "grad_norm": 0.937072217464447, + "learning_rate": 9.652475440156618e-06, + "loss": 0.2927, + "step": 2362 + }, + { + "epoch": 1.6803555555555556, + "grad_norm": 1.0652668476104736, + "learning_rate": 9.644204498531345e-06, + "loss": 0.287, + "step": 2363 + }, + { + "epoch": 1.6810666666666667, + "grad_norm": 0.9267204999923706, + "learning_rate": 9.635933800601722e-06, + "loss": 0.293, + "step": 2364 + }, + { + "epoch": 1.6817777777777778, + "grad_norm": 1.0353270769119263, + "learning_rate": 9.627663352032602e-06, + "loss": 0.3024, + "step": 2365 + }, + { + "epoch": 1.682488888888889, + "grad_norm": 0.9170510768890381, + "learning_rate": 9.619393158488684e-06, + "loss": 0.2947, + "step": 2366 + }, + { + "epoch": 1.6832, + "grad_norm": 1.1303682327270508, + "learning_rate": 9.611123225634486e-06, + "loss": 0.3066, + "step": 2367 + }, + { + "epoch": 1.6839111111111111, + "grad_norm": 1.2191321849822998, + "learning_rate": 9.602853559134345e-06, + "loss": 0.3005, + "step": 2368 + }, + { + "epoch": 1.6846222222222222, + "grad_norm": 1.095065951347351, + "learning_rate": 9.59458416465242e-06, + "loss": 0.3268, + "step": 2369 + }, + { + "epoch": 1.6853333333333333, + "grad_norm": 1.1531301736831665, + "learning_rate": 9.586315047852685e-06, + "loss": 0.3444, + "step": 2370 + }, + { + "epoch": 1.6860444444444445, + "grad_norm": 1.3774409294128418, + "learning_rate": 9.578046214398912e-06, + "loss": 0.3265, + "step": 2371 + }, + { + "epoch": 1.6867555555555556, + "grad_norm": 1.0347107648849487, + "learning_rate": 9.569777669954694e-06, + "loss": 0.3305, + "step": 2372 + }, + { + "epoch": 1.6874666666666667, + "grad_norm": 1.4924397468566895, + "learning_rate": 9.561509420183424e-06, + "loss": 0.2876, + "step": 2373 + }, + { + "epoch": 1.6881777777777778, + "grad_norm": 1.4535497426986694, + "learning_rate": 9.553241470748282e-06, + "loss": 0.3326, + "step": 2374 + }, + { + "epoch": 1.6888888888888889, + "grad_norm": 1.1733615398406982, + "learning_rate": 9.544973827312252e-06, + "loss": 0.3408, + "step": 2375 + }, + { + "epoch": 1.6896, + "grad_norm": 1.2991629838943481, + "learning_rate": 9.536706495538106e-06, + "loss": 0.2721, + "step": 2376 + }, + { + "epoch": 1.690311111111111, + "grad_norm": 0.8506069779396057, + "learning_rate": 9.5284394810884e-06, + "loss": 0.3019, + "step": 2377 + }, + { + "epoch": 1.6910222222222222, + "grad_norm": 1.1887857913970947, + "learning_rate": 9.520172789625478e-06, + "loss": 0.2995, + "step": 2378 + }, + { + "epoch": 1.6917333333333333, + "grad_norm": 0.9380427598953247, + "learning_rate": 9.51190642681146e-06, + "loss": 0.3459, + "step": 2379 + }, + { + "epoch": 1.6924444444444444, + "grad_norm": 1.0044454336166382, + "learning_rate": 9.503640398308232e-06, + "loss": 0.3077, + "step": 2380 + }, + { + "epoch": 1.6931555555555555, + "grad_norm": 1.033432960510254, + "learning_rate": 9.495374709777473e-06, + "loss": 0.3127, + "step": 2381 + }, + { + "epoch": 1.6938666666666666, + "grad_norm": 0.947831928730011, + "learning_rate": 9.487109366880604e-06, + "loss": 0.2666, + "step": 2382 + }, + { + "epoch": 1.6945777777777777, + "grad_norm": 1.2585060596466064, + "learning_rate": 9.47884437527883e-06, + "loss": 0.3226, + "step": 2383 + }, + { + "epoch": 1.6952888888888888, + "grad_norm": 1.0555614233016968, + "learning_rate": 9.470579740633096e-06, + "loss": 0.2649, + "step": 2384 + }, + { + "epoch": 1.696, + "grad_norm": 1.4933278560638428, + "learning_rate": 9.462315468604126e-06, + "loss": 0.3075, + "step": 2385 + }, + { + "epoch": 1.696711111111111, + "grad_norm": 1.4623199701309204, + "learning_rate": 9.454051564852368e-06, + "loss": 0.2375, + "step": 2386 + }, + { + "epoch": 1.6974222222222224, + "grad_norm": 0.8875882625579834, + "learning_rate": 9.445788035038047e-06, + "loss": 0.324, + "step": 2387 + }, + { + "epoch": 1.6981333333333333, + "grad_norm": 1.8792895078659058, + "learning_rate": 9.437524884821106e-06, + "loss": 0.2871, + "step": 2388 + }, + { + "epoch": 1.6988444444444446, + "grad_norm": 1.2965513467788696, + "learning_rate": 9.429262119861244e-06, + "loss": 0.2988, + "step": 2389 + }, + { + "epoch": 1.6995555555555555, + "grad_norm": 0.9470494389533997, + "learning_rate": 9.420999745817891e-06, + "loss": 0.2786, + "step": 2390 + }, + { + "epoch": 1.7002666666666668, + "grad_norm": 1.659560203552246, + "learning_rate": 9.412737768350205e-06, + "loss": 0.3061, + "step": 2391 + }, + { + "epoch": 1.7009777777777777, + "grad_norm": 1.8311232328414917, + "learning_rate": 9.404476193117085e-06, + "loss": 0.313, + "step": 2392 + }, + { + "epoch": 1.701688888888889, + "grad_norm": 1.0976054668426514, + "learning_rate": 9.39621502577714e-06, + "loss": 0.289, + "step": 2393 + }, + { + "epoch": 1.7024, + "grad_norm": 1.381176233291626, + "learning_rate": 9.38795427198871e-06, + "loss": 0.3101, + "step": 2394 + }, + { + "epoch": 1.7031111111111112, + "grad_norm": 1.7475749254226685, + "learning_rate": 9.379693937409842e-06, + "loss": 0.2947, + "step": 2395 + }, + { + "epoch": 1.7038222222222221, + "grad_norm": 1.4745254516601562, + "learning_rate": 9.371434027698309e-06, + "loss": 0.3351, + "step": 2396 + }, + { + "epoch": 1.7045333333333335, + "grad_norm": 1.1209315061569214, + "learning_rate": 9.363174548511578e-06, + "loss": 0.3014, + "step": 2397 + }, + { + "epoch": 1.7052444444444443, + "grad_norm": 1.703981876373291, + "learning_rate": 9.354915505506839e-06, + "loss": 0.2965, + "step": 2398 + }, + { + "epoch": 1.7059555555555557, + "grad_norm": 1.4203730821609497, + "learning_rate": 9.346656904340963e-06, + "loss": 0.3385, + "step": 2399 + }, + { + "epoch": 1.7066666666666666, + "grad_norm": 0.988740086555481, + "learning_rate": 9.338398750670533e-06, + "loss": 0.3006, + "step": 2400 + }, + { + "epoch": 1.7073777777777779, + "grad_norm": 1.5214227437973022, + "learning_rate": 9.330141050151817e-06, + "loss": 0.333, + "step": 2401 + }, + { + "epoch": 1.7080888888888888, + "grad_norm": 1.8870924711227417, + "learning_rate": 9.321883808440784e-06, + "loss": 0.2941, + "step": 2402 + }, + { + "epoch": 1.7088, + "grad_norm": 1.0826079845428467, + "learning_rate": 9.313627031193071e-06, + "loss": 0.2991, + "step": 2403 + }, + { + "epoch": 1.709511111111111, + "grad_norm": 1.83572256565094, + "learning_rate": 9.305370724064016e-06, + "loss": 0.3099, + "step": 2404 + }, + { + "epoch": 1.7102222222222223, + "grad_norm": 1.71575129032135, + "learning_rate": 9.29711489270862e-06, + "loss": 0.2863, + "step": 2405 + }, + { + "epoch": 1.7109333333333332, + "grad_norm": 1.1343121528625488, + "learning_rate": 9.288859542781572e-06, + "loss": 0.2929, + "step": 2406 + }, + { + "epoch": 1.7116444444444445, + "grad_norm": 1.2251414060592651, + "learning_rate": 9.280604679937213e-06, + "loss": 0.3684, + "step": 2407 + }, + { + "epoch": 1.7123555555555554, + "grad_norm": 1.9201714992523193, + "learning_rate": 9.272350309829568e-06, + "loss": 0.29, + "step": 2408 + }, + { + "epoch": 1.7130666666666667, + "grad_norm": 1.2975449562072754, + "learning_rate": 9.264096438112317e-06, + "loss": 0.3156, + "step": 2409 + }, + { + "epoch": 1.7137777777777776, + "grad_norm": 1.3028841018676758, + "learning_rate": 9.255843070438795e-06, + "loss": 0.2898, + "step": 2410 + }, + { + "epoch": 1.714488888888889, + "grad_norm": 1.7121455669403076, + "learning_rate": 9.247590212462001e-06, + "loss": 0.3146, + "step": 2411 + }, + { + "epoch": 1.7151999999999998, + "grad_norm": 1.2650948762893677, + "learning_rate": 9.239337869834574e-06, + "loss": 0.2822, + "step": 2412 + }, + { + "epoch": 1.7159111111111112, + "grad_norm": 0.9904010891914368, + "learning_rate": 9.231086048208813e-06, + "loss": 0.3075, + "step": 2413 + }, + { + "epoch": 1.7166222222222223, + "grad_norm": 1.6256237030029297, + "learning_rate": 9.222834753236643e-06, + "loss": 0.3048, + "step": 2414 + }, + { + "epoch": 1.7173333333333334, + "grad_norm": 1.478918433189392, + "learning_rate": 9.214583990569648e-06, + "loss": 0.2802, + "step": 2415 + }, + { + "epoch": 1.7180444444444445, + "grad_norm": 1.049234390258789, + "learning_rate": 9.20633376585903e-06, + "loss": 0.2939, + "step": 2416 + }, + { + "epoch": 1.7187555555555556, + "grad_norm": 1.3227264881134033, + "learning_rate": 9.198084084755635e-06, + "loss": 0.3203, + "step": 2417 + }, + { + "epoch": 1.7194666666666667, + "grad_norm": 1.3347034454345703, + "learning_rate": 9.189834952909927e-06, + "loss": 0.29, + "step": 2418 + }, + { + "epoch": 1.7201777777777778, + "grad_norm": 1.0748311281204224, + "learning_rate": 9.181586375972006e-06, + "loss": 0.3155, + "step": 2419 + }, + { + "epoch": 1.720888888888889, + "grad_norm": 1.403921127319336, + "learning_rate": 9.173338359591578e-06, + "loss": 0.2891, + "step": 2420 + }, + { + "epoch": 1.7216, + "grad_norm": 1.6384270191192627, + "learning_rate": 9.165090909417973e-06, + "loss": 0.2847, + "step": 2421 + }, + { + "epoch": 1.7223111111111111, + "grad_norm": 1.1754292249679565, + "learning_rate": 9.15684403110013e-06, + "loss": 0.3034, + "step": 2422 + }, + { + "epoch": 1.7230222222222222, + "grad_norm": 1.4970953464508057, + "learning_rate": 9.148597730286602e-06, + "loss": 0.2795, + "step": 2423 + }, + { + "epoch": 1.7237333333333333, + "grad_norm": 1.540498971939087, + "learning_rate": 9.140352012625538e-06, + "loss": 0.2716, + "step": 2424 + }, + { + "epoch": 1.7244444444444444, + "grad_norm": 1.2614885568618774, + "learning_rate": 9.132106883764692e-06, + "loss": 0.3069, + "step": 2425 + }, + { + "epoch": 1.7251555555555556, + "grad_norm": 1.3876653909683228, + "learning_rate": 9.123862349351423e-06, + "loss": 0.2867, + "step": 2426 + }, + { + "epoch": 1.7258666666666667, + "grad_norm": 1.7879736423492432, + "learning_rate": 9.115618415032663e-06, + "loss": 0.332, + "step": 2427 + }, + { + "epoch": 1.7265777777777778, + "grad_norm": 1.4935444593429565, + "learning_rate": 9.107375086454956e-06, + "loss": 0.3368, + "step": 2428 + }, + { + "epoch": 1.7272888888888889, + "grad_norm": 1.4804750680923462, + "learning_rate": 9.099132369264411e-06, + "loss": 0.3037, + "step": 2429 + }, + { + "epoch": 1.728, + "grad_norm": 1.0134989023208618, + "learning_rate": 9.09089026910673e-06, + "loss": 0.3319, + "step": 2430 + }, + { + "epoch": 1.728711111111111, + "grad_norm": 1.2008188962936401, + "learning_rate": 9.082648791627189e-06, + "loss": 0.3291, + "step": 2431 + }, + { + "epoch": 1.7294222222222222, + "grad_norm": 1.0581779479980469, + "learning_rate": 9.07440794247064e-06, + "loss": 0.3194, + "step": 2432 + }, + { + "epoch": 1.7301333333333333, + "grad_norm": 2.123607873916626, + "learning_rate": 9.066167727281495e-06, + "loss": 0.3115, + "step": 2433 + }, + { + "epoch": 1.7308444444444444, + "grad_norm": 1.8083082437515259, + "learning_rate": 9.05792815170375e-06, + "loss": 0.2884, + "step": 2434 + }, + { + "epoch": 1.7315555555555555, + "grad_norm": 1.3329499959945679, + "learning_rate": 9.04968922138094e-06, + "loss": 0.3027, + "step": 2435 + }, + { + "epoch": 1.7322666666666666, + "grad_norm": 1.245384693145752, + "learning_rate": 9.041450941956184e-06, + "loss": 0.3045, + "step": 2436 + }, + { + "epoch": 1.7329777777777777, + "grad_norm": 1.148603916168213, + "learning_rate": 9.033213319072126e-06, + "loss": 0.2939, + "step": 2437 + }, + { + "epoch": 1.7336888888888888, + "grad_norm": 1.2206472158432007, + "learning_rate": 9.024976358370992e-06, + "loss": 0.2676, + "step": 2438 + }, + { + "epoch": 1.7344, + "grad_norm": 1.5090903043746948, + "learning_rate": 9.016740065494523e-06, + "loss": 0.2605, + "step": 2439 + }, + { + "epoch": 1.7351111111111113, + "grad_norm": 1.0836009979248047, + "learning_rate": 9.00850444608403e-06, + "loss": 0.3079, + "step": 2440 + }, + { + "epoch": 1.7358222222222222, + "grad_norm": 1.4062063694000244, + "learning_rate": 9.000269505780342e-06, + "loss": 0.3398, + "step": 2441 + }, + { + "epoch": 1.7365333333333335, + "grad_norm": 1.6221246719360352, + "learning_rate": 8.992035250223831e-06, + "loss": 0.2787, + "step": 2442 + }, + { + "epoch": 1.7372444444444444, + "grad_norm": 1.2397183179855347, + "learning_rate": 8.98380168505441e-06, + "loss": 0.2685, + "step": 2443 + }, + { + "epoch": 1.7379555555555557, + "grad_norm": 0.8790352940559387, + "learning_rate": 8.975568815911497e-06, + "loss": 0.2836, + "step": 2444 + }, + { + "epoch": 1.7386666666666666, + "grad_norm": 1.8789798021316528, + "learning_rate": 8.967336648434055e-06, + "loss": 0.2971, + "step": 2445 + }, + { + "epoch": 1.739377777777778, + "grad_norm": 1.8194379806518555, + "learning_rate": 8.95910518826055e-06, + "loss": 0.3142, + "step": 2446 + }, + { + "epoch": 1.7400888888888888, + "grad_norm": 1.1137025356292725, + "learning_rate": 8.95087444102898e-06, + "loss": 0.3172, + "step": 2447 + }, + { + "epoch": 1.7408000000000001, + "grad_norm": 1.591629981994629, + "learning_rate": 8.942644412376833e-06, + "loss": 0.28, + "step": 2448 + }, + { + "epoch": 1.741511111111111, + "grad_norm": 2.1425628662109375, + "learning_rate": 8.934415107941128e-06, + "loss": 0.3115, + "step": 2449 + }, + { + "epoch": 1.7422222222222223, + "grad_norm": 1.7320064306259155, + "learning_rate": 8.92618653335837e-06, + "loss": 0.304, + "step": 2450 + }, + { + "epoch": 1.7429333333333332, + "grad_norm": 0.878272294998169, + "learning_rate": 8.917958694264575e-06, + "loss": 0.2909, + "step": 2451 + }, + { + "epoch": 1.7436444444444446, + "grad_norm": 1.5235697031021118, + "learning_rate": 8.90973159629525e-06, + "loss": 0.3302, + "step": 2452 + }, + { + "epoch": 1.7443555555555554, + "grad_norm": 1.8924115896224976, + "learning_rate": 8.901505245085398e-06, + "loss": 0.2458, + "step": 2453 + }, + { + "epoch": 1.7450666666666668, + "grad_norm": 1.2743295431137085, + "learning_rate": 8.893279646269502e-06, + "loss": 0.317, + "step": 2454 + }, + { + "epoch": 1.7457777777777777, + "grad_norm": 0.9779114723205566, + "learning_rate": 8.885054805481546e-06, + "loss": 0.3387, + "step": 2455 + }, + { + "epoch": 1.746488888888889, + "grad_norm": 2.1881232261657715, + "learning_rate": 8.876830728354973e-06, + "loss": 0.3126, + "step": 2456 + }, + { + "epoch": 1.7471999999999999, + "grad_norm": 2.037285804748535, + "learning_rate": 8.868607420522725e-06, + "loss": 0.2646, + "step": 2457 + }, + { + "epoch": 1.7479111111111112, + "grad_norm": 1.2113865613937378, + "learning_rate": 8.860384887617198e-06, + "loss": 0.324, + "step": 2458 + }, + { + "epoch": 1.748622222222222, + "grad_norm": 1.3858423233032227, + "learning_rate": 8.852163135270274e-06, + "loss": 0.2977, + "step": 2459 + }, + { + "epoch": 1.7493333333333334, + "grad_norm": 1.9002503156661987, + "learning_rate": 8.843942169113289e-06, + "loss": 0.2625, + "step": 2460 + }, + { + "epoch": 1.7500444444444443, + "grad_norm": 1.565624475479126, + "learning_rate": 8.835721994777043e-06, + "loss": 0.3177, + "step": 2461 + }, + { + "epoch": 1.7507555555555556, + "grad_norm": 0.9459837675094604, + "learning_rate": 8.827502617891799e-06, + "loss": 0.2931, + "step": 2462 + }, + { + "epoch": 1.7514666666666665, + "grad_norm": 1.2895097732543945, + "learning_rate": 8.81928404408726e-06, + "loss": 0.3493, + "step": 2463 + }, + { + "epoch": 1.7521777777777778, + "grad_norm": 1.5787327289581299, + "learning_rate": 8.8110662789926e-06, + "loss": 0.2967, + "step": 2464 + }, + { + "epoch": 1.752888888888889, + "grad_norm": 1.3125972747802734, + "learning_rate": 8.802849328236418e-06, + "loss": 0.2907, + "step": 2465 + }, + { + "epoch": 1.7536, + "grad_norm": 1.5768535137176514, + "learning_rate": 8.79463319744677e-06, + "loss": 0.2628, + "step": 2466 + }, + { + "epoch": 1.7543111111111112, + "grad_norm": 1.7689968347549438, + "learning_rate": 8.786417892251143e-06, + "loss": 0.3375, + "step": 2467 + }, + { + "epoch": 1.7550222222222223, + "grad_norm": 1.4944849014282227, + "learning_rate": 8.778203418276463e-06, + "loss": 0.3187, + "step": 2468 + }, + { + "epoch": 1.7557333333333334, + "grad_norm": 1.130882978439331, + "learning_rate": 8.769989781149077e-06, + "loss": 0.2962, + "step": 2469 + }, + { + "epoch": 1.7564444444444445, + "grad_norm": 1.7000775337219238, + "learning_rate": 8.761776986494778e-06, + "loss": 0.2988, + "step": 2470 + }, + { + "epoch": 1.7571555555555556, + "grad_norm": 1.1217436790466309, + "learning_rate": 8.75356503993876e-06, + "loss": 0.2468, + "step": 2471 + }, + { + "epoch": 1.7578666666666667, + "grad_norm": 1.2479653358459473, + "learning_rate": 8.745353947105649e-06, + "loss": 0.2673, + "step": 2472 + }, + { + "epoch": 1.7585777777777778, + "grad_norm": 1.5145286321640015, + "learning_rate": 8.737143713619482e-06, + "loss": 0.3561, + "step": 2473 + }, + { + "epoch": 1.759288888888889, + "grad_norm": 1.6627912521362305, + "learning_rate": 8.728934345103715e-06, + "loss": 0.3232, + "step": 2474 + }, + { + "epoch": 1.76, + "grad_norm": 1.3916614055633545, + "learning_rate": 8.720725847181198e-06, + "loss": 0.2988, + "step": 2475 + }, + { + "epoch": 1.7607111111111111, + "grad_norm": 1.4626548290252686, + "learning_rate": 8.712518225474191e-06, + "loss": 0.3005, + "step": 2476 + }, + { + "epoch": 1.7614222222222222, + "grad_norm": 1.141692042350769, + "learning_rate": 8.704311485604366e-06, + "loss": 0.2893, + "step": 2477 + }, + { + "epoch": 1.7621333333333333, + "grad_norm": 1.0191608667373657, + "learning_rate": 8.696105633192766e-06, + "loss": 0.2692, + "step": 2478 + }, + { + "epoch": 1.7628444444444444, + "grad_norm": 0.849615216255188, + "learning_rate": 8.687900673859853e-06, + "loss": 0.2876, + "step": 2479 + }, + { + "epoch": 1.7635555555555555, + "grad_norm": 1.1361091136932373, + "learning_rate": 8.679696613225452e-06, + "loss": 0.351, + "step": 2480 + }, + { + "epoch": 1.7642666666666666, + "grad_norm": 1.1185874938964844, + "learning_rate": 8.671493456908797e-06, + "loss": 0.3208, + "step": 2481 + }, + { + "epoch": 1.7649777777777778, + "grad_norm": 0.9764256477355957, + "learning_rate": 8.663291210528481e-06, + "loss": 0.3197, + "step": 2482 + }, + { + "epoch": 1.7656888888888889, + "grad_norm": 1.0133628845214844, + "learning_rate": 8.655089879702489e-06, + "loss": 0.2664, + "step": 2483 + }, + { + "epoch": 1.7664, + "grad_norm": 1.1122949123382568, + "learning_rate": 8.646889470048166e-06, + "loss": 0.2714, + "step": 2484 + }, + { + "epoch": 1.767111111111111, + "grad_norm": 1.0336740016937256, + "learning_rate": 8.638689987182244e-06, + "loss": 0.3482, + "step": 2485 + }, + { + "epoch": 1.7678222222222222, + "grad_norm": 1.3224053382873535, + "learning_rate": 8.630491436720794e-06, + "loss": 0.2887, + "step": 2486 + }, + { + "epoch": 1.7685333333333333, + "grad_norm": 1.1404019594192505, + "learning_rate": 8.62229382427928e-06, + "loss": 0.3028, + "step": 2487 + }, + { + "epoch": 1.7692444444444444, + "grad_norm": 1.0537240505218506, + "learning_rate": 8.614097155472496e-06, + "loss": 0.3355, + "step": 2488 + }, + { + "epoch": 1.7699555555555555, + "grad_norm": 0.7367929816246033, + "learning_rate": 8.605901435914608e-06, + "loss": 0.2693, + "step": 2489 + }, + { + "epoch": 1.7706666666666666, + "grad_norm": 1.1835496425628662, + "learning_rate": 8.597706671219116e-06, + "loss": 0.2938, + "step": 2490 + }, + { + "epoch": 1.771377777777778, + "grad_norm": 1.0701507329940796, + "learning_rate": 8.589512866998885e-06, + "loss": 0.2879, + "step": 2491 + }, + { + "epoch": 1.7720888888888888, + "grad_norm": 1.3152763843536377, + "learning_rate": 8.581320028866105e-06, + "loss": 0.3417, + "step": 2492 + }, + { + "epoch": 1.7728000000000002, + "grad_norm": 1.286181926727295, + "learning_rate": 8.573128162432314e-06, + "loss": 0.2553, + "step": 2493 + }, + { + "epoch": 1.773511111111111, + "grad_norm": 1.0779309272766113, + "learning_rate": 8.564937273308382e-06, + "loss": 0.2875, + "step": 2494 + }, + { + "epoch": 1.7742222222222224, + "grad_norm": 0.9301726222038269, + "learning_rate": 8.556747367104504e-06, + "loss": 0.3017, + "step": 2495 + }, + { + "epoch": 1.7749333333333333, + "grad_norm": 0.9442057013511658, + "learning_rate": 8.548558449430217e-06, + "loss": 0.3252, + "step": 2496 + }, + { + "epoch": 1.7756444444444446, + "grad_norm": 0.9042646884918213, + "learning_rate": 8.54037052589436e-06, + "loss": 0.2984, + "step": 2497 + }, + { + "epoch": 1.7763555555555555, + "grad_norm": 0.9317138195037842, + "learning_rate": 8.53218360210511e-06, + "loss": 0.2751, + "step": 2498 + }, + { + "epoch": 1.7770666666666668, + "grad_norm": 1.0915993452072144, + "learning_rate": 8.523997683669945e-06, + "loss": 0.2983, + "step": 2499 + }, + { + "epoch": 1.7777777777777777, + "grad_norm": 1.038307547569275, + "learning_rate": 8.515812776195667e-06, + "loss": 0.3099, + "step": 2500 + }, + { + "epoch": 1.778488888888889, + "grad_norm": 1.0920332670211792, + "learning_rate": 8.507628885288372e-06, + "loss": 0.304, + "step": 2501 + }, + { + "epoch": 1.7792, + "grad_norm": 1.6293336153030396, + "learning_rate": 8.499446016553475e-06, + "loss": 0.3095, + "step": 2502 + }, + { + "epoch": 1.7799111111111112, + "grad_norm": 1.5048739910125732, + "learning_rate": 8.491264175595677e-06, + "loss": 0.288, + "step": 2503 + }, + { + "epoch": 1.780622222222222, + "grad_norm": 1.1640102863311768, + "learning_rate": 8.483083368018984e-06, + "loss": 0.3189, + "step": 2504 + }, + { + "epoch": 1.7813333333333334, + "grad_norm": 1.1616272926330566, + "learning_rate": 8.474903599426686e-06, + "loss": 0.2899, + "step": 2505 + }, + { + "epoch": 1.7820444444444443, + "grad_norm": 1.1557422876358032, + "learning_rate": 8.466724875421374e-06, + "loss": 0.3034, + "step": 2506 + }, + { + "epoch": 1.7827555555555556, + "grad_norm": 1.143441081047058, + "learning_rate": 8.458547201604906e-06, + "loss": 0.3183, + "step": 2507 + }, + { + "epoch": 1.7834666666666665, + "grad_norm": 1.076912522315979, + "learning_rate": 8.450370583578444e-06, + "loss": 0.3406, + "step": 2508 + }, + { + "epoch": 1.7841777777777779, + "grad_norm": 0.87898188829422, + "learning_rate": 8.442195026942398e-06, + "loss": 0.2639, + "step": 2509 + }, + { + "epoch": 1.7848888888888887, + "grad_norm": 0.9954450726509094, + "learning_rate": 8.434020537296477e-06, + "loss": 0.3147, + "step": 2510 + }, + { + "epoch": 1.7856, + "grad_norm": 0.8699561953544617, + "learning_rate": 8.425847120239649e-06, + "loss": 0.2633, + "step": 2511 + }, + { + "epoch": 1.786311111111111, + "grad_norm": 1.1255155801773071, + "learning_rate": 8.417674781370143e-06, + "loss": 0.3036, + "step": 2512 + }, + { + "epoch": 1.7870222222222223, + "grad_norm": 1.2179157733917236, + "learning_rate": 8.409503526285456e-06, + "loss": 0.2755, + "step": 2513 + }, + { + "epoch": 1.7877333333333332, + "grad_norm": 0.9332045912742615, + "learning_rate": 8.401333360582336e-06, + "loss": 0.2913, + "step": 2514 + }, + { + "epoch": 1.7884444444444445, + "grad_norm": 1.0457414388656616, + "learning_rate": 8.393164289856797e-06, + "loss": 0.2923, + "step": 2515 + }, + { + "epoch": 1.7891555555555556, + "grad_norm": 1.020363688468933, + "learning_rate": 8.384996319704084e-06, + "loss": 0.2883, + "step": 2516 + }, + { + "epoch": 1.7898666666666667, + "grad_norm": 0.8262305855751038, + "learning_rate": 8.37682945571871e-06, + "loss": 0.3058, + "step": 2517 + }, + { + "epoch": 1.7905777777777778, + "grad_norm": 0.8562010526657104, + "learning_rate": 8.36866370349441e-06, + "loss": 0.2501, + "step": 2518 + }, + { + "epoch": 1.791288888888889, + "grad_norm": 0.916185200214386, + "learning_rate": 8.36049906862418e-06, + "loss": 0.2943, + "step": 2519 + }, + { + "epoch": 1.792, + "grad_norm": 0.9067898392677307, + "learning_rate": 8.352335556700221e-06, + "loss": 0.2841, + "step": 2520 + }, + { + "epoch": 1.7927111111111111, + "grad_norm": 1.052657127380371, + "learning_rate": 8.344173173313998e-06, + "loss": 0.2991, + "step": 2521 + }, + { + "epoch": 1.7934222222222223, + "grad_norm": 0.9242912530899048, + "learning_rate": 8.336011924056175e-06, + "loss": 0.2818, + "step": 2522 + }, + { + "epoch": 1.7941333333333334, + "grad_norm": 1.0334945917129517, + "learning_rate": 8.32785181451666e-06, + "loss": 0.2902, + "step": 2523 + }, + { + "epoch": 1.7948444444444445, + "grad_norm": 1.1962789297103882, + "learning_rate": 8.319692850284564e-06, + "loss": 0.337, + "step": 2524 + }, + { + "epoch": 1.7955555555555556, + "grad_norm": 0.8610202074050903, + "learning_rate": 8.311535036948223e-06, + "loss": 0.2958, + "step": 2525 + }, + { + "epoch": 1.7962666666666667, + "grad_norm": 0.736598014831543, + "learning_rate": 8.303378380095184e-06, + "loss": 0.2856, + "step": 2526 + }, + { + "epoch": 1.7969777777777778, + "grad_norm": 1.0694541931152344, + "learning_rate": 8.2952228853122e-06, + "loss": 0.2827, + "step": 2527 + }, + { + "epoch": 1.797688888888889, + "grad_norm": 0.8858334422111511, + "learning_rate": 8.287068558185225e-06, + "loss": 0.2987, + "step": 2528 + }, + { + "epoch": 1.7984, + "grad_norm": 1.3889973163604736, + "learning_rate": 8.27891540429942e-06, + "loss": 0.3022, + "step": 2529 + }, + { + "epoch": 1.799111111111111, + "grad_norm": 0.873847484588623, + "learning_rate": 8.270763429239144e-06, + "loss": 0.3009, + "step": 2530 + }, + { + "epoch": 1.7998222222222222, + "grad_norm": 0.9429922103881836, + "learning_rate": 8.262612638587933e-06, + "loss": 0.2925, + "step": 2531 + }, + { + "epoch": 1.8005333333333333, + "grad_norm": 0.7871550917625427, + "learning_rate": 8.254463037928534e-06, + "loss": 0.2979, + "step": 2532 + }, + { + "epoch": 1.8012444444444444, + "grad_norm": 0.8696495890617371, + "learning_rate": 8.246314632842856e-06, + "loss": 0.3125, + "step": 2533 + }, + { + "epoch": 1.8019555555555555, + "grad_norm": 0.8911949396133423, + "learning_rate": 8.238167428912009e-06, + "loss": 0.3035, + "step": 2534 + }, + { + "epoch": 1.8026666666666666, + "grad_norm": 0.9185874462127686, + "learning_rate": 8.23002143171627e-06, + "loss": 0.2733, + "step": 2535 + }, + { + "epoch": 1.8033777777777777, + "grad_norm": 0.9980380535125732, + "learning_rate": 8.22187664683509e-06, + "loss": 0.286, + "step": 2536 + }, + { + "epoch": 1.8040888888888889, + "grad_norm": 1.0439165830612183, + "learning_rate": 8.213733079847086e-06, + "loss": 0.3107, + "step": 2537 + }, + { + "epoch": 1.8048, + "grad_norm": 1.089730143547058, + "learning_rate": 8.205590736330058e-06, + "loss": 0.2891, + "step": 2538 + }, + { + "epoch": 1.805511111111111, + "grad_norm": 1.0048503875732422, + "learning_rate": 8.197449621860944e-06, + "loss": 0.3118, + "step": 2539 + }, + { + "epoch": 1.8062222222222222, + "grad_norm": 1.0420483350753784, + "learning_rate": 8.18930974201586e-06, + "loss": 0.3072, + "step": 2540 + }, + { + "epoch": 1.8069333333333333, + "grad_norm": 0.9612111449241638, + "learning_rate": 8.181171102370062e-06, + "loss": 0.2639, + "step": 2541 + }, + { + "epoch": 1.8076444444444446, + "grad_norm": 1.0710488557815552, + "learning_rate": 8.173033708497968e-06, + "loss": 0.3436, + "step": 2542 + }, + { + "epoch": 1.8083555555555555, + "grad_norm": 1.1218905448913574, + "learning_rate": 8.164897565973133e-06, + "loss": 0.3345, + "step": 2543 + }, + { + "epoch": 1.8090666666666668, + "grad_norm": 0.9367212653160095, + "learning_rate": 8.156762680368267e-06, + "loss": 0.296, + "step": 2544 + }, + { + "epoch": 1.8097777777777777, + "grad_norm": 0.9909266233444214, + "learning_rate": 8.148629057255203e-06, + "loss": 0.3067, + "step": 2545 + }, + { + "epoch": 1.810488888888889, + "grad_norm": 0.7594852447509766, + "learning_rate": 8.140496702204921e-06, + "loss": 0.2635, + "step": 2546 + }, + { + "epoch": 1.8112, + "grad_norm": 0.8196940422058105, + "learning_rate": 8.132365620787533e-06, + "loss": 0.2772, + "step": 2547 + }, + { + "epoch": 1.8119111111111112, + "grad_norm": 1.0279898643493652, + "learning_rate": 8.124235818572268e-06, + "loss": 0.2656, + "step": 2548 + }, + { + "epoch": 1.8126222222222221, + "grad_norm": 0.9127392768859863, + "learning_rate": 8.116107301127495e-06, + "loss": 0.3196, + "step": 2549 + }, + { + "epoch": 1.8133333333333335, + "grad_norm": 1.1290074586868286, + "learning_rate": 8.107980074020684e-06, + "loss": 0.3119, + "step": 2550 + }, + { + "epoch": 1.8140444444444443, + "grad_norm": 0.8611456155776978, + "learning_rate": 8.09985414281844e-06, + "loss": 0.284, + "step": 2551 + }, + { + "epoch": 1.8147555555555557, + "grad_norm": 1.0375441312789917, + "learning_rate": 8.091729513086462e-06, + "loss": 0.3311, + "step": 2552 + }, + { + "epoch": 1.8154666666666666, + "grad_norm": 0.8402802348136902, + "learning_rate": 8.083606190389579e-06, + "loss": 0.3261, + "step": 2553 + }, + { + "epoch": 1.816177777777778, + "grad_norm": 0.9757222533226013, + "learning_rate": 8.075484180291702e-06, + "loss": 0.3256, + "step": 2554 + }, + { + "epoch": 1.8168888888888888, + "grad_norm": 0.9266077876091003, + "learning_rate": 8.067363488355859e-06, + "loss": 0.326, + "step": 2555 + }, + { + "epoch": 1.8176, + "grad_norm": 1.0077120065689087, + "learning_rate": 8.059244120144167e-06, + "loss": 0.273, + "step": 2556 + }, + { + "epoch": 1.818311111111111, + "grad_norm": 1.115109920501709, + "learning_rate": 8.051126081217847e-06, + "loss": 0.29, + "step": 2557 + }, + { + "epoch": 1.8190222222222223, + "grad_norm": 1.077256202697754, + "learning_rate": 8.043009377137188e-06, + "loss": 0.3075, + "step": 2558 + }, + { + "epoch": 1.8197333333333332, + "grad_norm": 1.1593800783157349, + "learning_rate": 8.034894013461592e-06, + "loss": 0.2909, + "step": 2559 + }, + { + "epoch": 1.8204444444444445, + "grad_norm": 1.0799871683120728, + "learning_rate": 8.026779995749519e-06, + "loss": 0.2983, + "step": 2560 + }, + { + "epoch": 1.8211555555555554, + "grad_norm": 1.2153723239898682, + "learning_rate": 8.018667329558527e-06, + "loss": 0.2849, + "step": 2561 + }, + { + "epoch": 1.8218666666666667, + "grad_norm": 1.1019790172576904, + "learning_rate": 8.010556020445231e-06, + "loss": 0.304, + "step": 2562 + }, + { + "epoch": 1.8225777777777776, + "grad_norm": 1.0194923877716064, + "learning_rate": 8.002446073965325e-06, + "loss": 0.297, + "step": 2563 + }, + { + "epoch": 1.823288888888889, + "grad_norm": 1.6127690076828003, + "learning_rate": 7.99433749567358e-06, + "loss": 0.3343, + "step": 2564 + }, + { + "epoch": 1.8239999999999998, + "grad_norm": 1.566721796989441, + "learning_rate": 7.986230291123805e-06, + "loss": 0.2806, + "step": 2565 + }, + { + "epoch": 1.8247111111111112, + "grad_norm": 1.0992478132247925, + "learning_rate": 7.97812446586889e-06, + "loss": 0.313, + "step": 2566 + }, + { + "epoch": 1.825422222222222, + "grad_norm": 1.4223458766937256, + "learning_rate": 7.970020025460765e-06, + "loss": 0.2804, + "step": 2567 + }, + { + "epoch": 1.8261333333333334, + "grad_norm": 1.3181889057159424, + "learning_rate": 7.96191697545043e-06, + "loss": 0.2624, + "step": 2568 + }, + { + "epoch": 1.8268444444444445, + "grad_norm": 1.480263352394104, + "learning_rate": 7.953815321387908e-06, + "loss": 0.2961, + "step": 2569 + }, + { + "epoch": 1.8275555555555556, + "grad_norm": 1.1141407489776611, + "learning_rate": 7.945715068822291e-06, + "loss": 0.3112, + "step": 2570 + }, + { + "epoch": 1.8282666666666667, + "grad_norm": 1.4093608856201172, + "learning_rate": 7.937616223301692e-06, + "loss": 0.2752, + "step": 2571 + }, + { + "epoch": 1.8289777777777778, + "grad_norm": 1.2492815256118774, + "learning_rate": 7.929518790373274e-06, + "loss": 0.296, + "step": 2572 + }, + { + "epoch": 1.829688888888889, + "grad_norm": 1.0210415124893188, + "learning_rate": 7.921422775583218e-06, + "loss": 0.3055, + "step": 2573 + }, + { + "epoch": 1.8304, + "grad_norm": 1.0024794340133667, + "learning_rate": 7.913328184476748e-06, + "loss": 0.2858, + "step": 2574 + }, + { + "epoch": 1.8311111111111111, + "grad_norm": 1.078639030456543, + "learning_rate": 7.905235022598103e-06, + "loss": 0.2892, + "step": 2575 + }, + { + "epoch": 1.8318222222222222, + "grad_norm": 0.9358330368995667, + "learning_rate": 7.897143295490551e-06, + "loss": 0.3115, + "step": 2576 + }, + { + "epoch": 1.8325333333333333, + "grad_norm": 1.256665587425232, + "learning_rate": 7.88905300869637e-06, + "loss": 0.3277, + "step": 2577 + }, + { + "epoch": 1.8332444444444445, + "grad_norm": 0.9349859952926636, + "learning_rate": 7.880964167756855e-06, + "loss": 0.3185, + "step": 2578 + }, + { + "epoch": 1.8339555555555556, + "grad_norm": 0.9094565510749817, + "learning_rate": 7.872876778212308e-06, + "loss": 0.2866, + "step": 2579 + }, + { + "epoch": 1.8346666666666667, + "grad_norm": 1.216451644897461, + "learning_rate": 7.86479084560204e-06, + "loss": 0.3097, + "step": 2580 + }, + { + "epoch": 1.8353777777777778, + "grad_norm": 1.3017231225967407, + "learning_rate": 7.856706375464368e-06, + "loss": 0.3327, + "step": 2581 + }, + { + "epoch": 1.8360888888888889, + "grad_norm": 0.8567536473274231, + "learning_rate": 7.848623373336594e-06, + "loss": 0.2803, + "step": 2582 + }, + { + "epoch": 1.8368, + "grad_norm": 1.2454681396484375, + "learning_rate": 7.840541844755033e-06, + "loss": 0.2767, + "step": 2583 + }, + { + "epoch": 1.837511111111111, + "grad_norm": 1.3123332262039185, + "learning_rate": 7.83246179525497e-06, + "loss": 0.3342, + "step": 2584 + }, + { + "epoch": 1.8382222222222222, + "grad_norm": 0.891075849533081, + "learning_rate": 7.824383230370698e-06, + "loss": 0.2879, + "step": 2585 + }, + { + "epoch": 1.8389333333333333, + "grad_norm": 0.9105257391929626, + "learning_rate": 7.816306155635475e-06, + "loss": 0.2581, + "step": 2586 + }, + { + "epoch": 1.8396444444444444, + "grad_norm": 1.0493334531784058, + "learning_rate": 7.808230576581552e-06, + "loss": 0.326, + "step": 2587 + }, + { + "epoch": 1.8403555555555555, + "grad_norm": 1.0923950672149658, + "learning_rate": 7.800156498740143e-06, + "loss": 0.2827, + "step": 2588 + }, + { + "epoch": 1.8410666666666666, + "grad_norm": 0.8727123737335205, + "learning_rate": 7.792083927641447e-06, + "loss": 0.2933, + "step": 2589 + }, + { + "epoch": 1.8417777777777777, + "grad_norm": 1.1724317073822021, + "learning_rate": 7.784012868814618e-06, + "loss": 0.251, + "step": 2590 + }, + { + "epoch": 1.8424888888888888, + "grad_norm": 0.9697776436805725, + "learning_rate": 7.775943327787789e-06, + "loss": 0.2712, + "step": 2591 + }, + { + "epoch": 1.8432, + "grad_norm": 0.9038233757019043, + "learning_rate": 7.767875310088034e-06, + "loss": 0.3126, + "step": 2592 + }, + { + "epoch": 1.843911111111111, + "grad_norm": 1.128234624862671, + "learning_rate": 7.759808821241406e-06, + "loss": 0.3093, + "step": 2593 + }, + { + "epoch": 1.8446222222222222, + "grad_norm": 1.0924396514892578, + "learning_rate": 7.751743866772889e-06, + "loss": 0.2731, + "step": 2594 + }, + { + "epoch": 1.8453333333333335, + "grad_norm": 0.8447387218475342, + "learning_rate": 7.743680452206439e-06, + "loss": 0.2788, + "step": 2595 + }, + { + "epoch": 1.8460444444444444, + "grad_norm": 1.416035532951355, + "learning_rate": 7.735618583064931e-06, + "loss": 0.2785, + "step": 2596 + }, + { + "epoch": 1.8467555555555557, + "grad_norm": 1.6071412563323975, + "learning_rate": 7.727558264870205e-06, + "loss": 0.3291, + "step": 2597 + }, + { + "epoch": 1.8474666666666666, + "grad_norm": 0.920295000076294, + "learning_rate": 7.719499503143027e-06, + "loss": 0.312, + "step": 2598 + }, + { + "epoch": 1.848177777777778, + "grad_norm": 1.1171166896820068, + "learning_rate": 7.711442303403096e-06, + "loss": 0.3206, + "step": 2599 + }, + { + "epoch": 1.8488888888888888, + "grad_norm": 1.547060489654541, + "learning_rate": 7.70338667116905e-06, + "loss": 0.2807, + "step": 2600 + }, + { + "epoch": 1.8496000000000001, + "grad_norm": 1.3143491744995117, + "learning_rate": 7.695332611958438e-06, + "loss": 0.3016, + "step": 2601 + }, + { + "epoch": 1.850311111111111, + "grad_norm": 1.102298378944397, + "learning_rate": 7.68728013128775e-06, + "loss": 0.2967, + "step": 2602 + }, + { + "epoch": 1.8510222222222223, + "grad_norm": 1.5591912269592285, + "learning_rate": 7.67922923467238e-06, + "loss": 0.2968, + "step": 2603 + }, + { + "epoch": 1.8517333333333332, + "grad_norm": 1.603651762008667, + "learning_rate": 7.671179927626651e-06, + "loss": 0.2904, + "step": 2604 + }, + { + "epoch": 1.8524444444444446, + "grad_norm": 0.8563565611839294, + "learning_rate": 7.663132215663777e-06, + "loss": 0.2974, + "step": 2605 + }, + { + "epoch": 1.8531555555555554, + "grad_norm": 1.2999849319458008, + "learning_rate": 7.655086104295904e-06, + "loss": 0.3261, + "step": 2606 + }, + { + "epoch": 1.8538666666666668, + "grad_norm": 1.2167513370513916, + "learning_rate": 7.64704159903406e-06, + "loss": 0.334, + "step": 2607 + }, + { + "epoch": 1.8545777777777777, + "grad_norm": 1.0433200597763062, + "learning_rate": 7.638998705388188e-06, + "loss": 0.2856, + "step": 2608 + }, + { + "epoch": 1.855288888888889, + "grad_norm": 1.151815414428711, + "learning_rate": 7.630957428867117e-06, + "loss": 0.2754, + "step": 2609 + }, + { + "epoch": 1.8559999999999999, + "grad_norm": 1.5194209814071655, + "learning_rate": 7.622917774978579e-06, + "loss": 0.3192, + "step": 2610 + }, + { + "epoch": 1.8567111111111112, + "grad_norm": 1.3469945192337036, + "learning_rate": 7.614879749229178e-06, + "loss": 0.351, + "step": 2611 + }, + { + "epoch": 1.857422222222222, + "grad_norm": 1.3624012470245361, + "learning_rate": 7.606843357124426e-06, + "loss": 0.3079, + "step": 2612 + }, + { + "epoch": 1.8581333333333334, + "grad_norm": 1.3058192729949951, + "learning_rate": 7.598808604168691e-06, + "loss": 0.2848, + "step": 2613 + }, + { + "epoch": 1.8588444444444443, + "grad_norm": 1.0508201122283936, + "learning_rate": 7.5907754958652365e-06, + "loss": 0.2845, + "step": 2614 + }, + { + "epoch": 1.8595555555555556, + "grad_norm": 1.1531128883361816, + "learning_rate": 7.582744037716198e-06, + "loss": 0.2954, + "step": 2615 + }, + { + "epoch": 1.8602666666666665, + "grad_norm": 1.1618566513061523, + "learning_rate": 7.574714235222571e-06, + "loss": 0.2761, + "step": 2616 + }, + { + "epoch": 1.8609777777777778, + "grad_norm": 0.8415836691856384, + "learning_rate": 7.566686093884224e-06, + "loss": 0.3125, + "step": 2617 + }, + { + "epoch": 1.8616888888888887, + "grad_norm": 1.001941204071045, + "learning_rate": 7.558659619199884e-06, + "loss": 0.2664, + "step": 2618 + }, + { + "epoch": 1.8624, + "grad_norm": 1.1795772314071655, + "learning_rate": 7.550634816667142e-06, + "loss": 0.2903, + "step": 2619 + }, + { + "epoch": 1.8631111111111112, + "grad_norm": 0.8030235171318054, + "learning_rate": 7.542611691782433e-06, + "loss": 0.282, + "step": 2620 + }, + { + "epoch": 1.8638222222222223, + "grad_norm": 1.0577133893966675, + "learning_rate": 7.5345902500410614e-06, + "loss": 0.3067, + "step": 2621 + }, + { + "epoch": 1.8645333333333334, + "grad_norm": 0.9377948045730591, + "learning_rate": 7.526570496937157e-06, + "loss": 0.2602, + "step": 2622 + }, + { + "epoch": 1.8652444444444445, + "grad_norm": 0.773709237575531, + "learning_rate": 7.5185524379637105e-06, + "loss": 0.265, + "step": 2623 + }, + { + "epoch": 1.8659555555555556, + "grad_norm": 1.274538516998291, + "learning_rate": 7.5105360786125405e-06, + "loss": 0.2817, + "step": 2624 + }, + { + "epoch": 1.8666666666666667, + "grad_norm": 1.0782482624053955, + "learning_rate": 7.502521424374312e-06, + "loss": 0.2795, + "step": 2625 + }, + { + "epoch": 1.8673777777777778, + "grad_norm": 0.8996921181678772, + "learning_rate": 7.494508480738508e-06, + "loss": 0.3243, + "step": 2626 + }, + { + "epoch": 1.868088888888889, + "grad_norm": 1.2805826663970947, + "learning_rate": 7.486497253193458e-06, + "loss": 0.2834, + "step": 2627 + }, + { + "epoch": 1.8688, + "grad_norm": 0.8913900852203369, + "learning_rate": 7.4784877472262994e-06, + "loss": 0.2871, + "step": 2628 + }, + { + "epoch": 1.8695111111111111, + "grad_norm": 0.8884389996528625, + "learning_rate": 7.470479968322999e-06, + "loss": 0.2806, + "step": 2629 + }, + { + "epoch": 1.8702222222222222, + "grad_norm": 1.113813042640686, + "learning_rate": 7.462473921968338e-06, + "loss": 0.3255, + "step": 2630 + }, + { + "epoch": 1.8709333333333333, + "grad_norm": 0.8476263880729675, + "learning_rate": 7.454469613645912e-06, + "loss": 0.3251, + "step": 2631 + }, + { + "epoch": 1.8716444444444444, + "grad_norm": 0.8127641677856445, + "learning_rate": 7.446467048838131e-06, + "loss": 0.3046, + "step": 2632 + }, + { + "epoch": 1.8723555555555556, + "grad_norm": 1.1676257848739624, + "learning_rate": 7.438466233026198e-06, + "loss": 0.2753, + "step": 2633 + }, + { + "epoch": 1.8730666666666667, + "grad_norm": 0.8329217433929443, + "learning_rate": 7.430467171690134e-06, + "loss": 0.3158, + "step": 2634 + }, + { + "epoch": 1.8737777777777778, + "grad_norm": 1.330282211303711, + "learning_rate": 7.422469870308744e-06, + "loss": 0.3679, + "step": 2635 + }, + { + "epoch": 1.8744888888888889, + "grad_norm": 1.174935221672058, + "learning_rate": 7.4144743343596385e-06, + "loss": 0.2811, + "step": 2636 + }, + { + "epoch": 1.8752, + "grad_norm": 0.9553257822990417, + "learning_rate": 7.40648056931921e-06, + "loss": 0.3135, + "step": 2637 + }, + { + "epoch": 1.875911111111111, + "grad_norm": 1.4676322937011719, + "learning_rate": 7.398488580662644e-06, + "loss": 0.323, + "step": 2638 + }, + { + "epoch": 1.8766222222222222, + "grad_norm": 1.2959723472595215, + "learning_rate": 7.390498373863908e-06, + "loss": 0.2688, + "step": 2639 + }, + { + "epoch": 1.8773333333333333, + "grad_norm": 0.905664324760437, + "learning_rate": 7.382509954395749e-06, + "loss": 0.2712, + "step": 2640 + }, + { + "epoch": 1.8780444444444444, + "grad_norm": 1.2460429668426514, + "learning_rate": 7.374523327729685e-06, + "loss": 0.2705, + "step": 2641 + }, + { + "epoch": 1.8787555555555555, + "grad_norm": 1.084533929824829, + "learning_rate": 7.366538499336018e-06, + "loss": 0.2876, + "step": 2642 + }, + { + "epoch": 1.8794666666666666, + "grad_norm": 0.9086869955062866, + "learning_rate": 7.358555474683801e-06, + "loss": 0.3218, + "step": 2643 + }, + { + "epoch": 1.8801777777777777, + "grad_norm": 1.0477027893066406, + "learning_rate": 7.35057425924087e-06, + "loss": 0.3159, + "step": 2644 + }, + { + "epoch": 1.8808888888888888, + "grad_norm": 1.1287204027175903, + "learning_rate": 7.342594858473807e-06, + "loss": 0.286, + "step": 2645 + }, + { + "epoch": 1.8816000000000002, + "grad_norm": 0.9045525789260864, + "learning_rate": 7.334617277847963e-06, + "loss": 0.2583, + "step": 2646 + }, + { + "epoch": 1.882311111111111, + "grad_norm": 1.055514931678772, + "learning_rate": 7.326641522827429e-06, + "loss": 0.2843, + "step": 2647 + }, + { + "epoch": 1.8830222222222224, + "grad_norm": 1.299106478691101, + "learning_rate": 7.31866759887506e-06, + "loss": 0.2605, + "step": 2648 + }, + { + "epoch": 1.8837333333333333, + "grad_norm": 1.0591930150985718, + "learning_rate": 7.310695511452448e-06, + "loss": 0.281, + "step": 2649 + }, + { + "epoch": 1.8844444444444446, + "grad_norm": 1.056222677230835, + "learning_rate": 7.302725266019924e-06, + "loss": 0.3147, + "step": 2650 + }, + { + "epoch": 1.8851555555555555, + "grad_norm": 1.1425197124481201, + "learning_rate": 7.294756868036571e-06, + "loss": 0.2742, + "step": 2651 + }, + { + "epoch": 1.8858666666666668, + "grad_norm": 1.0755802392959595, + "learning_rate": 7.286790322960189e-06, + "loss": 0.3173, + "step": 2652 + }, + { + "epoch": 1.8865777777777777, + "grad_norm": 1.2457255125045776, + "learning_rate": 7.278825636247327e-06, + "loss": 0.2974, + "step": 2653 + }, + { + "epoch": 1.887288888888889, + "grad_norm": 1.376346230506897, + "learning_rate": 7.270862813353241e-06, + "loss": 0.2837, + "step": 2654 + }, + { + "epoch": 1.888, + "grad_norm": 1.096067190170288, + "learning_rate": 7.262901859731933e-06, + "loss": 0.3083, + "step": 2655 + }, + { + "epoch": 1.8887111111111112, + "grad_norm": 0.9530537128448486, + "learning_rate": 7.254942780836103e-06, + "loss": 0.2703, + "step": 2656 + }, + { + "epoch": 1.8894222222222221, + "grad_norm": 1.2835155725479126, + "learning_rate": 7.246985582117185e-06, + "loss": 0.2541, + "step": 2657 + }, + { + "epoch": 1.8901333333333334, + "grad_norm": 1.4047069549560547, + "learning_rate": 7.239030269025311e-06, + "loss": 0.3152, + "step": 2658 + }, + { + "epoch": 1.8908444444444443, + "grad_norm": 1.4884498119354248, + "learning_rate": 7.231076847009335e-06, + "loss": 0.3243, + "step": 2659 + }, + { + "epoch": 1.8915555555555557, + "grad_norm": 1.1223069429397583, + "learning_rate": 7.2231253215168e-06, + "loss": 0.255, + "step": 2660 + }, + { + "epoch": 1.8922666666666665, + "grad_norm": 1.3849148750305176, + "learning_rate": 7.2151756979939654e-06, + "loss": 0.313, + "step": 2661 + }, + { + "epoch": 1.8929777777777779, + "grad_norm": 1.2538187503814697, + "learning_rate": 7.2072279818857745e-06, + "loss": 0.2587, + "step": 2662 + }, + { + "epoch": 1.8936888888888888, + "grad_norm": 1.0633207559585571, + "learning_rate": 7.199282178635878e-06, + "loss": 0.304, + "step": 2663 + }, + { + "epoch": 1.8944, + "grad_norm": 1.4451664686203003, + "learning_rate": 7.191338293686601e-06, + "loss": 0.3014, + "step": 2664 + }, + { + "epoch": 1.895111111111111, + "grad_norm": 1.229330062866211, + "learning_rate": 7.183396332478971e-06, + "loss": 0.2894, + "step": 2665 + }, + { + "epoch": 1.8958222222222223, + "grad_norm": 0.8069056868553162, + "learning_rate": 7.175456300452681e-06, + "loss": 0.2724, + "step": 2666 + }, + { + "epoch": 1.8965333333333332, + "grad_norm": 1.3277325630187988, + "learning_rate": 7.1675182030461155e-06, + "loss": 0.2712, + "step": 2667 + }, + { + "epoch": 1.8972444444444445, + "grad_norm": 0.9606416821479797, + "learning_rate": 7.159582045696334e-06, + "loss": 0.3025, + "step": 2668 + }, + { + "epoch": 1.8979555555555554, + "grad_norm": 0.8697434067726135, + "learning_rate": 7.151647833839055e-06, + "loss": 0.2638, + "step": 2669 + }, + { + "epoch": 1.8986666666666667, + "grad_norm": 0.8371191620826721, + "learning_rate": 7.143715572908679e-06, + "loss": 0.2731, + "step": 2670 + }, + { + "epoch": 1.8993777777777778, + "grad_norm": 0.8524051308631897, + "learning_rate": 7.1357852683382565e-06, + "loss": 0.2866, + "step": 2671 + }, + { + "epoch": 1.900088888888889, + "grad_norm": 0.8527877926826477, + "learning_rate": 7.127856925559513e-06, + "loss": 0.2919, + "step": 2672 + }, + { + "epoch": 1.9008, + "grad_norm": 0.8782299757003784, + "learning_rate": 7.119930550002814e-06, + "loss": 0.2914, + "step": 2673 + }, + { + "epoch": 1.9015111111111112, + "grad_norm": 0.9655768871307373, + "learning_rate": 7.112006147097195e-06, + "loss": 0.2724, + "step": 2674 + }, + { + "epoch": 1.9022222222222223, + "grad_norm": 0.9094228148460388, + "learning_rate": 7.104083722270322e-06, + "loss": 0.2371, + "step": 2675 + }, + { + "epoch": 1.9029333333333334, + "grad_norm": 0.9312955737113953, + "learning_rate": 7.096163280948523e-06, + "loss": 0.2427, + "step": 2676 + }, + { + "epoch": 1.9036444444444445, + "grad_norm": 0.9036617279052734, + "learning_rate": 7.088244828556754e-06, + "loss": 0.2913, + "step": 2677 + }, + { + "epoch": 1.9043555555555556, + "grad_norm": 0.8126955628395081, + "learning_rate": 7.080328370518623e-06, + "loss": 0.2725, + "step": 2678 + }, + { + "epoch": 1.9050666666666667, + "grad_norm": 0.8704136610031128, + "learning_rate": 7.072413912256353e-06, + "loss": 0.2459, + "step": 2679 + }, + { + "epoch": 1.9057777777777778, + "grad_norm": 0.8371139764785767, + "learning_rate": 7.064501459190816e-06, + "loss": 0.3017, + "step": 2680 + }, + { + "epoch": 1.906488888888889, + "grad_norm": 1.0437263250350952, + "learning_rate": 7.056591016741501e-06, + "loss": 0.2833, + "step": 2681 + }, + { + "epoch": 1.9072, + "grad_norm": 1.0794475078582764, + "learning_rate": 7.048682590326519e-06, + "loss": 0.3207, + "step": 2682 + }, + { + "epoch": 1.9079111111111111, + "grad_norm": 0.9429819583892822, + "learning_rate": 7.040776185362601e-06, + "loss": 0.3154, + "step": 2683 + }, + { + "epoch": 1.9086222222222222, + "grad_norm": 1.1637533903121948, + "learning_rate": 7.032871807265097e-06, + "loss": 0.2932, + "step": 2684 + }, + { + "epoch": 1.9093333333333333, + "grad_norm": 0.8366556167602539, + "learning_rate": 7.024969461447973e-06, + "loss": 0.2643, + "step": 2685 + }, + { + "epoch": 1.9100444444444444, + "grad_norm": 0.9259528517723083, + "learning_rate": 7.017069153323785e-06, + "loss": 0.2542, + "step": 2686 + }, + { + "epoch": 1.9107555555555555, + "grad_norm": 0.8660364151000977, + "learning_rate": 7.0091708883037155e-06, + "loss": 0.3059, + "step": 2687 + }, + { + "epoch": 1.9114666666666666, + "grad_norm": 0.9208256006240845, + "learning_rate": 7.0012746717975284e-06, + "loss": 0.3402, + "step": 2688 + }, + { + "epoch": 1.9121777777777778, + "grad_norm": 0.8375102877616882, + "learning_rate": 6.993380509213603e-06, + "loss": 0.2795, + "step": 2689 + }, + { + "epoch": 1.9128888888888889, + "grad_norm": 0.8787679672241211, + "learning_rate": 6.9854884059588934e-06, + "loss": 0.2787, + "step": 2690 + }, + { + "epoch": 1.9136, + "grad_norm": 0.7850491404533386, + "learning_rate": 6.977598367438956e-06, + "loss": 0.2837, + "step": 2691 + }, + { + "epoch": 1.914311111111111, + "grad_norm": 0.7378684282302856, + "learning_rate": 6.969710399057927e-06, + "loss": 0.2731, + "step": 2692 + }, + { + "epoch": 1.9150222222222222, + "grad_norm": 0.9139169454574585, + "learning_rate": 6.961824506218531e-06, + "loss": 0.3059, + "step": 2693 + }, + { + "epoch": 1.9157333333333333, + "grad_norm": 0.8881707191467285, + "learning_rate": 6.953940694322059e-06, + "loss": 0.2929, + "step": 2694 + }, + { + "epoch": 1.9164444444444444, + "grad_norm": 0.8410139679908752, + "learning_rate": 6.946058968768392e-06, + "loss": 0.2715, + "step": 2695 + }, + { + "epoch": 1.9171555555555555, + "grad_norm": 0.866623044013977, + "learning_rate": 6.938179334955967e-06, + "loss": 0.2834, + "step": 2696 + }, + { + "epoch": 1.9178666666666668, + "grad_norm": 0.7775970697402954, + "learning_rate": 6.930301798281803e-06, + "loss": 0.2763, + "step": 2697 + }, + { + "epoch": 1.9185777777777777, + "grad_norm": 0.9200167655944824, + "learning_rate": 6.9224263641414675e-06, + "loss": 0.2907, + "step": 2698 + }, + { + "epoch": 1.919288888888889, + "grad_norm": 0.8392654061317444, + "learning_rate": 6.914553037929103e-06, + "loss": 0.2531, + "step": 2699 + }, + { + "epoch": 1.92, + "grad_norm": 0.94516521692276, + "learning_rate": 6.906681825037395e-06, + "loss": 0.2668, + "step": 2700 + }, + { + "epoch": 1.9207111111111113, + "grad_norm": 1.113713026046753, + "learning_rate": 6.898812730857589e-06, + "loss": 0.3026, + "step": 2701 + }, + { + "epoch": 1.9214222222222221, + "grad_norm": 0.9510433673858643, + "learning_rate": 6.8909457607794824e-06, + "loss": 0.3275, + "step": 2702 + }, + { + "epoch": 1.9221333333333335, + "grad_norm": 0.7645305395126343, + "learning_rate": 6.883080920191404e-06, + "loss": 0.2424, + "step": 2703 + }, + { + "epoch": 1.9228444444444444, + "grad_norm": 0.8973283171653748, + "learning_rate": 6.8752182144802415e-06, + "loss": 0.3068, + "step": 2704 + }, + { + "epoch": 1.9235555555555557, + "grad_norm": 0.9254751205444336, + "learning_rate": 6.867357649031405e-06, + "loss": 0.3021, + "step": 2705 + }, + { + "epoch": 1.9242666666666666, + "grad_norm": 0.8371214270591736, + "learning_rate": 6.859499229228852e-06, + "loss": 0.2808, + "step": 2706 + }, + { + "epoch": 1.924977777777778, + "grad_norm": 0.8484550714492798, + "learning_rate": 6.851642960455057e-06, + "loss": 0.2934, + "step": 2707 + }, + { + "epoch": 1.9256888888888888, + "grad_norm": 0.8045176267623901, + "learning_rate": 6.8437888480910355e-06, + "loss": 0.2879, + "step": 2708 + }, + { + "epoch": 1.9264000000000001, + "grad_norm": 0.9100943207740784, + "learning_rate": 6.835936897516311e-06, + "loss": 0.2855, + "step": 2709 + }, + { + "epoch": 1.927111111111111, + "grad_norm": 0.8397871255874634, + "learning_rate": 6.8280871141089415e-06, + "loss": 0.2968, + "step": 2710 + }, + { + "epoch": 1.9278222222222223, + "grad_norm": 0.7852779030799866, + "learning_rate": 6.820239503245487e-06, + "loss": 0.2767, + "step": 2711 + }, + { + "epoch": 1.9285333333333332, + "grad_norm": 0.855201005935669, + "learning_rate": 6.81239407030103e-06, + "loss": 0.2798, + "step": 2712 + }, + { + "epoch": 1.9292444444444445, + "grad_norm": 1.0833855867385864, + "learning_rate": 6.804550820649152e-06, + "loss": 0.2552, + "step": 2713 + }, + { + "epoch": 1.9299555555555554, + "grad_norm": 0.7380629181861877, + "learning_rate": 6.7967097596619495e-06, + "loss": 0.3015, + "step": 2714 + }, + { + "epoch": 1.9306666666666668, + "grad_norm": 0.931531548500061, + "learning_rate": 6.788870892710007e-06, + "loss": 0.2776, + "step": 2715 + }, + { + "epoch": 1.9313777777777776, + "grad_norm": 0.8180390000343323, + "learning_rate": 6.781034225162422e-06, + "loss": 0.2443, + "step": 2716 + }, + { + "epoch": 1.932088888888889, + "grad_norm": 0.76549232006073, + "learning_rate": 6.773199762386768e-06, + "loss": 0.2689, + "step": 2717 + }, + { + "epoch": 1.9327999999999999, + "grad_norm": 0.9245445132255554, + "learning_rate": 6.765367509749123e-06, + "loss": 0.2678, + "step": 2718 + }, + { + "epoch": 1.9335111111111112, + "grad_norm": 0.7454220652580261, + "learning_rate": 6.757537472614051e-06, + "loss": 0.2494, + "step": 2719 + }, + { + "epoch": 1.934222222222222, + "grad_norm": 0.9364449977874756, + "learning_rate": 6.749709656344584e-06, + "loss": 0.3108, + "step": 2720 + }, + { + "epoch": 1.9349333333333334, + "grad_norm": 0.883269190788269, + "learning_rate": 6.741884066302245e-06, + "loss": 0.3019, + "step": 2721 + }, + { + "epoch": 1.9356444444444443, + "grad_norm": 0.9224371910095215, + "learning_rate": 6.73406070784703e-06, + "loss": 0.337, + "step": 2722 + }, + { + "epoch": 1.9363555555555556, + "grad_norm": 0.964763343334198, + "learning_rate": 6.726239586337408e-06, + "loss": 0.2804, + "step": 2723 + }, + { + "epoch": 1.9370666666666667, + "grad_norm": 0.8234248757362366, + "learning_rate": 6.7184207071303065e-06, + "loss": 0.266, + "step": 2724 + }, + { + "epoch": 1.9377777777777778, + "grad_norm": 0.7993099689483643, + "learning_rate": 6.710604075581133e-06, + "loss": 0.3053, + "step": 2725 + }, + { + "epoch": 1.938488888888889, + "grad_norm": 0.9199758172035217, + "learning_rate": 6.702789697043739e-06, + "loss": 0.3027, + "step": 2726 + }, + { + "epoch": 1.9392, + "grad_norm": 0.8635696768760681, + "learning_rate": 6.6949775768704486e-06, + "loss": 0.2676, + "step": 2727 + }, + { + "epoch": 1.9399111111111111, + "grad_norm": 1.056146502494812, + "learning_rate": 6.687167720412022e-06, + "loss": 0.3014, + "step": 2728 + }, + { + "epoch": 1.9406222222222222, + "grad_norm": 0.8131300210952759, + "learning_rate": 6.679360133017687e-06, + "loss": 0.2845, + "step": 2729 + }, + { + "epoch": 1.9413333333333334, + "grad_norm": 0.8043224811553955, + "learning_rate": 6.6715548200351025e-06, + "loss": 0.2796, + "step": 2730 + }, + { + "epoch": 1.9420444444444445, + "grad_norm": 0.7607710957527161, + "learning_rate": 6.6637517868103774e-06, + "loss": 0.2777, + "step": 2731 + }, + { + "epoch": 1.9427555555555556, + "grad_norm": 0.8876447677612305, + "learning_rate": 6.6559510386880575e-06, + "loss": 0.2687, + "step": 2732 + }, + { + "epoch": 1.9434666666666667, + "grad_norm": 0.9118863344192505, + "learning_rate": 6.648152581011121e-06, + "loss": 0.2895, + "step": 2733 + }, + { + "epoch": 1.9441777777777778, + "grad_norm": 0.8442678451538086, + "learning_rate": 6.640356419120981e-06, + "loss": 0.286, + "step": 2734 + }, + { + "epoch": 1.944888888888889, + "grad_norm": 0.8080199956893921, + "learning_rate": 6.632562558357474e-06, + "loss": 0.3023, + "step": 2735 + }, + { + "epoch": 1.9456, + "grad_norm": 0.9762502908706665, + "learning_rate": 6.624771004058869e-06, + "loss": 0.3013, + "step": 2736 + }, + { + "epoch": 1.946311111111111, + "grad_norm": 0.8486790657043457, + "learning_rate": 6.61698176156184e-06, + "loss": 0.2936, + "step": 2737 + }, + { + "epoch": 1.9470222222222222, + "grad_norm": 0.7586075663566589, + "learning_rate": 6.609194836201498e-06, + "loss": 0.266, + "step": 2738 + }, + { + "epoch": 1.9477333333333333, + "grad_norm": 0.8430595993995667, + "learning_rate": 6.601410233311344e-06, + "loss": 0.2796, + "step": 2739 + }, + { + "epoch": 1.9484444444444444, + "grad_norm": 0.8738642334938049, + "learning_rate": 6.59362795822331e-06, + "loss": 0.2526, + "step": 2740 + }, + { + "epoch": 1.9491555555555555, + "grad_norm": 0.9492768049240112, + "learning_rate": 6.585848016267713e-06, + "loss": 0.2537, + "step": 2741 + }, + { + "epoch": 1.9498666666666666, + "grad_norm": 1.400539755821228, + "learning_rate": 6.578070412773294e-06, + "loss": 0.3073, + "step": 2742 + }, + { + "epoch": 1.9505777777777777, + "grad_norm": 0.9515194296836853, + "learning_rate": 6.570295153067172e-06, + "loss": 0.2815, + "step": 2743 + }, + { + "epoch": 1.9512888888888889, + "grad_norm": 0.9849377274513245, + "learning_rate": 6.562522242474873e-06, + "loss": 0.2659, + "step": 2744 + }, + { + "epoch": 1.952, + "grad_norm": 1.1352920532226562, + "learning_rate": 6.554751686320309e-06, + "loss": 0.2602, + "step": 2745 + }, + { + "epoch": 1.952711111111111, + "grad_norm": 0.9879097938537598, + "learning_rate": 6.546983489925783e-06, + "loss": 0.2714, + "step": 2746 + }, + { + "epoch": 1.9534222222222222, + "grad_norm": 0.87734454870224, + "learning_rate": 6.539217658611975e-06, + "loss": 0.2265, + "step": 2747 + }, + { + "epoch": 1.9541333333333335, + "grad_norm": 1.2946680784225464, + "learning_rate": 6.531454197697956e-06, + "loss": 0.2823, + "step": 2748 + }, + { + "epoch": 1.9548444444444444, + "grad_norm": 1.0743485689163208, + "learning_rate": 6.52369311250116e-06, + "loss": 0.2856, + "step": 2749 + }, + { + "epoch": 1.9555555555555557, + "grad_norm": 0.8076424598693848, + "learning_rate": 6.51593440833741e-06, + "loss": 0.2436, + "step": 2750 + }, + { + "epoch": 1.9562666666666666, + "grad_norm": 0.9477840065956116, + "learning_rate": 6.508178090520878e-06, + "loss": 0.2796, + "step": 2751 + }, + { + "epoch": 1.956977777777778, + "grad_norm": 1.0164532661437988, + "learning_rate": 6.5004241643641204e-06, + "loss": 0.2865, + "step": 2752 + }, + { + "epoch": 1.9576888888888888, + "grad_norm": 0.9441478252410889, + "learning_rate": 6.492672635178048e-06, + "loss": 0.2807, + "step": 2753 + }, + { + "epoch": 1.9584000000000001, + "grad_norm": 0.9758086204528809, + "learning_rate": 6.484923508271921e-06, + "loss": 0.2722, + "step": 2754 + }, + { + "epoch": 1.959111111111111, + "grad_norm": 1.1609536409378052, + "learning_rate": 6.477176788953373e-06, + "loss": 0.2788, + "step": 2755 + }, + { + "epoch": 1.9598222222222224, + "grad_norm": 1.0413261651992798, + "learning_rate": 6.469432482528371e-06, + "loss": 0.3388, + "step": 2756 + }, + { + "epoch": 1.9605333333333332, + "grad_norm": 1.0004788637161255, + "learning_rate": 6.461690594301242e-06, + "loss": 0.2958, + "step": 2757 + }, + { + "epoch": 1.9612444444444446, + "grad_norm": 1.1758333444595337, + "learning_rate": 6.453951129574644e-06, + "loss": 0.2858, + "step": 2758 + }, + { + "epoch": 1.9619555555555555, + "grad_norm": 1.0547860860824585, + "learning_rate": 6.446214093649591e-06, + "loss": 0.3057, + "step": 2759 + }, + { + "epoch": 1.9626666666666668, + "grad_norm": 1.0639702081680298, + "learning_rate": 6.4384794918254155e-06, + "loss": 0.2565, + "step": 2760 + }, + { + "epoch": 1.9633777777777777, + "grad_norm": 1.0899959802627563, + "learning_rate": 6.430747329399802e-06, + "loss": 0.2962, + "step": 2761 + }, + { + "epoch": 1.964088888888889, + "grad_norm": 1.0920106172561646, + "learning_rate": 6.423017611668745e-06, + "loss": 0.3034, + "step": 2762 + }, + { + "epoch": 1.9647999999999999, + "grad_norm": 0.991718590259552, + "learning_rate": 6.415290343926582e-06, + "loss": 0.2673, + "step": 2763 + }, + { + "epoch": 1.9655111111111112, + "grad_norm": 0.9146506786346436, + "learning_rate": 6.407565531465958e-06, + "loss": 0.2897, + "step": 2764 + }, + { + "epoch": 1.966222222222222, + "grad_norm": 1.1896411180496216, + "learning_rate": 6.399843179577846e-06, + "loss": 0.2873, + "step": 2765 + }, + { + "epoch": 1.9669333333333334, + "grad_norm": 0.7962215542793274, + "learning_rate": 6.392123293551524e-06, + "loss": 0.253, + "step": 2766 + }, + { + "epoch": 1.9676444444444443, + "grad_norm": 1.2339417934417725, + "learning_rate": 6.384405878674598e-06, + "loss": 0.322, + "step": 2767 + }, + { + "epoch": 1.9683555555555556, + "grad_norm": 0.98467618227005, + "learning_rate": 6.3766909402329595e-06, + "loss": 0.294, + "step": 2768 + }, + { + "epoch": 1.9690666666666665, + "grad_norm": 0.7929223775863647, + "learning_rate": 6.3689784835108205e-06, + "loss": 0.2824, + "step": 2769 + }, + { + "epoch": 1.9697777777777778, + "grad_norm": 1.0007925033569336, + "learning_rate": 6.36126851379069e-06, + "loss": 0.2922, + "step": 2770 + }, + { + "epoch": 1.9704888888888887, + "grad_norm": 1.0785821676254272, + "learning_rate": 6.353561036353366e-06, + "loss": 0.2565, + "step": 2771 + }, + { + "epoch": 1.9712, + "grad_norm": 0.8675540089607239, + "learning_rate": 6.3458560564779506e-06, + "loss": 0.3265, + "step": 2772 + }, + { + "epoch": 1.971911111111111, + "grad_norm": 0.8996270895004272, + "learning_rate": 6.338153579441823e-06, + "loss": 0.2562, + "step": 2773 + }, + { + "epoch": 1.9726222222222223, + "grad_norm": 1.0816200971603394, + "learning_rate": 6.330453610520659e-06, + "loss": 0.2971, + "step": 2774 + }, + { + "epoch": 1.9733333333333334, + "grad_norm": 0.848760724067688, + "learning_rate": 6.32275615498841e-06, + "loss": 0.2468, + "step": 2775 + }, + { + "epoch": 1.9740444444444445, + "grad_norm": 0.9875853657722473, + "learning_rate": 6.315061218117311e-06, + "loss": 0.2675, + "step": 2776 + }, + { + "epoch": 1.9747555555555556, + "grad_norm": 0.9403297901153564, + "learning_rate": 6.30736880517786e-06, + "loss": 0.2796, + "step": 2777 + }, + { + "epoch": 1.9754666666666667, + "grad_norm": 1.011225700378418, + "learning_rate": 6.299678921438845e-06, + "loss": 0.2961, + "step": 2778 + }, + { + "epoch": 1.9761777777777778, + "grad_norm": 0.960138201713562, + "learning_rate": 6.2919915721673055e-06, + "loss": 0.2834, + "step": 2779 + }, + { + "epoch": 1.976888888888889, + "grad_norm": 0.9259982705116272, + "learning_rate": 6.284306762628556e-06, + "loss": 0.2921, + "step": 2780 + }, + { + "epoch": 1.9776, + "grad_norm": 1.3185292482376099, + "learning_rate": 6.276624498086161e-06, + "loss": 0.2661, + "step": 2781 + }, + { + "epoch": 1.9783111111111111, + "grad_norm": 1.274665117263794, + "learning_rate": 6.268944783801954e-06, + "loss": 0.2423, + "step": 2782 + }, + { + "epoch": 1.9790222222222222, + "grad_norm": 0.8618654012680054, + "learning_rate": 6.2612676250360085e-06, + "loss": 0.2706, + "step": 2783 + }, + { + "epoch": 1.9797333333333333, + "grad_norm": 0.8603076934814453, + "learning_rate": 6.253593027046663e-06, + "loss": 0.2655, + "step": 2784 + }, + { + "epoch": 1.9804444444444445, + "grad_norm": 1.1205596923828125, + "learning_rate": 6.245920995090484e-06, + "loss": 0.2965, + "step": 2785 + }, + { + "epoch": 1.9811555555555556, + "grad_norm": 1.2991588115692139, + "learning_rate": 6.238251534422295e-06, + "loss": 0.2762, + "step": 2786 + }, + { + "epoch": 1.9818666666666667, + "grad_norm": 0.9293467402458191, + "learning_rate": 6.230584650295155e-06, + "loss": 0.2753, + "step": 2787 + }, + { + "epoch": 1.9825777777777778, + "grad_norm": 1.3786507844924927, + "learning_rate": 6.22292034796035e-06, + "loss": 0.2429, + "step": 2788 + }, + { + "epoch": 1.9832888888888889, + "grad_norm": 1.0292890071868896, + "learning_rate": 6.215258632667415e-06, + "loss": 0.2943, + "step": 2789 + }, + { + "epoch": 1.984, + "grad_norm": 0.7866095304489136, + "learning_rate": 6.20759950966409e-06, + "loss": 0.2366, + "step": 2790 + }, + { + "epoch": 1.984711111111111, + "grad_norm": 0.9173397421836853, + "learning_rate": 6.199942984196362e-06, + "loss": 0.2404, + "step": 2791 + }, + { + "epoch": 1.9854222222222222, + "grad_norm": 0.9784284234046936, + "learning_rate": 6.19228906150842e-06, + "loss": 0.2738, + "step": 2792 + }, + { + "epoch": 1.9861333333333333, + "grad_norm": 0.892300546169281, + "learning_rate": 6.184637746842686e-06, + "loss": 0.2852, + "step": 2793 + }, + { + "epoch": 1.9868444444444444, + "grad_norm": 1.168915033340454, + "learning_rate": 6.176989045439785e-06, + "loss": 0.3182, + "step": 2794 + }, + { + "epoch": 1.9875555555555555, + "grad_norm": 1.120707631111145, + "learning_rate": 6.169342962538555e-06, + "loss": 0.2649, + "step": 2795 + }, + { + "epoch": 1.9882666666666666, + "grad_norm": 0.8865916132926941, + "learning_rate": 6.161699503376042e-06, + "loss": 0.3065, + "step": 2796 + }, + { + "epoch": 1.9889777777777777, + "grad_norm": 1.1394810676574707, + "learning_rate": 6.154058673187497e-06, + "loss": 0.2935, + "step": 2797 + }, + { + "epoch": 1.9896888888888888, + "grad_norm": 0.9089357256889343, + "learning_rate": 6.14642047720636e-06, + "loss": 0.3013, + "step": 2798 + }, + { + "epoch": 1.9904, + "grad_norm": 0.9499253630638123, + "learning_rate": 6.138784920664282e-06, + "loss": 0.2691, + "step": 2799 + }, + { + "epoch": 1.991111111111111, + "grad_norm": 0.8945969939231873, + "learning_rate": 6.131152008791092e-06, + "loss": 0.2812, + "step": 2800 + }, + { + "epoch": 1.9918222222222224, + "grad_norm": 0.8009551167488098, + "learning_rate": 6.12352174681482e-06, + "loss": 0.2626, + "step": 2801 + }, + { + "epoch": 1.9925333333333333, + "grad_norm": 0.8699034452438354, + "learning_rate": 6.115894139961668e-06, + "loss": 0.2833, + "step": 2802 + }, + { + "epoch": 1.9932444444444446, + "grad_norm": 0.7742673754692078, + "learning_rate": 6.108269193456033e-06, + "loss": 0.2331, + "step": 2803 + }, + { + "epoch": 1.9939555555555555, + "grad_norm": 0.8745961785316467, + "learning_rate": 6.1006469125204785e-06, + "loss": 0.2716, + "step": 2804 + }, + { + "epoch": 1.9946666666666668, + "grad_norm": 0.9071557521820068, + "learning_rate": 6.093027302375748e-06, + "loss": 0.3022, + "step": 2805 + }, + { + "epoch": 1.9953777777777777, + "grad_norm": 0.8307523131370544, + "learning_rate": 6.085410368240759e-06, + "loss": 0.2389, + "step": 2806 + }, + { + "epoch": 1.996088888888889, + "grad_norm": 1.084177851676941, + "learning_rate": 6.077796115332582e-06, + "loss": 0.3196, + "step": 2807 + }, + { + "epoch": 1.9968, + "grad_norm": 0.9772243499755859, + "learning_rate": 6.070184548866471e-06, + "loss": 0.2645, + "step": 2808 + }, + { + "epoch": 1.9975111111111112, + "grad_norm": 0.8574574589729309, + "learning_rate": 6.062575674055822e-06, + "loss": 0.2969, + "step": 2809 + }, + { + "epoch": 1.9982222222222221, + "grad_norm": 0.8409656286239624, + "learning_rate": 6.054969496112202e-06, + "loss": 0.2867, + "step": 2810 + }, + { + "epoch": 1.9989333333333335, + "grad_norm": 0.9655705094337463, + "learning_rate": 6.047366020245316e-06, + "loss": 0.2786, + "step": 2811 + }, + { + "epoch": 1.9996444444444443, + "grad_norm": 1.1685421466827393, + "learning_rate": 6.039765251663034e-06, + "loss": 0.2639, + "step": 2812 + }, + { + "epoch": 2.0003555555555557, + "grad_norm": 0.9600791931152344, + "learning_rate": 6.032167195571357e-06, + "loss": 0.2766, + "step": 2813 + }, + { + "epoch": 2.0010666666666665, + "grad_norm": 0.8358759880065918, + "learning_rate": 6.024571857174443e-06, + "loss": 0.239, + "step": 2814 + }, + { + "epoch": 2.001777777777778, + "grad_norm": 0.9488192796707153, + "learning_rate": 6.016979241674574e-06, + "loss": 0.3134, + "step": 2815 + }, + { + "epoch": 2.0024888888888888, + "grad_norm": 0.8311015367507935, + "learning_rate": 6.009389354272175e-06, + "loss": 0.251, + "step": 2816 + }, + { + "epoch": 2.0032, + "grad_norm": 1.1861385107040405, + "learning_rate": 6.0018022001658e-06, + "loss": 0.2406, + "step": 2817 + }, + { + "epoch": 2.003911111111111, + "grad_norm": 0.8547160029411316, + "learning_rate": 5.994217784552139e-06, + "loss": 0.2876, + "step": 2818 + }, + { + "epoch": 2.0046222222222223, + "grad_norm": 1.082413673400879, + "learning_rate": 5.986636112625987e-06, + "loss": 0.2553, + "step": 2819 + }, + { + "epoch": 2.005333333333333, + "grad_norm": 0.7898240685462952, + "learning_rate": 5.979057189580284e-06, + "loss": 0.2338, + "step": 2820 + }, + { + "epoch": 2.0060444444444445, + "grad_norm": 1.168239951133728, + "learning_rate": 5.971481020606067e-06, + "loss": 0.2826, + "step": 2821 + }, + { + "epoch": 2.0067555555555554, + "grad_norm": 0.8325672745704651, + "learning_rate": 5.963907610892497e-06, + "loss": 0.2909, + "step": 2822 + }, + { + "epoch": 2.0074666666666667, + "grad_norm": 1.4545872211456299, + "learning_rate": 5.956336965626848e-06, + "loss": 0.2745, + "step": 2823 + }, + { + "epoch": 2.0081777777777776, + "grad_norm": 1.6204991340637207, + "learning_rate": 5.948769089994486e-06, + "loss": 0.2902, + "step": 2824 + }, + { + "epoch": 2.008888888888889, + "grad_norm": 1.0876424312591553, + "learning_rate": 5.9412039891789e-06, + "loss": 0.2458, + "step": 2825 + }, + { + "epoch": 2.0096, + "grad_norm": 0.8750013709068298, + "learning_rate": 5.93364166836166e-06, + "loss": 0.2855, + "step": 2826 + }, + { + "epoch": 2.010311111111111, + "grad_norm": 1.0177582502365112, + "learning_rate": 5.926082132722445e-06, + "loss": 0.2785, + "step": 2827 + }, + { + "epoch": 2.011022222222222, + "grad_norm": 0.9281204342842102, + "learning_rate": 5.918525387439014e-06, + "loss": 0.248, + "step": 2828 + }, + { + "epoch": 2.0117333333333334, + "grad_norm": 1.063319206237793, + "learning_rate": 5.910971437687232e-06, + "loss": 0.2679, + "step": 2829 + }, + { + "epoch": 2.0124444444444443, + "grad_norm": 0.8693380355834961, + "learning_rate": 5.903420288641029e-06, + "loss": 0.2691, + "step": 2830 + }, + { + "epoch": 2.0131555555555556, + "grad_norm": 1.1312817335128784, + "learning_rate": 5.895871945472434e-06, + "loss": 0.2957, + "step": 2831 + }, + { + "epoch": 2.0138666666666665, + "grad_norm": 0.9749444723129272, + "learning_rate": 5.888326413351542e-06, + "loss": 0.2399, + "step": 2832 + }, + { + "epoch": 2.014577777777778, + "grad_norm": 0.8435316681861877, + "learning_rate": 5.880783697446535e-06, + "loss": 0.2892, + "step": 2833 + }, + { + "epoch": 2.0152888888888887, + "grad_norm": 0.8448778390884399, + "learning_rate": 5.873243802923651e-06, + "loss": 0.269, + "step": 2834 + }, + { + "epoch": 2.016, + "grad_norm": 0.9506697058677673, + "learning_rate": 5.865706734947209e-06, + "loss": 0.257, + "step": 2835 + }, + { + "epoch": 2.016711111111111, + "grad_norm": 0.8200880885124207, + "learning_rate": 5.85817249867958e-06, + "loss": 0.248, + "step": 2836 + }, + { + "epoch": 2.0174222222222222, + "grad_norm": 0.8496691584587097, + "learning_rate": 5.85064109928121e-06, + "loss": 0.2639, + "step": 2837 + }, + { + "epoch": 2.018133333333333, + "grad_norm": 0.9549514651298523, + "learning_rate": 5.843112541910587e-06, + "loss": 0.3222, + "step": 2838 + }, + { + "epoch": 2.0188444444444444, + "grad_norm": 0.9337554574012756, + "learning_rate": 5.835586831724258e-06, + "loss": 0.2476, + "step": 2839 + }, + { + "epoch": 2.0195555555555558, + "grad_norm": 1.0488146543502808, + "learning_rate": 5.828063973876834e-06, + "loss": 0.3193, + "step": 2840 + }, + { + "epoch": 2.0202666666666667, + "grad_norm": 0.9039694666862488, + "learning_rate": 5.8205439735209445e-06, + "loss": 0.2847, + "step": 2841 + }, + { + "epoch": 2.020977777777778, + "grad_norm": 0.8752442002296448, + "learning_rate": 5.813026835807282e-06, + "loss": 0.3244, + "step": 2842 + }, + { + "epoch": 2.021688888888889, + "grad_norm": 1.0878524780273438, + "learning_rate": 5.805512565884574e-06, + "loss": 0.2795, + "step": 2843 + }, + { + "epoch": 2.0224, + "grad_norm": 1.1186189651489258, + "learning_rate": 5.7980011688995806e-06, + "loss": 0.2861, + "step": 2844 + }, + { + "epoch": 2.023111111111111, + "grad_norm": 1.0342803001403809, + "learning_rate": 5.790492649997098e-06, + "loss": 0.2567, + "step": 2845 + }, + { + "epoch": 2.0238222222222224, + "grad_norm": 0.8002597689628601, + "learning_rate": 5.78298701431995e-06, + "loss": 0.2645, + "step": 2846 + }, + { + "epoch": 2.0245333333333333, + "grad_norm": 1.3833887577056885, + "learning_rate": 5.775484267008982e-06, + "loss": 0.2773, + "step": 2847 + }, + { + "epoch": 2.0252444444444446, + "grad_norm": 1.1828534603118896, + "learning_rate": 5.7679844132030674e-06, + "loss": 0.2719, + "step": 2848 + }, + { + "epoch": 2.0259555555555555, + "grad_norm": 0.9774261116981506, + "learning_rate": 5.760487458039092e-06, + "loss": 0.2809, + "step": 2849 + }, + { + "epoch": 2.026666666666667, + "grad_norm": 1.071334719657898, + "learning_rate": 5.7529934066519585e-06, + "loss": 0.2816, + "step": 2850 + }, + { + "epoch": 2.0273777777777777, + "grad_norm": 1.1346110105514526, + "learning_rate": 5.745502264174583e-06, + "loss": 0.3044, + "step": 2851 + }, + { + "epoch": 2.028088888888889, + "grad_norm": 1.3303066492080688, + "learning_rate": 5.738014035737885e-06, + "loss": 0.281, + "step": 2852 + }, + { + "epoch": 2.0288, + "grad_norm": 0.9083513617515564, + "learning_rate": 5.730528726470792e-06, + "loss": 0.2572, + "step": 2853 + }, + { + "epoch": 2.0295111111111113, + "grad_norm": 0.949312686920166, + "learning_rate": 5.723046341500231e-06, + "loss": 0.2869, + "step": 2854 + }, + { + "epoch": 2.030222222222222, + "grad_norm": 1.0865838527679443, + "learning_rate": 5.715566885951117e-06, + "loss": 0.2856, + "step": 2855 + }, + { + "epoch": 2.0309333333333335, + "grad_norm": 0.8760380744934082, + "learning_rate": 5.708090364946376e-06, + "loss": 0.2546, + "step": 2856 + }, + { + "epoch": 2.0316444444444444, + "grad_norm": 0.8068846464157104, + "learning_rate": 5.7006167836069115e-06, + "loss": 0.2563, + "step": 2857 + }, + { + "epoch": 2.0323555555555557, + "grad_norm": 0.8824520707130432, + "learning_rate": 5.693146147051618e-06, + "loss": 0.3047, + "step": 2858 + }, + { + "epoch": 2.0330666666666666, + "grad_norm": 1.2335296869277954, + "learning_rate": 5.68567846039737e-06, + "loss": 0.2609, + "step": 2859 + }, + { + "epoch": 2.033777777777778, + "grad_norm": 1.141453742980957, + "learning_rate": 5.678213728759024e-06, + "loss": 0.2914, + "step": 2860 + }, + { + "epoch": 2.034488888888889, + "grad_norm": 0.9362562894821167, + "learning_rate": 5.6707519572494115e-06, + "loss": 0.2941, + "step": 2861 + }, + { + "epoch": 2.0352, + "grad_norm": 0.941246509552002, + "learning_rate": 5.663293150979339e-06, + "loss": 0.2436, + "step": 2862 + }, + { + "epoch": 2.035911111111111, + "grad_norm": 1.1526888608932495, + "learning_rate": 5.655837315057577e-06, + "loss": 0.2364, + "step": 2863 + }, + { + "epoch": 2.0366222222222223, + "grad_norm": 1.0874254703521729, + "learning_rate": 5.648384454590867e-06, + "loss": 0.2686, + "step": 2864 + }, + { + "epoch": 2.037333333333333, + "grad_norm": 0.8492470383644104, + "learning_rate": 5.640934574683913e-06, + "loss": 0.2504, + "step": 2865 + }, + { + "epoch": 2.0380444444444445, + "grad_norm": 1.1354244947433472, + "learning_rate": 5.633487680439362e-06, + "loss": 0.2571, + "step": 2866 + }, + { + "epoch": 2.0387555555555554, + "grad_norm": 1.1935374736785889, + "learning_rate": 5.626043776957844e-06, + "loss": 0.2684, + "step": 2867 + }, + { + "epoch": 2.0394666666666668, + "grad_norm": 1.0071823596954346, + "learning_rate": 5.618602869337909e-06, + "loss": 0.2746, + "step": 2868 + }, + { + "epoch": 2.0401777777777776, + "grad_norm": 1.1181355714797974, + "learning_rate": 5.611164962676088e-06, + "loss": 0.2811, + "step": 2869 + }, + { + "epoch": 2.040888888888889, + "grad_norm": 1.510095477104187, + "learning_rate": 5.6037300620668235e-06, + "loss": 0.3018, + "step": 2870 + }, + { + "epoch": 2.0416, + "grad_norm": 1.186015248298645, + "learning_rate": 5.5962981726025325e-06, + "loss": 0.2686, + "step": 2871 + }, + { + "epoch": 2.042311111111111, + "grad_norm": 1.0770788192749023, + "learning_rate": 5.588869299373533e-06, + "loss": 0.3159, + "step": 2872 + }, + { + "epoch": 2.043022222222222, + "grad_norm": 1.0861624479293823, + "learning_rate": 5.581443447468111e-06, + "loss": 0.2774, + "step": 2873 + }, + { + "epoch": 2.0437333333333334, + "grad_norm": 1.0733383893966675, + "learning_rate": 5.574020621972467e-06, + "loss": 0.263, + "step": 2874 + }, + { + "epoch": 2.0444444444444443, + "grad_norm": 1.1308029890060425, + "learning_rate": 5.566600827970718e-06, + "loss": 0.2335, + "step": 2875 + }, + { + "epoch": 2.0451555555555556, + "grad_norm": 0.9299684762954712, + "learning_rate": 5.559184070544933e-06, + "loss": 0.2881, + "step": 2876 + }, + { + "epoch": 2.0458666666666665, + "grad_norm": 0.8135644793510437, + "learning_rate": 5.5517703547750705e-06, + "loss": 0.2685, + "step": 2877 + }, + { + "epoch": 2.046577777777778, + "grad_norm": 1.1901438236236572, + "learning_rate": 5.544359685739033e-06, + "loss": 0.2536, + "step": 2878 + }, + { + "epoch": 2.0472888888888887, + "grad_norm": 1.0705538988113403, + "learning_rate": 5.536952068512609e-06, + "loss": 0.2598, + "step": 2879 + }, + { + "epoch": 2.048, + "grad_norm": 0.9520405530929565, + "learning_rate": 5.529547508169526e-06, + "loss": 0.2583, + "step": 2880 + }, + { + "epoch": 2.048711111111111, + "grad_norm": 1.019066572189331, + "learning_rate": 5.522146009781388e-06, + "loss": 0.2626, + "step": 2881 + }, + { + "epoch": 2.0494222222222223, + "grad_norm": 0.9103976488113403, + "learning_rate": 5.514747578417731e-06, + "loss": 0.2847, + "step": 2882 + }, + { + "epoch": 2.050133333333333, + "grad_norm": 0.9804105758666992, + "learning_rate": 5.507352219145965e-06, + "loss": 0.2639, + "step": 2883 + }, + { + "epoch": 2.0508444444444445, + "grad_norm": 0.9668303728103638, + "learning_rate": 5.4999599370314085e-06, + "loss": 0.212, + "step": 2884 + }, + { + "epoch": 2.0515555555555554, + "grad_norm": 1.0453071594238281, + "learning_rate": 5.4925707371372725e-06, + "loss": 0.2761, + "step": 2885 + }, + { + "epoch": 2.0522666666666667, + "grad_norm": 1.2840666770935059, + "learning_rate": 5.4851846245246545e-06, + "loss": 0.2998, + "step": 2886 + }, + { + "epoch": 2.0529777777777776, + "grad_norm": 1.0720361471176147, + "learning_rate": 5.477801604252536e-06, + "loss": 0.2895, + "step": 2887 + }, + { + "epoch": 2.053688888888889, + "grad_norm": 0.9466294646263123, + "learning_rate": 5.470421681377785e-06, + "loss": 0.2701, + "step": 2888 + }, + { + "epoch": 2.0544, + "grad_norm": 1.1145267486572266, + "learning_rate": 5.463044860955142e-06, + "loss": 0.2368, + "step": 2889 + }, + { + "epoch": 2.055111111111111, + "grad_norm": 1.2340354919433594, + "learning_rate": 5.455671148037225e-06, + "loss": 0.2898, + "step": 2890 + }, + { + "epoch": 2.0558222222222224, + "grad_norm": 1.0386790037155151, + "learning_rate": 5.448300547674532e-06, + "loss": 0.2791, + "step": 2891 + }, + { + "epoch": 2.0565333333333333, + "grad_norm": 0.9507420063018799, + "learning_rate": 5.440933064915414e-06, + "loss": 0.2632, + "step": 2892 + }, + { + "epoch": 2.0572444444444447, + "grad_norm": 1.0850605964660645, + "learning_rate": 5.433568704806095e-06, + "loss": 0.2368, + "step": 2893 + }, + { + "epoch": 2.0579555555555555, + "grad_norm": 1.0632293224334717, + "learning_rate": 5.426207472390661e-06, + "loss": 0.2496, + "step": 2894 + }, + { + "epoch": 2.058666666666667, + "grad_norm": 1.1629679203033447, + "learning_rate": 5.4188493727110525e-06, + "loss": 0.2381, + "step": 2895 + }, + { + "epoch": 2.0593777777777778, + "grad_norm": 1.1950361728668213, + "learning_rate": 5.411494410807065e-06, + "loss": 0.2906, + "step": 2896 + }, + { + "epoch": 2.060088888888889, + "grad_norm": 1.058361291885376, + "learning_rate": 5.404142591716348e-06, + "loss": 0.2511, + "step": 2897 + }, + { + "epoch": 2.0608, + "grad_norm": 0.9299618005752563, + "learning_rate": 5.396793920474397e-06, + "loss": 0.266, + "step": 2898 + }, + { + "epoch": 2.0615111111111113, + "grad_norm": 1.1203224658966064, + "learning_rate": 5.389448402114546e-06, + "loss": 0.2132, + "step": 2899 + }, + { + "epoch": 2.062222222222222, + "grad_norm": 1.1637037992477417, + "learning_rate": 5.382106041667976e-06, + "loss": 0.2745, + "step": 2900 + }, + { + "epoch": 2.0629333333333335, + "grad_norm": 0.8635064363479614, + "learning_rate": 5.3747668441637046e-06, + "loss": 0.241, + "step": 2901 + }, + { + "epoch": 2.0636444444444444, + "grad_norm": 1.1659942865371704, + "learning_rate": 5.3674308146285804e-06, + "loss": 0.261, + "step": 2902 + }, + { + "epoch": 2.0643555555555557, + "grad_norm": 1.0490862131118774, + "learning_rate": 5.36009795808728e-06, + "loss": 0.3504, + "step": 2903 + }, + { + "epoch": 2.0650666666666666, + "grad_norm": 0.8469544649124146, + "learning_rate": 5.352768279562315e-06, + "loss": 0.2691, + "step": 2904 + }, + { + "epoch": 2.065777777777778, + "grad_norm": 1.0089105367660522, + "learning_rate": 5.3454417840740125e-06, + "loss": 0.2799, + "step": 2905 + }, + { + "epoch": 2.066488888888889, + "grad_norm": 1.0562530755996704, + "learning_rate": 5.3381184766405215e-06, + "loss": 0.2752, + "step": 2906 + }, + { + "epoch": 2.0672, + "grad_norm": 0.9492927193641663, + "learning_rate": 5.3307983622778095e-06, + "loss": 0.2731, + "step": 2907 + }, + { + "epoch": 2.067911111111111, + "grad_norm": 0.9867216944694519, + "learning_rate": 5.323481445999654e-06, + "loss": 0.2732, + "step": 2908 + }, + { + "epoch": 2.0686222222222224, + "grad_norm": 1.0084117650985718, + "learning_rate": 5.316167732817643e-06, + "loss": 0.2735, + "step": 2909 + }, + { + "epoch": 2.0693333333333332, + "grad_norm": 0.8792888522148132, + "learning_rate": 5.308857227741173e-06, + "loss": 0.2799, + "step": 2910 + }, + { + "epoch": 2.0700444444444446, + "grad_norm": 0.7637631893157959, + "learning_rate": 5.301549935777441e-06, + "loss": 0.291, + "step": 2911 + }, + { + "epoch": 2.0707555555555555, + "grad_norm": 0.923477292060852, + "learning_rate": 5.294245861931442e-06, + "loss": 0.2897, + "step": 2912 + }, + { + "epoch": 2.071466666666667, + "grad_norm": 1.0692672729492188, + "learning_rate": 5.286945011205968e-06, + "loss": 0.2406, + "step": 2913 + }, + { + "epoch": 2.0721777777777777, + "grad_norm": 0.8462406992912292, + "learning_rate": 5.2796473886016075e-06, + "loss": 0.2639, + "step": 2914 + }, + { + "epoch": 2.072888888888889, + "grad_norm": 0.8960204720497131, + "learning_rate": 5.272352999116728e-06, + "loss": 0.2921, + "step": 2915 + }, + { + "epoch": 2.0736, + "grad_norm": 1.1324297189712524, + "learning_rate": 5.265061847747497e-06, + "loss": 0.2746, + "step": 2916 + }, + { + "epoch": 2.074311111111111, + "grad_norm": 1.0957542657852173, + "learning_rate": 5.257773939487842e-06, + "loss": 0.255, + "step": 2917 + }, + { + "epoch": 2.075022222222222, + "grad_norm": 0.8507300615310669, + "learning_rate": 5.250489279329501e-06, + "loss": 0.2647, + "step": 2918 + }, + { + "epoch": 2.0757333333333334, + "grad_norm": 0.7929480075836182, + "learning_rate": 5.243207872261951e-06, + "loss": 0.249, + "step": 2919 + }, + { + "epoch": 2.0764444444444443, + "grad_norm": 1.2949143648147583, + "learning_rate": 5.235929723272475e-06, + "loss": 0.2667, + "step": 2920 + }, + { + "epoch": 2.0771555555555556, + "grad_norm": 1.1818195581436157, + "learning_rate": 5.228654837346094e-06, + "loss": 0.2603, + "step": 2921 + }, + { + "epoch": 2.0778666666666665, + "grad_norm": 0.9909701943397522, + "learning_rate": 5.221383219465623e-06, + "loss": 0.2741, + "step": 2922 + }, + { + "epoch": 2.078577777777778, + "grad_norm": 0.7823296785354614, + "learning_rate": 5.214114874611609e-06, + "loss": 0.244, + "step": 2923 + }, + { + "epoch": 2.0792888888888887, + "grad_norm": 1.1611075401306152, + "learning_rate": 5.2068498077623845e-06, + "loss": 0.2344, + "step": 2924 + }, + { + "epoch": 2.08, + "grad_norm": 1.3629710674285889, + "learning_rate": 5.199588023894024e-06, + "loss": 0.2702, + "step": 2925 + }, + { + "epoch": 2.080711111111111, + "grad_norm": 1.0419052839279175, + "learning_rate": 5.192329527980342e-06, + "loss": 0.2478, + "step": 2926 + }, + { + "epoch": 2.0814222222222223, + "grad_norm": 0.8881648778915405, + "learning_rate": 5.18507432499293e-06, + "loss": 0.2392, + "step": 2927 + }, + { + "epoch": 2.082133333333333, + "grad_norm": 1.2188395261764526, + "learning_rate": 5.1778224199010905e-06, + "loss": 0.2673, + "step": 2928 + }, + { + "epoch": 2.0828444444444445, + "grad_norm": 1.469802737236023, + "learning_rate": 5.170573817671899e-06, + "loss": 0.2618, + "step": 2929 + }, + { + "epoch": 2.0835555555555554, + "grad_norm": 1.0384467840194702, + "learning_rate": 5.163328523270138e-06, + "loss": 0.2848, + "step": 2930 + }, + { + "epoch": 2.0842666666666667, + "grad_norm": 0.8417734503746033, + "learning_rate": 5.156086541658356e-06, + "loss": 0.2631, + "step": 2931 + }, + { + "epoch": 2.0849777777777776, + "grad_norm": 1.0232224464416504, + "learning_rate": 5.1488478777968e-06, + "loss": 0.2676, + "step": 2932 + }, + { + "epoch": 2.085688888888889, + "grad_norm": 1.0539512634277344, + "learning_rate": 5.141612536643476e-06, + "loss": 0.2594, + "step": 2933 + }, + { + "epoch": 2.0864, + "grad_norm": 0.9278348684310913, + "learning_rate": 5.134380523154086e-06, + "loss": 0.2631, + "step": 2934 + }, + { + "epoch": 2.087111111111111, + "grad_norm": 0.8888529539108276, + "learning_rate": 5.127151842282079e-06, + "loss": 0.2492, + "step": 2935 + }, + { + "epoch": 2.087822222222222, + "grad_norm": 0.8310458064079285, + "learning_rate": 5.1199264989785976e-06, + "loss": 0.2828, + "step": 2936 + }, + { + "epoch": 2.0885333333333334, + "grad_norm": 0.8618475794792175, + "learning_rate": 5.112704498192512e-06, + "loss": 0.3197, + "step": 2937 + }, + { + "epoch": 2.0892444444444442, + "grad_norm": 1.1704217195510864, + "learning_rate": 5.105485844870401e-06, + "loss": 0.2956, + "step": 2938 + }, + { + "epoch": 2.0899555555555556, + "grad_norm": 0.8167716264724731, + "learning_rate": 5.09827054395655e-06, + "loss": 0.2473, + "step": 2939 + }, + { + "epoch": 2.0906666666666665, + "grad_norm": 0.9328314661979675, + "learning_rate": 5.091058600392944e-06, + "loss": 0.2809, + "step": 2940 + }, + { + "epoch": 2.091377777777778, + "grad_norm": 0.8080491423606873, + "learning_rate": 5.083850019119277e-06, + "loss": 0.2622, + "step": 2941 + }, + { + "epoch": 2.092088888888889, + "grad_norm": 0.9174755215644836, + "learning_rate": 5.076644805072934e-06, + "loss": 0.2409, + "step": 2942 + }, + { + "epoch": 2.0928, + "grad_norm": 0.9131911993026733, + "learning_rate": 5.069442963188991e-06, + "loss": 0.2655, + "step": 2943 + }, + { + "epoch": 2.0935111111111113, + "grad_norm": 0.8674187064170837, + "learning_rate": 5.062244498400228e-06, + "loss": 0.2492, + "step": 2944 + }, + { + "epoch": 2.094222222222222, + "grad_norm": 0.7955570816993713, + "learning_rate": 5.055049415637094e-06, + "loss": 0.254, + "step": 2945 + }, + { + "epoch": 2.0949333333333335, + "grad_norm": 0.8933150768280029, + "learning_rate": 5.0478577198277325e-06, + "loss": 0.2684, + "step": 2946 + }, + { + "epoch": 2.0956444444444444, + "grad_norm": 0.9143484234809875, + "learning_rate": 5.040669415897963e-06, + "loss": 0.2082, + "step": 2947 + }, + { + "epoch": 2.0963555555555557, + "grad_norm": 0.8864539265632629, + "learning_rate": 5.033484508771285e-06, + "loss": 0.2824, + "step": 2948 + }, + { + "epoch": 2.0970666666666666, + "grad_norm": 0.9682989120483398, + "learning_rate": 5.026303003368869e-06, + "loss": 0.2809, + "step": 2949 + }, + { + "epoch": 2.097777777777778, + "grad_norm": 0.7518890500068665, + "learning_rate": 5.019124904609555e-06, + "loss": 0.2353, + "step": 2950 + }, + { + "epoch": 2.098488888888889, + "grad_norm": 0.8197988271713257, + "learning_rate": 5.011950217409852e-06, + "loss": 0.2385, + "step": 2951 + }, + { + "epoch": 2.0992, + "grad_norm": 0.8093001246452332, + "learning_rate": 5.004778946683932e-06, + "loss": 0.2444, + "step": 2952 + }, + { + "epoch": 2.099911111111111, + "grad_norm": 0.8627079725265503, + "learning_rate": 4.997611097343625e-06, + "loss": 0.2776, + "step": 2953 + }, + { + "epoch": 2.1006222222222224, + "grad_norm": 0.7928313612937927, + "learning_rate": 4.99044667429842e-06, + "loss": 0.2672, + "step": 2954 + }, + { + "epoch": 2.1013333333333333, + "grad_norm": 0.825973391532898, + "learning_rate": 4.983285682455458e-06, + "loss": 0.285, + "step": 2955 + }, + { + "epoch": 2.1020444444444446, + "grad_norm": 0.8718094229698181, + "learning_rate": 4.976128126719527e-06, + "loss": 0.2893, + "step": 2956 + }, + { + "epoch": 2.1027555555555555, + "grad_norm": 0.8293783068656921, + "learning_rate": 4.968974011993067e-06, + "loss": 0.2469, + "step": 2957 + }, + { + "epoch": 2.103466666666667, + "grad_norm": 0.8567909002304077, + "learning_rate": 4.9618233431761584e-06, + "loss": 0.2821, + "step": 2958 + }, + { + "epoch": 2.1041777777777777, + "grad_norm": 0.851294219493866, + "learning_rate": 4.954676125166521e-06, + "loss": 0.2708, + "step": 2959 + }, + { + "epoch": 2.104888888888889, + "grad_norm": 0.7553427815437317, + "learning_rate": 4.947532362859511e-06, + "loss": 0.2502, + "step": 2960 + }, + { + "epoch": 2.1056, + "grad_norm": 0.8868500590324402, + "learning_rate": 4.940392061148119e-06, + "loss": 0.2741, + "step": 2961 + }, + { + "epoch": 2.1063111111111112, + "grad_norm": 0.8573623895645142, + "learning_rate": 4.933255224922964e-06, + "loss": 0.2388, + "step": 2962 + }, + { + "epoch": 2.107022222222222, + "grad_norm": 0.8989682793617249, + "learning_rate": 4.926121859072293e-06, + "loss": 0.2667, + "step": 2963 + }, + { + "epoch": 2.1077333333333335, + "grad_norm": 0.8471580743789673, + "learning_rate": 4.918991968481973e-06, + "loss": 0.3188, + "step": 2964 + }, + { + "epoch": 2.1084444444444443, + "grad_norm": 0.9526674747467041, + "learning_rate": 4.911865558035494e-06, + "loss": 0.2843, + "step": 2965 + }, + { + "epoch": 2.1091555555555557, + "grad_norm": 0.9104166626930237, + "learning_rate": 4.90474263261396e-06, + "loss": 0.2315, + "step": 2966 + }, + { + "epoch": 2.1098666666666666, + "grad_norm": 0.897904098033905, + "learning_rate": 4.89762319709609e-06, + "loss": 0.2852, + "step": 2967 + }, + { + "epoch": 2.110577777777778, + "grad_norm": 0.9402526617050171, + "learning_rate": 4.890507256358211e-06, + "loss": 0.233, + "step": 2968 + }, + { + "epoch": 2.1112888888888888, + "grad_norm": 0.7319994568824768, + "learning_rate": 4.883394815274261e-06, + "loss": 0.259, + "step": 2969 + }, + { + "epoch": 2.112, + "grad_norm": 0.8777678608894348, + "learning_rate": 4.876285878715764e-06, + "loss": 0.2622, + "step": 2970 + }, + { + "epoch": 2.112711111111111, + "grad_norm": 0.8315487504005432, + "learning_rate": 4.8691804515518735e-06, + "loss": 0.2971, + "step": 2971 + }, + { + "epoch": 2.1134222222222223, + "grad_norm": 0.8097047805786133, + "learning_rate": 4.862078538649308e-06, + "loss": 0.3087, + "step": 2972 + }, + { + "epoch": 2.114133333333333, + "grad_norm": 0.8531997799873352, + "learning_rate": 4.854980144872405e-06, + "loss": 0.2666, + "step": 2973 + }, + { + "epoch": 2.1148444444444445, + "grad_norm": 0.9076053500175476, + "learning_rate": 4.847885275083068e-06, + "loss": 0.2789, + "step": 2974 + }, + { + "epoch": 2.1155555555555554, + "grad_norm": 0.9119158983230591, + "learning_rate": 4.840793934140813e-06, + "loss": 0.2567, + "step": 2975 + }, + { + "epoch": 2.1162666666666667, + "grad_norm": 0.8400464653968811, + "learning_rate": 4.833706126902709e-06, + "loss": 0.2731, + "step": 2976 + }, + { + "epoch": 2.1169777777777776, + "grad_norm": 0.814409077167511, + "learning_rate": 4.826621858223431e-06, + "loss": 0.2847, + "step": 2977 + }, + { + "epoch": 2.117688888888889, + "grad_norm": 0.8196157217025757, + "learning_rate": 4.819541132955222e-06, + "loss": 0.2766, + "step": 2978 + }, + { + "epoch": 2.1184, + "grad_norm": 0.8371366262435913, + "learning_rate": 4.812463955947884e-06, + "loss": 0.2711, + "step": 2979 + }, + { + "epoch": 2.119111111111111, + "grad_norm": 0.831911563873291, + "learning_rate": 4.805390332048813e-06, + "loss": 0.3105, + "step": 2980 + }, + { + "epoch": 2.119822222222222, + "grad_norm": 0.676557719707489, + "learning_rate": 4.798320266102946e-06, + "loss": 0.2268, + "step": 2981 + }, + { + "epoch": 2.1205333333333334, + "grad_norm": 0.8764379024505615, + "learning_rate": 4.79125376295281e-06, + "loss": 0.2728, + "step": 2982 + }, + { + "epoch": 2.1212444444444443, + "grad_norm": 0.8233434557914734, + "learning_rate": 4.784190827438462e-06, + "loss": 0.2666, + "step": 2983 + }, + { + "epoch": 2.1219555555555556, + "grad_norm": 0.8884332776069641, + "learning_rate": 4.7771314643975475e-06, + "loss": 0.2764, + "step": 2984 + }, + { + "epoch": 2.1226666666666665, + "grad_norm": 0.9366398453712463, + "learning_rate": 4.770075678665232e-06, + "loss": 0.2424, + "step": 2985 + }, + { + "epoch": 2.123377777777778, + "grad_norm": 0.8557524085044861, + "learning_rate": 4.763023475074264e-06, + "loss": 0.2668, + "step": 2986 + }, + { + "epoch": 2.1240888888888887, + "grad_norm": 0.851570188999176, + "learning_rate": 4.755974858454909e-06, + "loss": 0.2899, + "step": 2987 + }, + { + "epoch": 2.1248, + "grad_norm": 0.8133741617202759, + "learning_rate": 4.748929833634993e-06, + "loss": 0.2801, + "step": 2988 + }, + { + "epoch": 2.125511111111111, + "grad_norm": 0.7966433763504028, + "learning_rate": 4.741888405439877e-06, + "loss": 0.2768, + "step": 2989 + }, + { + "epoch": 2.1262222222222222, + "grad_norm": 0.8084270358085632, + "learning_rate": 4.734850578692461e-06, + "loss": 0.2284, + "step": 2990 + }, + { + "epoch": 2.1269333333333336, + "grad_norm": 0.7600616812705994, + "learning_rate": 4.727816358213171e-06, + "loss": 0.2664, + "step": 2991 + }, + { + "epoch": 2.1276444444444444, + "grad_norm": 1.0215866565704346, + "learning_rate": 4.720785748819973e-06, + "loss": 0.2447, + "step": 2992 + }, + { + "epoch": 2.1283555555555553, + "grad_norm": 0.7621174454689026, + "learning_rate": 4.713758755328352e-06, + "loss": 0.2367, + "step": 2993 + }, + { + "epoch": 2.1290666666666667, + "grad_norm": 0.8965351581573486, + "learning_rate": 4.706735382551318e-06, + "loss": 0.2828, + "step": 2994 + }, + { + "epoch": 2.129777777777778, + "grad_norm": 0.887436032295227, + "learning_rate": 4.699715635299411e-06, + "loss": 0.2642, + "step": 2995 + }, + { + "epoch": 2.130488888888889, + "grad_norm": 0.947029709815979, + "learning_rate": 4.692699518380664e-06, + "loss": 0.2824, + "step": 2996 + }, + { + "epoch": 2.1312, + "grad_norm": 0.9152494072914124, + "learning_rate": 4.6856870366006555e-06, + "loss": 0.2798, + "step": 2997 + }, + { + "epoch": 2.131911111111111, + "grad_norm": 0.7610787749290466, + "learning_rate": 4.678678194762444e-06, + "loss": 0.2566, + "step": 2998 + }, + { + "epoch": 2.1326222222222224, + "grad_norm": 0.8071789145469666, + "learning_rate": 4.671672997666613e-06, + "loss": 0.2801, + "step": 2999 + }, + { + "epoch": 2.1333333333333333, + "grad_norm": 0.9690770506858826, + "learning_rate": 4.6646714501112425e-06, + "loss": 0.2543, + "step": 3000 + }, + { + "epoch": 2.1340444444444446, + "grad_norm": 0.7485826015472412, + "learning_rate": 4.657673556891918e-06, + "loss": 0.2411, + "step": 3001 + }, + { + "epoch": 2.1347555555555555, + "grad_norm": 0.9378333687782288, + "learning_rate": 4.6506793228017145e-06, + "loss": 0.2384, + "step": 3002 + }, + { + "epoch": 2.135466666666667, + "grad_norm": 0.8194955587387085, + "learning_rate": 4.643688752631209e-06, + "loss": 0.2566, + "step": 3003 + }, + { + "epoch": 2.1361777777777777, + "grad_norm": 0.7976048588752747, + "learning_rate": 4.636701851168465e-06, + "loss": 0.2478, + "step": 3004 + }, + { + "epoch": 2.136888888888889, + "grad_norm": 0.8040217161178589, + "learning_rate": 4.629718623199031e-06, + "loss": 0.243, + "step": 3005 + }, + { + "epoch": 2.1376, + "grad_norm": 0.9101088047027588, + "learning_rate": 4.622739073505943e-06, + "loss": 0.2794, + "step": 3006 + }, + { + "epoch": 2.1383111111111113, + "grad_norm": 0.8490354418754578, + "learning_rate": 4.615763206869715e-06, + "loss": 0.251, + "step": 3007 + }, + { + "epoch": 2.139022222222222, + "grad_norm": 0.8418045043945312, + "learning_rate": 4.608791028068344e-06, + "loss": 0.2282, + "step": 3008 + }, + { + "epoch": 2.1397333333333335, + "grad_norm": 0.8690575957298279, + "learning_rate": 4.601822541877291e-06, + "loss": 0.2761, + "step": 3009 + }, + { + "epoch": 2.1404444444444444, + "grad_norm": 0.9156320691108704, + "learning_rate": 4.594857753069497e-06, + "loss": 0.2608, + "step": 3010 + }, + { + "epoch": 2.1411555555555557, + "grad_norm": 0.8230268359184265, + "learning_rate": 4.587896666415368e-06, + "loss": 0.3139, + "step": 3011 + }, + { + "epoch": 2.1418666666666666, + "grad_norm": 0.9055205583572388, + "learning_rate": 4.5809392866827704e-06, + "loss": 0.2535, + "step": 3012 + }, + { + "epoch": 2.142577777777778, + "grad_norm": 0.8776367902755737, + "learning_rate": 4.573985618637036e-06, + "loss": 0.2945, + "step": 3013 + }, + { + "epoch": 2.143288888888889, + "grad_norm": 0.8618184328079224, + "learning_rate": 4.5670356670409544e-06, + "loss": 0.2985, + "step": 3014 + }, + { + "epoch": 2.144, + "grad_norm": 0.820305585861206, + "learning_rate": 4.560089436654766e-06, + "loss": 0.2334, + "step": 3015 + }, + { + "epoch": 2.144711111111111, + "grad_norm": 0.790127694606781, + "learning_rate": 4.5531469322361656e-06, + "loss": 0.2816, + "step": 3016 + }, + { + "epoch": 2.1454222222222223, + "grad_norm": 0.8079758882522583, + "learning_rate": 4.546208158540296e-06, + "loss": 0.2472, + "step": 3017 + }, + { + "epoch": 2.1461333333333332, + "grad_norm": 0.9431691765785217, + "learning_rate": 4.539273120319743e-06, + "loss": 0.3025, + "step": 3018 + }, + { + "epoch": 2.1468444444444446, + "grad_norm": 0.7194493412971497, + "learning_rate": 4.532341822324533e-06, + "loss": 0.2318, + "step": 3019 + }, + { + "epoch": 2.1475555555555554, + "grad_norm": 0.8484240770339966, + "learning_rate": 4.525414269302138e-06, + "loss": 0.2659, + "step": 3020 + }, + { + "epoch": 2.1482666666666668, + "grad_norm": 0.8221232295036316, + "learning_rate": 4.5184904659974495e-06, + "loss": 0.227, + "step": 3021 + }, + { + "epoch": 2.1489777777777777, + "grad_norm": 0.7435066103935242, + "learning_rate": 4.5115704171528105e-06, + "loss": 0.2634, + "step": 3022 + }, + { + "epoch": 2.149688888888889, + "grad_norm": 1.0502856969833374, + "learning_rate": 4.504654127507973e-06, + "loss": 0.2497, + "step": 3023 + }, + { + "epoch": 2.1504, + "grad_norm": 1.1580634117126465, + "learning_rate": 4.497741601800134e-06, + "loss": 0.2329, + "step": 3024 + }, + { + "epoch": 2.151111111111111, + "grad_norm": 0.8244979977607727, + "learning_rate": 4.490832844763889e-06, + "loss": 0.245, + "step": 3025 + }, + { + "epoch": 2.151822222222222, + "grad_norm": 0.8312277793884277, + "learning_rate": 4.48392786113128e-06, + "loss": 0.2659, + "step": 3026 + }, + { + "epoch": 2.1525333333333334, + "grad_norm": 1.1239759922027588, + "learning_rate": 4.477026655631734e-06, + "loss": 0.2598, + "step": 3027 + }, + { + "epoch": 2.1532444444444443, + "grad_norm": 0.8912851810455322, + "learning_rate": 4.470129232992118e-06, + "loss": 0.2526, + "step": 3028 + }, + { + "epoch": 2.1539555555555556, + "grad_norm": 0.862596869468689, + "learning_rate": 4.4632355979366915e-06, + "loss": 0.2687, + "step": 3029 + }, + { + "epoch": 2.1546666666666665, + "grad_norm": 0.9846833348274231, + "learning_rate": 4.456345755187126e-06, + "loss": 0.277, + "step": 3030 + }, + { + "epoch": 2.155377777777778, + "grad_norm": 1.0585483312606812, + "learning_rate": 4.449459709462495e-06, + "loss": 0.2409, + "step": 3031 + }, + { + "epoch": 2.1560888888888887, + "grad_norm": 0.8098119497299194, + "learning_rate": 4.44257746547926e-06, + "loss": 0.2534, + "step": 3032 + }, + { + "epoch": 2.1568, + "grad_norm": 0.7816736102104187, + "learning_rate": 4.435699027951303e-06, + "loss": 0.2427, + "step": 3033 + }, + { + "epoch": 2.157511111111111, + "grad_norm": 0.9749840497970581, + "learning_rate": 4.428824401589871e-06, + "loss": 0.2779, + "step": 3034 + }, + { + "epoch": 2.1582222222222223, + "grad_norm": 0.9001795649528503, + "learning_rate": 4.421953591103627e-06, + "loss": 0.2578, + "step": 3035 + }, + { + "epoch": 2.158933333333333, + "grad_norm": 0.7714693546295166, + "learning_rate": 4.415086601198592e-06, + "loss": 0.2721, + "step": 3036 + }, + { + "epoch": 2.1596444444444445, + "grad_norm": 1.1203774213790894, + "learning_rate": 4.408223436578204e-06, + "loss": 0.2463, + "step": 3037 + }, + { + "epoch": 2.1603555555555554, + "grad_norm": 0.8661901354789734, + "learning_rate": 4.401364101943244e-06, + "loss": 0.2434, + "step": 3038 + }, + { + "epoch": 2.1610666666666667, + "grad_norm": 0.8025822639465332, + "learning_rate": 4.394508601991906e-06, + "loss": 0.2922, + "step": 3039 + }, + { + "epoch": 2.1617777777777776, + "grad_norm": 0.8846868872642517, + "learning_rate": 4.387656941419726e-06, + "loss": 0.2903, + "step": 3040 + }, + { + "epoch": 2.162488888888889, + "grad_norm": 0.9835827946662903, + "learning_rate": 4.380809124919628e-06, + "loss": 0.2594, + "step": 3041 + }, + { + "epoch": 2.1632, + "grad_norm": 0.8251909613609314, + "learning_rate": 4.373965157181901e-06, + "loss": 0.2168, + "step": 3042 + }, + { + "epoch": 2.163911111111111, + "grad_norm": 0.7906753420829773, + "learning_rate": 4.367125042894195e-06, + "loss": 0.2376, + "step": 3043 + }, + { + "epoch": 2.1646222222222224, + "grad_norm": 0.7942435145378113, + "learning_rate": 4.360288786741521e-06, + "loss": 0.2539, + "step": 3044 + }, + { + "epoch": 2.1653333333333333, + "grad_norm": 1.0452114343643188, + "learning_rate": 4.353456393406244e-06, + "loss": 0.2639, + "step": 3045 + }, + { + "epoch": 2.166044444444444, + "grad_norm": 0.8110646605491638, + "learning_rate": 4.3466278675681e-06, + "loss": 0.2918, + "step": 3046 + }, + { + "epoch": 2.1667555555555555, + "grad_norm": 0.8953929543495178, + "learning_rate": 4.339803213904149e-06, + "loss": 0.2478, + "step": 3047 + }, + { + "epoch": 2.167466666666667, + "grad_norm": 0.7873168587684631, + "learning_rate": 4.332982437088825e-06, + "loss": 0.2902, + "step": 3048 + }, + { + "epoch": 2.1681777777777778, + "grad_norm": 0.8493070006370544, + "learning_rate": 4.326165541793886e-06, + "loss": 0.294, + "step": 3049 + }, + { + "epoch": 2.168888888888889, + "grad_norm": 0.9186519980430603, + "learning_rate": 4.319352532688444e-06, + "loss": 0.2633, + "step": 3050 + }, + { + "epoch": 2.1696, + "grad_norm": 0.838941216468811, + "learning_rate": 4.3125434144389434e-06, + "loss": 0.2285, + "step": 3051 + }, + { + "epoch": 2.1703111111111113, + "grad_norm": 0.7792643904685974, + "learning_rate": 4.305738191709167e-06, + "loss": 0.2398, + "step": 3052 + }, + { + "epoch": 2.171022222222222, + "grad_norm": 0.9897586107254028, + "learning_rate": 4.298936869160228e-06, + "loss": 0.2382, + "step": 3053 + }, + { + "epoch": 2.1717333333333335, + "grad_norm": 0.7836630344390869, + "learning_rate": 4.292139451450569e-06, + "loss": 0.2838, + "step": 3054 + }, + { + "epoch": 2.1724444444444444, + "grad_norm": 0.7678898572921753, + "learning_rate": 4.285345943235956e-06, + "loss": 0.2255, + "step": 3055 + }, + { + "epoch": 2.1731555555555557, + "grad_norm": 0.8258121609687805, + "learning_rate": 4.2785563491694785e-06, + "loss": 0.278, + "step": 3056 + }, + { + "epoch": 2.1738666666666666, + "grad_norm": 0.9479091167449951, + "learning_rate": 4.2717706739015464e-06, + "loss": 0.2825, + "step": 3057 + }, + { + "epoch": 2.174577777777778, + "grad_norm": 1.059945821762085, + "learning_rate": 4.264988922079885e-06, + "loss": 0.2394, + "step": 3058 + }, + { + "epoch": 2.175288888888889, + "grad_norm": 0.8792638182640076, + "learning_rate": 4.2582110983495285e-06, + "loss": 0.2842, + "step": 3059 + }, + { + "epoch": 2.176, + "grad_norm": 0.9157023429870605, + "learning_rate": 4.251437207352826e-06, + "loss": 0.3037, + "step": 3060 + }, + { + "epoch": 2.176711111111111, + "grad_norm": 0.7653731107711792, + "learning_rate": 4.244667253729431e-06, + "loss": 0.316, + "step": 3061 + }, + { + "epoch": 2.1774222222222224, + "grad_norm": 1.0746371746063232, + "learning_rate": 4.237901242116299e-06, + "loss": 0.2958, + "step": 3062 + }, + { + "epoch": 2.1781333333333333, + "grad_norm": 0.8187775015830994, + "learning_rate": 4.231139177147687e-06, + "loss": 0.2608, + "step": 3063 + }, + { + "epoch": 2.1788444444444446, + "grad_norm": 0.8333412408828735, + "learning_rate": 4.224381063455147e-06, + "loss": 0.2467, + "step": 3064 + }, + { + "epoch": 2.1795555555555555, + "grad_norm": 0.9642142653465271, + "learning_rate": 4.217626905667527e-06, + "loss": 0.2556, + "step": 3065 + }, + { + "epoch": 2.180266666666667, + "grad_norm": 0.8618711233139038, + "learning_rate": 4.210876708410962e-06, + "loss": 0.2937, + "step": 3066 + }, + { + "epoch": 2.1809777777777777, + "grad_norm": 0.8279787302017212, + "learning_rate": 4.204130476308881e-06, + "loss": 0.2713, + "step": 3067 + }, + { + "epoch": 2.181688888888889, + "grad_norm": 0.8522974848747253, + "learning_rate": 4.1973882139819875e-06, + "loss": 0.2844, + "step": 3068 + }, + { + "epoch": 2.1824, + "grad_norm": 0.7352008819580078, + "learning_rate": 4.190649926048274e-06, + "loss": 0.2461, + "step": 3069 + }, + { + "epoch": 2.1831111111111112, + "grad_norm": 0.9881961941719055, + "learning_rate": 4.18391561712301e-06, + "loss": 0.2525, + "step": 3070 + }, + { + "epoch": 2.183822222222222, + "grad_norm": 0.9412872195243835, + "learning_rate": 4.177185291818732e-06, + "loss": 0.2944, + "step": 3071 + }, + { + "epoch": 2.1845333333333334, + "grad_norm": 0.7667661309242249, + "learning_rate": 4.17045895474526e-06, + "loss": 0.2517, + "step": 3072 + }, + { + "epoch": 2.1852444444444443, + "grad_norm": 1.0345288515090942, + "learning_rate": 4.163736610509676e-06, + "loss": 0.3253, + "step": 3073 + }, + { + "epoch": 2.1859555555555557, + "grad_norm": 0.9666414260864258, + "learning_rate": 4.1570182637163155e-06, + "loss": 0.2394, + "step": 3074 + }, + { + "epoch": 2.1866666666666665, + "grad_norm": 0.9668805003166199, + "learning_rate": 4.150303918966805e-06, + "loss": 0.2906, + "step": 3075 + }, + { + "epoch": 2.187377777777778, + "grad_norm": 0.7751978039741516, + "learning_rate": 4.143593580859995e-06, + "loss": 0.2647, + "step": 3076 + }, + { + "epoch": 2.1880888888888888, + "grad_norm": 0.8382031917572021, + "learning_rate": 4.136887253992027e-06, + "loss": 0.2415, + "step": 3077 + }, + { + "epoch": 2.1888, + "grad_norm": 0.9293338656425476, + "learning_rate": 4.13018494295626e-06, + "loss": 0.2726, + "step": 3078 + }, + { + "epoch": 2.189511111111111, + "grad_norm": 1.1024454832077026, + "learning_rate": 4.123486652343336e-06, + "loss": 0.2636, + "step": 3079 + }, + { + "epoch": 2.1902222222222223, + "grad_norm": 0.9387472867965698, + "learning_rate": 4.116792386741112e-06, + "loss": 0.2182, + "step": 3080 + }, + { + "epoch": 2.190933333333333, + "grad_norm": 0.8202537298202515, + "learning_rate": 4.110102150734714e-06, + "loss": 0.2814, + "step": 3081 + }, + { + "epoch": 2.1916444444444445, + "grad_norm": 0.8242006897926331, + "learning_rate": 4.103415948906498e-06, + "loss": 0.2466, + "step": 3082 + }, + { + "epoch": 2.1923555555555554, + "grad_norm": 1.0712981224060059, + "learning_rate": 4.0967337858360445e-06, + "loss": 0.2623, + "step": 3083 + }, + { + "epoch": 2.1930666666666667, + "grad_norm": 0.9024361968040466, + "learning_rate": 4.090055666100193e-06, + "loss": 0.2342, + "step": 3084 + }, + { + "epoch": 2.1937777777777776, + "grad_norm": 0.8322995901107788, + "learning_rate": 4.083381594272987e-06, + "loss": 0.2716, + "step": 3085 + }, + { + "epoch": 2.194488888888889, + "grad_norm": 1.003254771232605, + "learning_rate": 4.076711574925725e-06, + "loss": 0.2982, + "step": 3086 + }, + { + "epoch": 2.1952, + "grad_norm": 0.9898303747177124, + "learning_rate": 4.070045612626898e-06, + "loss": 0.2607, + "step": 3087 + }, + { + "epoch": 2.195911111111111, + "grad_norm": 0.9332098960876465, + "learning_rate": 4.063383711942253e-06, + "loss": 0.2619, + "step": 3088 + }, + { + "epoch": 2.196622222222222, + "grad_norm": 1.0703798532485962, + "learning_rate": 4.05672587743472e-06, + "loss": 0.2432, + "step": 3089 + }, + { + "epoch": 2.1973333333333334, + "grad_norm": 1.0419824123382568, + "learning_rate": 4.050072113664479e-06, + "loss": 0.2541, + "step": 3090 + }, + { + "epoch": 2.1980444444444442, + "grad_norm": 0.8271690011024475, + "learning_rate": 4.043422425188889e-06, + "loss": 0.2657, + "step": 3091 + }, + { + "epoch": 2.1987555555555556, + "grad_norm": 0.9126859307289124, + "learning_rate": 4.036776816562546e-06, + "loss": 0.2718, + "step": 3092 + }, + { + "epoch": 2.1994666666666665, + "grad_norm": 1.1578059196472168, + "learning_rate": 4.030135292337227e-06, + "loss": 0.2477, + "step": 3093 + }, + { + "epoch": 2.200177777777778, + "grad_norm": 0.9815757870674133, + "learning_rate": 4.023497857061929e-06, + "loss": 0.2617, + "step": 3094 + }, + { + "epoch": 2.2008888888888887, + "grad_norm": 0.7827691435813904, + "learning_rate": 4.016864515282841e-06, + "loss": 0.2367, + "step": 3095 + }, + { + "epoch": 2.2016, + "grad_norm": 0.9900698065757751, + "learning_rate": 4.010235271543349e-06, + "loss": 0.2567, + "step": 3096 + }, + { + "epoch": 2.2023111111111113, + "grad_norm": 1.0248678922653198, + "learning_rate": 4.003610130384035e-06, + "loss": 0.2564, + "step": 3097 + }, + { + "epoch": 2.203022222222222, + "grad_norm": 0.9633218050003052, + "learning_rate": 3.9969890963426615e-06, + "loss": 0.2723, + "step": 3098 + }, + { + "epoch": 2.203733333333333, + "grad_norm": 0.9647420048713684, + "learning_rate": 3.9903721739542e-06, + "loss": 0.2542, + "step": 3099 + }, + { + "epoch": 2.2044444444444444, + "grad_norm": 0.8966769576072693, + "learning_rate": 3.983759367750772e-06, + "loss": 0.2575, + "step": 3100 + }, + { + "epoch": 2.2051555555555558, + "grad_norm": 0.8884574174880981, + "learning_rate": 3.977150682261718e-06, + "loss": 0.2316, + "step": 3101 + }, + { + "epoch": 2.2058666666666666, + "grad_norm": 0.7535945773124695, + "learning_rate": 3.970546122013521e-06, + "loss": 0.2186, + "step": 3102 + }, + { + "epoch": 2.206577777777778, + "grad_norm": 0.8327714800834656, + "learning_rate": 3.9639456915298616e-06, + "loss": 0.2534, + "step": 3103 + }, + { + "epoch": 2.207288888888889, + "grad_norm": 0.8839426040649414, + "learning_rate": 3.957349395331581e-06, + "loss": 0.2705, + "step": 3104 + }, + { + "epoch": 2.208, + "grad_norm": 0.846763551235199, + "learning_rate": 3.950757237936696e-06, + "loss": 0.2944, + "step": 3105 + }, + { + "epoch": 2.208711111111111, + "grad_norm": 0.9097889065742493, + "learning_rate": 3.94416922386038e-06, + "loss": 0.2801, + "step": 3106 + }, + { + "epoch": 2.2094222222222224, + "grad_norm": 0.8919770121574402, + "learning_rate": 3.937585357614977e-06, + "loss": 0.2275, + "step": 3107 + }, + { + "epoch": 2.2101333333333333, + "grad_norm": 0.8784799575805664, + "learning_rate": 3.931005643709985e-06, + "loss": 0.2823, + "step": 3108 + }, + { + "epoch": 2.2108444444444446, + "grad_norm": 0.7169600129127502, + "learning_rate": 3.924430086652058e-06, + "loss": 0.2523, + "step": 3109 + }, + { + "epoch": 2.2115555555555555, + "grad_norm": 1.3660938739776611, + "learning_rate": 3.917858690945006e-06, + "loss": 0.2711, + "step": 3110 + }, + { + "epoch": 2.212266666666667, + "grad_norm": 0.8828915357589722, + "learning_rate": 3.911291461089784e-06, + "loss": 0.2293, + "step": 3111 + }, + { + "epoch": 2.2129777777777777, + "grad_norm": 0.8699429631233215, + "learning_rate": 3.9047284015844985e-06, + "loss": 0.2682, + "step": 3112 + }, + { + "epoch": 2.213688888888889, + "grad_norm": 0.8156993389129639, + "learning_rate": 3.8981695169243986e-06, + "loss": 0.267, + "step": 3113 + }, + { + "epoch": 2.2144, + "grad_norm": 0.8794909119606018, + "learning_rate": 3.891614811601869e-06, + "loss": 0.2545, + "step": 3114 + }, + { + "epoch": 2.2151111111111113, + "grad_norm": 0.8401433229446411, + "learning_rate": 3.885064290106438e-06, + "loss": 0.2834, + "step": 3115 + }, + { + "epoch": 2.215822222222222, + "grad_norm": 0.9932928681373596, + "learning_rate": 3.8785179569247666e-06, + "loss": 0.2354, + "step": 3116 + }, + { + "epoch": 2.2165333333333335, + "grad_norm": 0.9610915184020996, + "learning_rate": 3.871975816540645e-06, + "loss": 0.2791, + "step": 3117 + }, + { + "epoch": 2.2172444444444444, + "grad_norm": 0.8093314170837402, + "learning_rate": 3.865437873434994e-06, + "loss": 0.2528, + "step": 3118 + }, + { + "epoch": 2.2179555555555557, + "grad_norm": 0.8110005855560303, + "learning_rate": 3.858904132085857e-06, + "loss": 0.2585, + "step": 3119 + }, + { + "epoch": 2.2186666666666666, + "grad_norm": 0.764506459236145, + "learning_rate": 3.852374596968402e-06, + "loss": 0.2772, + "step": 3120 + }, + { + "epoch": 2.219377777777778, + "grad_norm": 0.986409604549408, + "learning_rate": 3.845849272554917e-06, + "loss": 0.2451, + "step": 3121 + }, + { + "epoch": 2.220088888888889, + "grad_norm": 0.8642110824584961, + "learning_rate": 3.8393281633148e-06, + "loss": 0.2469, + "step": 3122 + }, + { + "epoch": 2.2208, + "grad_norm": 1.031758427619934, + "learning_rate": 3.832811273714569e-06, + "loss": 0.2755, + "step": 3123 + }, + { + "epoch": 2.221511111111111, + "grad_norm": 0.8378773927688599, + "learning_rate": 3.8262986082178485e-06, + "loss": 0.2802, + "step": 3124 + }, + { + "epoch": 2.2222222222222223, + "grad_norm": 0.7949985861778259, + "learning_rate": 3.819790171285369e-06, + "loss": 0.2107, + "step": 3125 + }, + { + "epoch": 2.222933333333333, + "grad_norm": 0.7673428058624268, + "learning_rate": 3.8132859673749688e-06, + "loss": 0.2415, + "step": 3126 + }, + { + "epoch": 2.2236444444444445, + "grad_norm": 0.7918609380722046, + "learning_rate": 3.8067860009415747e-06, + "loss": 0.2493, + "step": 3127 + }, + { + "epoch": 2.2243555555555554, + "grad_norm": 0.8651521801948547, + "learning_rate": 3.800290276437234e-06, + "loss": 0.2383, + "step": 3128 + }, + { + "epoch": 2.2250666666666667, + "grad_norm": 0.7404313683509827, + "learning_rate": 3.7937987983110602e-06, + "loss": 0.2372, + "step": 3129 + }, + { + "epoch": 2.2257777777777776, + "grad_norm": 0.8397415280342102, + "learning_rate": 3.787311571009288e-06, + "loss": 0.281, + "step": 3130 + }, + { + "epoch": 2.226488888888889, + "grad_norm": 0.9788802266120911, + "learning_rate": 3.780828598975209e-06, + "loss": 0.2578, + "step": 3131 + }, + { + "epoch": 2.2272, + "grad_norm": 0.94599848985672, + "learning_rate": 3.77434988664923e-06, + "loss": 0.2443, + "step": 3132 + }, + { + "epoch": 2.227911111111111, + "grad_norm": 0.8825621008872986, + "learning_rate": 3.767875438468822e-06, + "loss": 0.2768, + "step": 3133 + }, + { + "epoch": 2.228622222222222, + "grad_norm": 0.7951595783233643, + "learning_rate": 3.761405258868541e-06, + "loss": 0.2797, + "step": 3134 + }, + { + "epoch": 2.2293333333333334, + "grad_norm": 1.0284886360168457, + "learning_rate": 3.7549393522800213e-06, + "loss": 0.2747, + "step": 3135 + }, + { + "epoch": 2.2300444444444443, + "grad_norm": 0.8169663548469543, + "learning_rate": 3.748477723131958e-06, + "loss": 0.2312, + "step": 3136 + }, + { + "epoch": 2.2307555555555556, + "grad_norm": 0.8403474688529968, + "learning_rate": 3.7420203758501405e-06, + "loss": 0.267, + "step": 3137 + }, + { + "epoch": 2.2314666666666665, + "grad_norm": 0.797010064125061, + "learning_rate": 3.7355673148573956e-06, + "loss": 0.2889, + "step": 3138 + }, + { + "epoch": 2.232177777777778, + "grad_norm": 0.8306525945663452, + "learning_rate": 3.7291185445736445e-06, + "loss": 0.3052, + "step": 3139 + }, + { + "epoch": 2.2328888888888887, + "grad_norm": 0.7202008962631226, + "learning_rate": 3.7226740694158413e-06, + "loss": 0.2268, + "step": 3140 + }, + { + "epoch": 2.2336, + "grad_norm": 0.9269846677780151, + "learning_rate": 3.716233893798026e-06, + "loss": 0.2587, + "step": 3141 + }, + { + "epoch": 2.234311111111111, + "grad_norm": 0.7902366518974304, + "learning_rate": 3.7097980221312645e-06, + "loss": 0.2525, + "step": 3142 + }, + { + "epoch": 2.2350222222222222, + "grad_norm": 0.8937496542930603, + "learning_rate": 3.703366458823705e-06, + "loss": 0.2791, + "step": 3143 + }, + { + "epoch": 2.235733333333333, + "grad_norm": 0.7632637023925781, + "learning_rate": 3.696939208280518e-06, + "loss": 0.2533, + "step": 3144 + }, + { + "epoch": 2.2364444444444445, + "grad_norm": 0.8601859211921692, + "learning_rate": 3.690516274903936e-06, + "loss": 0.2471, + "step": 3145 + }, + { + "epoch": 2.237155555555556, + "grad_norm": 0.8494678139686584, + "learning_rate": 3.6840976630932292e-06, + "loss": 0.2792, + "step": 3146 + }, + { + "epoch": 2.2378666666666667, + "grad_norm": 0.8214401006698608, + "learning_rate": 3.6776833772447096e-06, + "loss": 0.2629, + "step": 3147 + }, + { + "epoch": 2.2385777777777776, + "grad_norm": 0.7197944521903992, + "learning_rate": 3.6712734217517256e-06, + "loss": 0.2488, + "step": 3148 + }, + { + "epoch": 2.239288888888889, + "grad_norm": 0.9130542278289795, + "learning_rate": 3.6648678010046547e-06, + "loss": 0.2477, + "step": 3149 + }, + { + "epoch": 2.24, + "grad_norm": 0.9376085996627808, + "learning_rate": 3.658466519390921e-06, + "loss": 0.2593, + "step": 3150 + }, + { + "epoch": 2.240711111111111, + "grad_norm": 0.7309895753860474, + "learning_rate": 3.6520695812949515e-06, + "loss": 0.2418, + "step": 3151 + }, + { + "epoch": 2.2414222222222224, + "grad_norm": 1.075785517692566, + "learning_rate": 3.645676991098227e-06, + "loss": 0.2525, + "step": 3152 + }, + { + "epoch": 2.2421333333333333, + "grad_norm": 0.7741522789001465, + "learning_rate": 3.639288753179225e-06, + "loss": 0.2244, + "step": 3153 + }, + { + "epoch": 2.2428444444444446, + "grad_norm": 0.8463334441184998, + "learning_rate": 3.6329048719134564e-06, + "loss": 0.2518, + "step": 3154 + }, + { + "epoch": 2.2435555555555555, + "grad_norm": 0.9422615170478821, + "learning_rate": 3.6265253516734434e-06, + "loss": 0.2612, + "step": 3155 + }, + { + "epoch": 2.244266666666667, + "grad_norm": 0.7748128175735474, + "learning_rate": 3.6201501968287225e-06, + "loss": 0.2915, + "step": 3156 + }, + { + "epoch": 2.2449777777777777, + "grad_norm": 1.0357054471969604, + "learning_rate": 3.6137794117458404e-06, + "loss": 0.2478, + "step": 3157 + }, + { + "epoch": 2.245688888888889, + "grad_norm": 1.0638400316238403, + "learning_rate": 3.60741300078835e-06, + "loss": 0.2326, + "step": 3158 + }, + { + "epoch": 2.2464, + "grad_norm": 0.7581081986427307, + "learning_rate": 3.601050968316807e-06, + "loss": 0.2304, + "step": 3159 + }, + { + "epoch": 2.2471111111111113, + "grad_norm": 0.9062352776527405, + "learning_rate": 3.5946933186887722e-06, + "loss": 0.2801, + "step": 3160 + }, + { + "epoch": 2.247822222222222, + "grad_norm": 0.8353803753852844, + "learning_rate": 3.5883400562587987e-06, + "loss": 0.2685, + "step": 3161 + }, + { + "epoch": 2.2485333333333335, + "grad_norm": 0.942915678024292, + "learning_rate": 3.581991185378442e-06, + "loss": 0.2789, + "step": 3162 + }, + { + "epoch": 2.2492444444444444, + "grad_norm": 0.7766583561897278, + "learning_rate": 3.575646710396241e-06, + "loss": 0.2294, + "step": 3163 + }, + { + "epoch": 2.2499555555555557, + "grad_norm": 0.7703956365585327, + "learning_rate": 3.56930663565773e-06, + "loss": 0.251, + "step": 3164 + }, + { + "epoch": 2.2506666666666666, + "grad_norm": 0.8264899253845215, + "learning_rate": 3.562970965505429e-06, + "loss": 0.2457, + "step": 3165 + }, + { + "epoch": 2.251377777777778, + "grad_norm": 0.7759608030319214, + "learning_rate": 3.556639704278838e-06, + "loss": 0.2516, + "step": 3166 + }, + { + "epoch": 2.252088888888889, + "grad_norm": 0.8582772016525269, + "learning_rate": 3.5503128563144396e-06, + "loss": 0.2477, + "step": 3167 + }, + { + "epoch": 2.2528, + "grad_norm": 0.7496066093444824, + "learning_rate": 3.543990425945694e-06, + "loss": 0.2642, + "step": 3168 + }, + { + "epoch": 2.253511111111111, + "grad_norm": 0.8818457126617432, + "learning_rate": 3.537672417503032e-06, + "loss": 0.27, + "step": 3169 + }, + { + "epoch": 2.2542222222222223, + "grad_norm": 0.7885164022445679, + "learning_rate": 3.5313588353138605e-06, + "loss": 0.2338, + "step": 3170 + }, + { + "epoch": 2.2549333333333332, + "grad_norm": 0.8105672597885132, + "learning_rate": 3.525049683702553e-06, + "loss": 0.3127, + "step": 3171 + }, + { + "epoch": 2.2556444444444446, + "grad_norm": 0.8236939311027527, + "learning_rate": 3.518744966990446e-06, + "loss": 0.2714, + "step": 3172 + }, + { + "epoch": 2.2563555555555554, + "grad_norm": 0.950911283493042, + "learning_rate": 3.512444689495842e-06, + "loss": 0.2778, + "step": 3173 + }, + { + "epoch": 2.2570666666666668, + "grad_norm": 0.966240644454956, + "learning_rate": 3.5061488555339997e-06, + "loss": 0.2587, + "step": 3174 + }, + { + "epoch": 2.2577777777777777, + "grad_norm": 0.8750878572463989, + "learning_rate": 3.499857469417136e-06, + "loss": 0.2323, + "step": 3175 + }, + { + "epoch": 2.258488888888889, + "grad_norm": 0.8230910897254944, + "learning_rate": 3.4935705354544224e-06, + "loss": 0.2624, + "step": 3176 + }, + { + "epoch": 2.2592, + "grad_norm": 0.8285713195800781, + "learning_rate": 3.487288057951983e-06, + "loss": 0.2674, + "step": 3177 + }, + { + "epoch": 2.259911111111111, + "grad_norm": 1.1207103729248047, + "learning_rate": 3.4810100412128743e-06, + "loss": 0.2599, + "step": 3178 + }, + { + "epoch": 2.260622222222222, + "grad_norm": 0.8603689074516296, + "learning_rate": 3.474736489537125e-06, + "loss": 0.2282, + "step": 3179 + }, + { + "epoch": 2.2613333333333334, + "grad_norm": 0.7914976477622986, + "learning_rate": 3.468467407221676e-06, + "loss": 0.2504, + "step": 3180 + }, + { + "epoch": 2.2620444444444443, + "grad_norm": 0.8054518699645996, + "learning_rate": 3.4622027985604333e-06, + "loss": 0.2961, + "step": 3181 + }, + { + "epoch": 2.2627555555555556, + "grad_norm": 0.7858434915542603, + "learning_rate": 3.455942667844214e-06, + "loss": 0.2532, + "step": 3182 + }, + { + "epoch": 2.2634666666666665, + "grad_norm": 0.7867281436920166, + "learning_rate": 3.44968701936079e-06, + "loss": 0.267, + "step": 3183 + }, + { + "epoch": 2.264177777777778, + "grad_norm": 0.873385488986969, + "learning_rate": 3.443435857394851e-06, + "loss": 0.2503, + "step": 3184 + }, + { + "epoch": 2.2648888888888887, + "grad_norm": 1.1095027923583984, + "learning_rate": 3.4371891862280173e-06, + "loss": 0.2615, + "step": 3185 + }, + { + "epoch": 2.2656, + "grad_norm": 0.7151523232460022, + "learning_rate": 3.430947010138833e-06, + "loss": 0.2597, + "step": 3186 + }, + { + "epoch": 2.266311111111111, + "grad_norm": 0.7378624677658081, + "learning_rate": 3.424709333402756e-06, + "loss": 0.2238, + "step": 3187 + }, + { + "epoch": 2.2670222222222223, + "grad_norm": 0.9891671538352966, + "learning_rate": 3.41847616029218e-06, + "loss": 0.2483, + "step": 3188 + }, + { + "epoch": 2.267733333333333, + "grad_norm": 0.8297561407089233, + "learning_rate": 3.412247495076392e-06, + "loss": 0.2723, + "step": 3189 + }, + { + "epoch": 2.2684444444444445, + "grad_norm": 0.7724572420120239, + "learning_rate": 3.4060233420216136e-06, + "loss": 0.2736, + "step": 3190 + }, + { + "epoch": 2.2691555555555554, + "grad_norm": 0.7192564010620117, + "learning_rate": 3.399803705390955e-06, + "loss": 0.236, + "step": 3191 + }, + { + "epoch": 2.2698666666666667, + "grad_norm": 0.7411932349205017, + "learning_rate": 3.393588589444453e-06, + "loss": 0.228, + "step": 3192 + }, + { + "epoch": 2.2705777777777776, + "grad_norm": 0.802090048789978, + "learning_rate": 3.3873779984390277e-06, + "loss": 0.2573, + "step": 3193 + }, + { + "epoch": 2.271288888888889, + "grad_norm": 0.9224371314048767, + "learning_rate": 3.381171936628521e-06, + "loss": 0.2604, + "step": 3194 + }, + { + "epoch": 2.2720000000000002, + "grad_norm": 0.7495435476303101, + "learning_rate": 3.3749704082636493e-06, + "loss": 0.2601, + "step": 3195 + }, + { + "epoch": 2.272711111111111, + "grad_norm": 0.6880185604095459, + "learning_rate": 3.3687734175920505e-06, + "loss": 0.2176, + "step": 3196 + }, + { + "epoch": 2.273422222222222, + "grad_norm": 1.1581192016601562, + "learning_rate": 3.362580968858229e-06, + "loss": 0.2679, + "step": 3197 + }, + { + "epoch": 2.2741333333333333, + "grad_norm": 0.8019190430641174, + "learning_rate": 3.356393066303595e-06, + "loss": 0.2657, + "step": 3198 + }, + { + "epoch": 2.2748444444444447, + "grad_norm": 0.9648621082305908, + "learning_rate": 3.350209714166437e-06, + "loss": 0.2677, + "step": 3199 + }, + { + "epoch": 2.2755555555555556, + "grad_norm": 1.003495693206787, + "learning_rate": 3.3440309166819284e-06, + "loss": 0.2758, + "step": 3200 + }, + { + "epoch": 2.2762666666666664, + "grad_norm": 0.7539083361625671, + "learning_rate": 3.337856678082133e-06, + "loss": 0.267, + "step": 3201 + }, + { + "epoch": 2.2769777777777778, + "grad_norm": 0.816024124622345, + "learning_rate": 3.3316870025959693e-06, + "loss": 0.2772, + "step": 3202 + }, + { + "epoch": 2.277688888888889, + "grad_norm": 0.9146069884300232, + "learning_rate": 3.3255218944492584e-06, + "loss": 0.2594, + "step": 3203 + }, + { + "epoch": 2.2784, + "grad_norm": 0.8772939443588257, + "learning_rate": 3.3193613578646633e-06, + "loss": 0.2115, + "step": 3204 + }, + { + "epoch": 2.279111111111111, + "grad_norm": 1.1568586826324463, + "learning_rate": 3.3132053970617473e-06, + "loss": 0.2779, + "step": 3205 + }, + { + "epoch": 2.279822222222222, + "grad_norm": 0.874474048614502, + "learning_rate": 3.307054016256912e-06, + "loss": 0.2725, + "step": 3206 + }, + { + "epoch": 2.2805333333333335, + "grad_norm": 0.7519720792770386, + "learning_rate": 3.3009072196634373e-06, + "loss": 0.2744, + "step": 3207 + }, + { + "epoch": 2.2812444444444444, + "grad_norm": 1.579458236694336, + "learning_rate": 3.2947650114914587e-06, + "loss": 0.2547, + "step": 3208 + }, + { + "epoch": 2.2819555555555557, + "grad_norm": 0.9574062824249268, + "learning_rate": 3.288627395947971e-06, + "loss": 0.2461, + "step": 3209 + }, + { + "epoch": 2.2826666666666666, + "grad_norm": 0.8056530356407166, + "learning_rate": 3.2824943772368213e-06, + "loss": 0.2599, + "step": 3210 + }, + { + "epoch": 2.283377777777778, + "grad_norm": 0.827707052230835, + "learning_rate": 3.2763659595587084e-06, + "loss": 0.2827, + "step": 3211 + }, + { + "epoch": 2.284088888888889, + "grad_norm": 1.1128686666488647, + "learning_rate": 3.270242147111182e-06, + "loss": 0.269, + "step": 3212 + }, + { + "epoch": 2.2848, + "grad_norm": 0.7721088528633118, + "learning_rate": 3.264122944088635e-06, + "loss": 0.2648, + "step": 3213 + }, + { + "epoch": 2.285511111111111, + "grad_norm": 1.0476008653640747, + "learning_rate": 3.258008354682303e-06, + "loss": 0.2612, + "step": 3214 + }, + { + "epoch": 2.2862222222222224, + "grad_norm": 0.7364855408668518, + "learning_rate": 3.2518983830802643e-06, + "loss": 0.2632, + "step": 3215 + }, + { + "epoch": 2.2869333333333333, + "grad_norm": 0.76447594165802, + "learning_rate": 3.2457930334674304e-06, + "loss": 0.2888, + "step": 3216 + }, + { + "epoch": 2.2876444444444446, + "grad_norm": 0.7832509279251099, + "learning_rate": 3.2396923100255515e-06, + "loss": 0.2749, + "step": 3217 + }, + { + "epoch": 2.2883555555555555, + "grad_norm": 0.792245626449585, + "learning_rate": 3.233596216933206e-06, + "loss": 0.2705, + "step": 3218 + }, + { + "epoch": 2.289066666666667, + "grad_norm": 0.8710457682609558, + "learning_rate": 3.2275047583658015e-06, + "loss": 0.2381, + "step": 3219 + }, + { + "epoch": 2.2897777777777777, + "grad_norm": 0.8367807865142822, + "learning_rate": 3.2214179384955713e-06, + "loss": 0.2824, + "step": 3220 + }, + { + "epoch": 2.290488888888889, + "grad_norm": 0.8836548924446106, + "learning_rate": 3.2153357614915726e-06, + "loss": 0.2397, + "step": 3221 + }, + { + "epoch": 2.2912, + "grad_norm": 0.8373388648033142, + "learning_rate": 3.209258231519682e-06, + "loss": 0.2477, + "step": 3222 + }, + { + "epoch": 2.2919111111111112, + "grad_norm": 0.9489513635635376, + "learning_rate": 3.2031853527425903e-06, + "loss": 0.2386, + "step": 3223 + }, + { + "epoch": 2.292622222222222, + "grad_norm": 1.0359413623809814, + "learning_rate": 3.197117129319808e-06, + "loss": 0.226, + "step": 3224 + }, + { + "epoch": 2.2933333333333334, + "grad_norm": 0.7910504341125488, + "learning_rate": 3.1910535654076535e-06, + "loss": 0.2472, + "step": 3225 + }, + { + "epoch": 2.2940444444444443, + "grad_norm": 1.1991521120071411, + "learning_rate": 3.1849946651592532e-06, + "loss": 0.2589, + "step": 3226 + }, + { + "epoch": 2.2947555555555557, + "grad_norm": 1.1095600128173828, + "learning_rate": 3.178940432724541e-06, + "loss": 0.2472, + "step": 3227 + }, + { + "epoch": 2.2954666666666665, + "grad_norm": 1.0395678281784058, + "learning_rate": 3.172890872250254e-06, + "loss": 0.2921, + "step": 3228 + }, + { + "epoch": 2.296177777777778, + "grad_norm": 0.8721950650215149, + "learning_rate": 3.1668459878799274e-06, + "loss": 0.2457, + "step": 3229 + }, + { + "epoch": 2.2968888888888888, + "grad_norm": 1.0965291261672974, + "learning_rate": 3.1608057837538976e-06, + "loss": 0.2666, + "step": 3230 + }, + { + "epoch": 2.2976, + "grad_norm": 1.2514506578445435, + "learning_rate": 3.1547702640092835e-06, + "loss": 0.2473, + "step": 3231 + }, + { + "epoch": 2.298311111111111, + "grad_norm": 0.7312183976173401, + "learning_rate": 3.1487394327800156e-06, + "loss": 0.2262, + "step": 3232 + }, + { + "epoch": 2.2990222222222223, + "grad_norm": 0.784202516078949, + "learning_rate": 3.142713294196791e-06, + "loss": 0.2351, + "step": 3233 + }, + { + "epoch": 2.299733333333333, + "grad_norm": 0.7031396627426147, + "learning_rate": 3.136691852387116e-06, + "loss": 0.2195, + "step": 3234 + }, + { + "epoch": 2.3004444444444445, + "grad_norm": 0.8046204447746277, + "learning_rate": 3.1306751114752532e-06, + "loss": 0.2433, + "step": 3235 + }, + { + "epoch": 2.3011555555555554, + "grad_norm": 1.0289037227630615, + "learning_rate": 3.1246630755822703e-06, + "loss": 0.2707, + "step": 3236 + }, + { + "epoch": 2.3018666666666667, + "grad_norm": 0.9371544122695923, + "learning_rate": 3.1186557488259985e-06, + "loss": 0.304, + "step": 3237 + }, + { + "epoch": 2.3025777777777776, + "grad_norm": 0.7565621733665466, + "learning_rate": 3.1126531353210456e-06, + "loss": 0.2727, + "step": 3238 + }, + { + "epoch": 2.303288888888889, + "grad_norm": 0.7751135230064392, + "learning_rate": 3.1066552391787974e-06, + "loss": 0.2627, + "step": 3239 + }, + { + "epoch": 2.304, + "grad_norm": 1.2136616706848145, + "learning_rate": 3.1006620645073925e-06, + "loss": 0.2758, + "step": 3240 + }, + { + "epoch": 2.304711111111111, + "grad_norm": 0.9637894034385681, + "learning_rate": 3.094673615411761e-06, + "loss": 0.2353, + "step": 3241 + }, + { + "epoch": 2.305422222222222, + "grad_norm": 0.735862672328949, + "learning_rate": 3.0886898959935663e-06, + "loss": 0.2596, + "step": 3242 + }, + { + "epoch": 2.3061333333333334, + "grad_norm": 0.9959772229194641, + "learning_rate": 3.0827109103512643e-06, + "loss": 0.26, + "step": 3243 + }, + { + "epoch": 2.3068444444444443, + "grad_norm": 1.185125708580017, + "learning_rate": 3.0767366625800366e-06, + "loss": 0.2805, + "step": 3244 + }, + { + "epoch": 2.3075555555555556, + "grad_norm": 1.0286190509796143, + "learning_rate": 3.070767156771849e-06, + "loss": 0.2562, + "step": 3245 + }, + { + "epoch": 2.3082666666666665, + "grad_norm": 0.9136818051338196, + "learning_rate": 3.064802397015394e-06, + "loss": 0.2581, + "step": 3246 + }, + { + "epoch": 2.308977777777778, + "grad_norm": 1.1909195184707642, + "learning_rate": 3.058842387396136e-06, + "loss": 0.2679, + "step": 3247 + }, + { + "epoch": 2.309688888888889, + "grad_norm": 1.0981003046035767, + "learning_rate": 3.052887131996267e-06, + "loss": 0.2842, + "step": 3248 + }, + { + "epoch": 2.3104, + "grad_norm": 1.1652390956878662, + "learning_rate": 3.0469366348947337e-06, + "loss": 0.2581, + "step": 3249 + }, + { + "epoch": 2.311111111111111, + "grad_norm": 0.8155806064605713, + "learning_rate": 3.040990900167219e-06, + "loss": 0.2393, + "step": 3250 + }, + { + "epoch": 2.3118222222222222, + "grad_norm": 1.0329827070236206, + "learning_rate": 3.035049931886147e-06, + "loss": 0.2563, + "step": 3251 + }, + { + "epoch": 2.3125333333333336, + "grad_norm": 1.264308214187622, + "learning_rate": 3.0291137341206755e-06, + "loss": 0.2765, + "step": 3252 + }, + { + "epoch": 2.3132444444444444, + "grad_norm": 1.3689857721328735, + "learning_rate": 3.023182310936692e-06, + "loss": 0.2431, + "step": 3253 + }, + { + "epoch": 2.3139555555555553, + "grad_norm": 1.3724125623703003, + "learning_rate": 3.0172556663968254e-06, + "loss": 0.2851, + "step": 3254 + }, + { + "epoch": 2.3146666666666667, + "grad_norm": 0.8553822040557861, + "learning_rate": 3.0113338045604102e-06, + "loss": 0.2165, + "step": 3255 + }, + { + "epoch": 2.315377777777778, + "grad_norm": 1.1060270071029663, + "learning_rate": 3.0054167294835314e-06, + "loss": 0.2796, + "step": 3256 + }, + { + "epoch": 2.316088888888889, + "grad_norm": 1.4455970525741577, + "learning_rate": 2.999504445218967e-06, + "loss": 0.2357, + "step": 3257 + }, + { + "epoch": 2.3168, + "grad_norm": 1.568426251411438, + "learning_rate": 2.993596955816244e-06, + "loss": 0.2563, + "step": 3258 + }, + { + "epoch": 2.317511111111111, + "grad_norm": 1.1558290719985962, + "learning_rate": 2.987694265321578e-06, + "loss": 0.2482, + "step": 3259 + }, + { + "epoch": 2.3182222222222224, + "grad_norm": 0.7898767590522766, + "learning_rate": 2.9817963777779124e-06, + "loss": 0.2502, + "step": 3260 + }, + { + "epoch": 2.3189333333333333, + "grad_norm": 1.456722617149353, + "learning_rate": 2.9759032972248993e-06, + "loss": 0.2588, + "step": 3261 + }, + { + "epoch": 2.3196444444444446, + "grad_norm": 1.6885344982147217, + "learning_rate": 2.970015027698895e-06, + "loss": 0.252, + "step": 3262 + }, + { + "epoch": 2.3203555555555555, + "grad_norm": 1.4213353395462036, + "learning_rate": 2.9641315732329635e-06, + "loss": 0.2945, + "step": 3263 + }, + { + "epoch": 2.321066666666667, + "grad_norm": 0.9834480285644531, + "learning_rate": 2.958252937856869e-06, + "loss": 0.2571, + "step": 3264 + }, + { + "epoch": 2.3217777777777777, + "grad_norm": 0.9098924994468689, + "learning_rate": 2.9523791255970768e-06, + "loss": 0.2797, + "step": 3265 + }, + { + "epoch": 2.322488888888889, + "grad_norm": 1.3113248348236084, + "learning_rate": 2.946510140476747e-06, + "loss": 0.2383, + "step": 3266 + }, + { + "epoch": 2.3232, + "grad_norm": 1.1416916847229004, + "learning_rate": 2.9406459865157334e-06, + "loss": 0.2501, + "step": 3267 + }, + { + "epoch": 2.3239111111111113, + "grad_norm": 0.999570906162262, + "learning_rate": 2.9347866677305814e-06, + "loss": 0.2407, + "step": 3268 + }, + { + "epoch": 2.324622222222222, + "grad_norm": 0.8495328426361084, + "learning_rate": 2.9289321881345257e-06, + "loss": 0.216, + "step": 3269 + }, + { + "epoch": 2.3253333333333335, + "grad_norm": 1.1027122735977173, + "learning_rate": 2.923082551737484e-06, + "loss": 0.2674, + "step": 3270 + }, + { + "epoch": 2.3260444444444444, + "grad_norm": 1.230595350265503, + "learning_rate": 2.917237762546059e-06, + "loss": 0.2597, + "step": 3271 + }, + { + "epoch": 2.3267555555555557, + "grad_norm": 0.8269692063331604, + "learning_rate": 2.911397824563533e-06, + "loss": 0.2757, + "step": 3272 + }, + { + "epoch": 2.3274666666666666, + "grad_norm": 0.9271746277809143, + "learning_rate": 2.9055627417898633e-06, + "loss": 0.2668, + "step": 3273 + }, + { + "epoch": 2.328177777777778, + "grad_norm": 0.8485908508300781, + "learning_rate": 2.899732518221685e-06, + "loss": 0.2393, + "step": 3274 + }, + { + "epoch": 2.328888888888889, + "grad_norm": 0.7049943208694458, + "learning_rate": 2.8939071578523036e-06, + "loss": 0.2171, + "step": 3275 + }, + { + "epoch": 2.3296, + "grad_norm": 0.7197278141975403, + "learning_rate": 2.888086664671693e-06, + "loss": 0.2709, + "step": 3276 + }, + { + "epoch": 2.330311111111111, + "grad_norm": 0.7895232439041138, + "learning_rate": 2.8822710426664935e-06, + "loss": 0.2377, + "step": 3277 + }, + { + "epoch": 2.3310222222222223, + "grad_norm": 1.3004909753799438, + "learning_rate": 2.8764602958200096e-06, + "loss": 0.2892, + "step": 3278 + }, + { + "epoch": 2.331733333333333, + "grad_norm": 0.7833949327468872, + "learning_rate": 2.870654428112206e-06, + "loss": 0.2394, + "step": 3279 + }, + { + "epoch": 2.3324444444444445, + "grad_norm": 0.7928273677825928, + "learning_rate": 2.8648534435197086e-06, + "loss": 0.2241, + "step": 3280 + }, + { + "epoch": 2.3331555555555554, + "grad_norm": 1.1372963190078735, + "learning_rate": 2.8590573460157954e-06, + "loss": 0.2606, + "step": 3281 + }, + { + "epoch": 2.3338666666666668, + "grad_norm": 0.9634580612182617, + "learning_rate": 2.853266139570391e-06, + "loss": 0.2404, + "step": 3282 + }, + { + "epoch": 2.3345777777777776, + "grad_norm": 0.8429052829742432, + "learning_rate": 2.847479828150087e-06, + "loss": 0.2505, + "step": 3283 + }, + { + "epoch": 2.335288888888889, + "grad_norm": 0.924619197845459, + "learning_rate": 2.841698415718103e-06, + "loss": 0.2565, + "step": 3284 + }, + { + "epoch": 2.336, + "grad_norm": 1.3895686864852905, + "learning_rate": 2.83592190623432e-06, + "loss": 0.2549, + "step": 3285 + }, + { + "epoch": 2.336711111111111, + "grad_norm": 1.2345738410949707, + "learning_rate": 2.8301503036552446e-06, + "loss": 0.2668, + "step": 3286 + }, + { + "epoch": 2.337422222222222, + "grad_norm": 0.93150794506073, + "learning_rate": 2.824383611934037e-06, + "loss": 0.2619, + "step": 3287 + }, + { + "epoch": 2.3381333333333334, + "grad_norm": 0.8868767619132996, + "learning_rate": 2.8186218350204865e-06, + "loss": 0.2574, + "step": 3288 + }, + { + "epoch": 2.3388444444444443, + "grad_norm": 1.0643558502197266, + "learning_rate": 2.812864976861016e-06, + "loss": 0.2113, + "step": 3289 + }, + { + "epoch": 2.3395555555555556, + "grad_norm": 0.8496482968330383, + "learning_rate": 2.8071130413986814e-06, + "loss": 0.2517, + "step": 3290 + }, + { + "epoch": 2.3402666666666665, + "grad_norm": 1.7619436979293823, + "learning_rate": 2.801366032573165e-06, + "loss": 0.2435, + "step": 3291 + }, + { + "epoch": 2.340977777777778, + "grad_norm": 1.0897668600082397, + "learning_rate": 2.795623954320781e-06, + "loss": 0.2581, + "step": 3292 + }, + { + "epoch": 2.3416888888888887, + "grad_norm": 0.7200884222984314, + "learning_rate": 2.7898868105744492e-06, + "loss": 0.2595, + "step": 3293 + }, + { + "epoch": 2.3424, + "grad_norm": 1.0589113235473633, + "learning_rate": 2.7841546052637346e-06, + "loss": 0.2747, + "step": 3294 + }, + { + "epoch": 2.343111111111111, + "grad_norm": 1.0766690969467163, + "learning_rate": 2.7784273423147966e-06, + "loss": 0.2983, + "step": 3295 + }, + { + "epoch": 2.3438222222222223, + "grad_norm": 0.9183441996574402, + "learning_rate": 2.7727050256504295e-06, + "loss": 0.2593, + "step": 3296 + }, + { + "epoch": 2.3445333333333336, + "grad_norm": 0.8860029578208923, + "learning_rate": 2.7669876591900193e-06, + "loss": 0.2384, + "step": 3297 + }, + { + "epoch": 2.3452444444444445, + "grad_norm": 0.7585933804512024, + "learning_rate": 2.761275246849582e-06, + "loss": 0.2205, + "step": 3298 + }, + { + "epoch": 2.3459555555555553, + "grad_norm": 1.1659044027328491, + "learning_rate": 2.7555677925417203e-06, + "loss": 0.2817, + "step": 3299 + }, + { + "epoch": 2.3466666666666667, + "grad_norm": 0.7577102184295654, + "learning_rate": 2.7498653001756615e-06, + "loss": 0.2271, + "step": 3300 + }, + { + "epoch": 2.347377777777778, + "grad_norm": 0.8996542096138, + "learning_rate": 2.744167773657216e-06, + "loss": 0.2462, + "step": 3301 + }, + { + "epoch": 2.348088888888889, + "grad_norm": 0.8135519623756409, + "learning_rate": 2.738475216888802e-06, + "loss": 0.2414, + "step": 3302 + }, + { + "epoch": 2.3487999999999998, + "grad_norm": 0.925624430179596, + "learning_rate": 2.732787633769435e-06, + "loss": 0.2674, + "step": 3303 + }, + { + "epoch": 2.349511111111111, + "grad_norm": 0.8324899673461914, + "learning_rate": 2.7271050281947165e-06, + "loss": 0.288, + "step": 3304 + }, + { + "epoch": 2.3502222222222224, + "grad_norm": 1.013461709022522, + "learning_rate": 2.7214274040568535e-06, + "loss": 0.2448, + "step": 3305 + }, + { + "epoch": 2.3509333333333333, + "grad_norm": 0.9989730715751648, + "learning_rate": 2.7157547652446193e-06, + "loss": 0.2541, + "step": 3306 + }, + { + "epoch": 2.351644444444444, + "grad_norm": 0.8985846638679504, + "learning_rate": 2.710087115643395e-06, + "loss": 0.2694, + "step": 3307 + }, + { + "epoch": 2.3523555555555555, + "grad_norm": 1.0478732585906982, + "learning_rate": 2.704424459135123e-06, + "loss": 0.2313, + "step": 3308 + }, + { + "epoch": 2.353066666666667, + "grad_norm": 1.2894492149353027, + "learning_rate": 2.698766799598349e-06, + "loss": 0.2631, + "step": 3309 + }, + { + "epoch": 2.3537777777777777, + "grad_norm": 0.9589313268661499, + "learning_rate": 2.6931141409081753e-06, + "loss": 0.2708, + "step": 3310 + }, + { + "epoch": 2.354488888888889, + "grad_norm": 0.9271663427352905, + "learning_rate": 2.687466486936289e-06, + "loss": 0.2563, + "step": 3311 + }, + { + "epoch": 2.3552, + "grad_norm": 1.0242267847061157, + "learning_rate": 2.681823841550947e-06, + "loss": 0.2553, + "step": 3312 + }, + { + "epoch": 2.3559111111111113, + "grad_norm": 1.1025840044021606, + "learning_rate": 2.6761862086169786e-06, + "loss": 0.2657, + "step": 3313 + }, + { + "epoch": 2.356622222222222, + "grad_norm": 0.7767345309257507, + "learning_rate": 2.6705535919957772e-06, + "loss": 0.2249, + "step": 3314 + }, + { + "epoch": 2.3573333333333335, + "grad_norm": 1.0700736045837402, + "learning_rate": 2.6649259955453e-06, + "loss": 0.2196, + "step": 3315 + }, + { + "epoch": 2.3580444444444444, + "grad_norm": 0.8243629932403564, + "learning_rate": 2.6593034231200664e-06, + "loss": 0.2974, + "step": 3316 + }, + { + "epoch": 2.3587555555555557, + "grad_norm": 0.860111653804779, + "learning_rate": 2.653685878571157e-06, + "loss": 0.27, + "step": 3317 + }, + { + "epoch": 2.3594666666666666, + "grad_norm": 1.04812753200531, + "learning_rate": 2.648073365746204e-06, + "loss": 0.2745, + "step": 3318 + }, + { + "epoch": 2.360177777777778, + "grad_norm": 0.7864211797714233, + "learning_rate": 2.642465888489397e-06, + "loss": 0.251, + "step": 3319 + }, + { + "epoch": 2.360888888888889, + "grad_norm": 0.7566313147544861, + "learning_rate": 2.6368634506414757e-06, + "loss": 0.2582, + "step": 3320 + }, + { + "epoch": 2.3616, + "grad_norm": 0.937096357345581, + "learning_rate": 2.6312660560397273e-06, + "loss": 0.2423, + "step": 3321 + }, + { + "epoch": 2.362311111111111, + "grad_norm": 0.9344249367713928, + "learning_rate": 2.6256737085179852e-06, + "loss": 0.2364, + "step": 3322 + }, + { + "epoch": 2.3630222222222224, + "grad_norm": 0.8246772289276123, + "learning_rate": 2.620086411906626e-06, + "loss": 0.2248, + "step": 3323 + }, + { + "epoch": 2.3637333333333332, + "grad_norm": 0.8359557390213013, + "learning_rate": 2.614504170032567e-06, + "loss": 0.2744, + "step": 3324 + }, + { + "epoch": 2.3644444444444446, + "grad_norm": 0.7598713636398315, + "learning_rate": 2.6089269867192622e-06, + "loss": 0.2296, + "step": 3325 + }, + { + "epoch": 2.3651555555555555, + "grad_norm": 0.8006106019020081, + "learning_rate": 2.6033548657867013e-06, + "loss": 0.2687, + "step": 3326 + }, + { + "epoch": 2.365866666666667, + "grad_norm": 0.7429032325744629, + "learning_rate": 2.597787811051409e-06, + "loss": 0.2416, + "step": 3327 + }, + { + "epoch": 2.3665777777777777, + "grad_norm": 0.8860127925872803, + "learning_rate": 2.5922258263264366e-06, + "loss": 0.2199, + "step": 3328 + }, + { + "epoch": 2.367288888888889, + "grad_norm": 0.7257380485534668, + "learning_rate": 2.5866689154213643e-06, + "loss": 0.2455, + "step": 3329 + }, + { + "epoch": 2.368, + "grad_norm": 0.797292947769165, + "learning_rate": 2.581117082142296e-06, + "loss": 0.2434, + "step": 3330 + }, + { + "epoch": 2.368711111111111, + "grad_norm": 0.8133886456489563, + "learning_rate": 2.575570330291861e-06, + "loss": 0.2768, + "step": 3331 + }, + { + "epoch": 2.369422222222222, + "grad_norm": 0.7348941564559937, + "learning_rate": 2.570028663669204e-06, + "loss": 0.2731, + "step": 3332 + }, + { + "epoch": 2.3701333333333334, + "grad_norm": 1.2016780376434326, + "learning_rate": 2.5644920860699883e-06, + "loss": 0.2882, + "step": 3333 + }, + { + "epoch": 2.3708444444444443, + "grad_norm": 1.0127344131469727, + "learning_rate": 2.5589606012863968e-06, + "loss": 0.2987, + "step": 3334 + }, + { + "epoch": 2.3715555555555556, + "grad_norm": 0.8229749202728271, + "learning_rate": 2.5534342131071076e-06, + "loss": 0.2772, + "step": 3335 + }, + { + "epoch": 2.3722666666666665, + "grad_norm": 0.8499463200569153, + "learning_rate": 2.547912925317334e-06, + "loss": 0.247, + "step": 3336 + }, + { + "epoch": 2.372977777777778, + "grad_norm": 0.7293373942375183, + "learning_rate": 2.5423967416987682e-06, + "loss": 0.2381, + "step": 3337 + }, + { + "epoch": 2.3736888888888887, + "grad_norm": 0.8224334716796875, + "learning_rate": 2.5368856660296327e-06, + "loss": 0.3194, + "step": 3338 + }, + { + "epoch": 2.3744, + "grad_norm": 0.8041974902153015, + "learning_rate": 2.5313797020846244e-06, + "loss": 0.2217, + "step": 3339 + }, + { + "epoch": 2.375111111111111, + "grad_norm": 0.8642458915710449, + "learning_rate": 2.5258788536349622e-06, + "loss": 0.2573, + "step": 3340 + }, + { + "epoch": 2.3758222222222223, + "grad_norm": 0.9705987572669983, + "learning_rate": 2.5203831244483513e-06, + "loss": 0.2226, + "step": 3341 + }, + { + "epoch": 2.376533333333333, + "grad_norm": 0.9231569170951843, + "learning_rate": 2.514892518288988e-06, + "loss": 0.2982, + "step": 3342 + }, + { + "epoch": 2.3772444444444445, + "grad_norm": 0.7860121130943298, + "learning_rate": 2.5094070389175696e-06, + "loss": 0.2476, + "step": 3343 + }, + { + "epoch": 2.3779555555555554, + "grad_norm": 1.2037911415100098, + "learning_rate": 2.503926690091263e-06, + "loss": 0.2656, + "step": 3344 + }, + { + "epoch": 2.3786666666666667, + "grad_norm": 0.905065655708313, + "learning_rate": 2.4984514755637467e-06, + "loss": 0.2489, + "step": 3345 + }, + { + "epoch": 2.3793777777777776, + "grad_norm": 0.8085951209068298, + "learning_rate": 2.492981399085157e-06, + "loss": 0.269, + "step": 3346 + }, + { + "epoch": 2.380088888888889, + "grad_norm": 1.0715622901916504, + "learning_rate": 2.4875164644021343e-06, + "loss": 0.246, + "step": 3347 + }, + { + "epoch": 2.3808, + "grad_norm": 0.845039427280426, + "learning_rate": 2.482056675257776e-06, + "loss": 0.2722, + "step": 3348 + }, + { + "epoch": 2.381511111111111, + "grad_norm": 0.842200756072998, + "learning_rate": 2.4766020353916744e-06, + "loss": 0.2831, + "step": 3349 + }, + { + "epoch": 2.3822222222222225, + "grad_norm": 0.7901564836502075, + "learning_rate": 2.471152548539876e-06, + "loss": 0.2461, + "step": 3350 + }, + { + "epoch": 2.3829333333333333, + "grad_norm": 0.9355416893959045, + "learning_rate": 2.465708218434918e-06, + "loss": 0.2941, + "step": 3351 + }, + { + "epoch": 2.3836444444444442, + "grad_norm": 0.7924145460128784, + "learning_rate": 2.4602690488057836e-06, + "loss": 0.2771, + "step": 3352 + }, + { + "epoch": 2.3843555555555556, + "grad_norm": 0.788975179195404, + "learning_rate": 2.4548350433779443e-06, + "loss": 0.2331, + "step": 3353 + }, + { + "epoch": 2.385066666666667, + "grad_norm": 0.838315486907959, + "learning_rate": 2.4494062058733157e-06, + "loss": 0.2202, + "step": 3354 + }, + { + "epoch": 2.3857777777777778, + "grad_norm": 0.7680338621139526, + "learning_rate": 2.4439825400102834e-06, + "loss": 0.243, + "step": 3355 + }, + { + "epoch": 2.3864888888888887, + "grad_norm": 1.1031882762908936, + "learning_rate": 2.438564049503688e-06, + "loss": 0.3309, + "step": 3356 + }, + { + "epoch": 2.3872, + "grad_norm": 0.9838505387306213, + "learning_rate": 2.4331507380648222e-06, + "loss": 0.2483, + "step": 3357 + }, + { + "epoch": 2.3879111111111113, + "grad_norm": 0.7379828691482544, + "learning_rate": 2.4277426094014457e-06, + "loss": 0.2136, + "step": 3358 + }, + { + "epoch": 2.388622222222222, + "grad_norm": 0.7614619135856628, + "learning_rate": 2.4223396672177435e-06, + "loss": 0.2294, + "step": 3359 + }, + { + "epoch": 2.389333333333333, + "grad_norm": 0.8441638946533203, + "learning_rate": 2.416941915214377e-06, + "loss": 0.2421, + "step": 3360 + }, + { + "epoch": 2.3900444444444444, + "grad_norm": 0.8255077004432678, + "learning_rate": 2.4115493570884242e-06, + "loss": 0.2462, + "step": 3361 + }, + { + "epoch": 2.3907555555555557, + "grad_norm": 1.2116557359695435, + "learning_rate": 2.4061619965334314e-06, + "loss": 0.2599, + "step": 3362 + }, + { + "epoch": 2.3914666666666666, + "grad_norm": 0.8518403768539429, + "learning_rate": 2.4007798372393643e-06, + "loss": 0.251, + "step": 3363 + }, + { + "epoch": 2.392177777777778, + "grad_norm": 0.9359354376792908, + "learning_rate": 2.395402882892639e-06, + "loss": 0.3044, + "step": 3364 + }, + { + "epoch": 2.392888888888889, + "grad_norm": 0.951191246509552, + "learning_rate": 2.390031137176101e-06, + "loss": 0.254, + "step": 3365 + }, + { + "epoch": 2.3936, + "grad_norm": 0.7701266407966614, + "learning_rate": 2.3846646037690304e-06, + "loss": 0.2363, + "step": 3366 + }, + { + "epoch": 2.394311111111111, + "grad_norm": 0.8670124411582947, + "learning_rate": 2.3793032863471354e-06, + "loss": 0.1983, + "step": 3367 + }, + { + "epoch": 2.3950222222222224, + "grad_norm": 1.036278486251831, + "learning_rate": 2.3739471885825536e-06, + "loss": 0.2727, + "step": 3368 + }, + { + "epoch": 2.3957333333333333, + "grad_norm": 0.8210153579711914, + "learning_rate": 2.3685963141438474e-06, + "loss": 0.2564, + "step": 3369 + }, + { + "epoch": 2.3964444444444446, + "grad_norm": 0.9106695652008057, + "learning_rate": 2.363250666695999e-06, + "loss": 0.2534, + "step": 3370 + }, + { + "epoch": 2.3971555555555555, + "grad_norm": 0.8458957672119141, + "learning_rate": 2.3579102499004127e-06, + "loss": 0.2281, + "step": 3371 + }, + { + "epoch": 2.397866666666667, + "grad_norm": 0.8055317997932434, + "learning_rate": 2.3525750674149094e-06, + "loss": 0.2546, + "step": 3372 + }, + { + "epoch": 2.3985777777777777, + "grad_norm": 0.8271927237510681, + "learning_rate": 2.3472451228937254e-06, + "loss": 0.278, + "step": 3373 + }, + { + "epoch": 2.399288888888889, + "grad_norm": 0.729041576385498, + "learning_rate": 2.34192041998751e-06, + "loss": 0.2602, + "step": 3374 + }, + { + "epoch": 2.4, + "grad_norm": 0.7766134738922119, + "learning_rate": 2.3366009623433195e-06, + "loss": 0.2221, + "step": 3375 + }, + { + "epoch": 2.4007111111111112, + "grad_norm": 0.9583573341369629, + "learning_rate": 2.331286753604621e-06, + "loss": 0.2531, + "step": 3376 + }, + { + "epoch": 2.401422222222222, + "grad_norm": 0.8477525115013123, + "learning_rate": 2.325977797411284e-06, + "loss": 0.2377, + "step": 3377 + }, + { + "epoch": 2.4021333333333335, + "grad_norm": 0.7192270159721375, + "learning_rate": 2.3206740973995823e-06, + "loss": 0.2066, + "step": 3378 + }, + { + "epoch": 2.4028444444444443, + "grad_norm": 0.7586358189582825, + "learning_rate": 2.315375657202188e-06, + "loss": 0.2604, + "step": 3379 + }, + { + "epoch": 2.4035555555555557, + "grad_norm": 0.9883622527122498, + "learning_rate": 2.3100824804481703e-06, + "loss": 0.2513, + "step": 3380 + }, + { + "epoch": 2.4042666666666666, + "grad_norm": 0.7067659497261047, + "learning_rate": 2.3047945707629958e-06, + "loss": 0.2558, + "step": 3381 + }, + { + "epoch": 2.404977777777778, + "grad_norm": 0.8114823698997498, + "learning_rate": 2.29951193176852e-06, + "loss": 0.2829, + "step": 3382 + }, + { + "epoch": 2.4056888888888888, + "grad_norm": 1.2475537061691284, + "learning_rate": 2.294234567082991e-06, + "loss": 0.2249, + "step": 3383 + }, + { + "epoch": 2.4064, + "grad_norm": 0.7526620626449585, + "learning_rate": 2.2889624803210453e-06, + "loss": 0.2512, + "step": 3384 + }, + { + "epoch": 2.407111111111111, + "grad_norm": 0.8872572779655457, + "learning_rate": 2.283695675093699e-06, + "loss": 0.2277, + "step": 3385 + }, + { + "epoch": 2.4078222222222223, + "grad_norm": 0.7882347702980042, + "learning_rate": 2.2784341550083577e-06, + "loss": 0.263, + "step": 3386 + }, + { + "epoch": 2.408533333333333, + "grad_norm": 0.7618403434753418, + "learning_rate": 2.2731779236688047e-06, + "loss": 0.2327, + "step": 3387 + }, + { + "epoch": 2.4092444444444445, + "grad_norm": 0.6914927363395691, + "learning_rate": 2.2679269846751915e-06, + "loss": 0.2422, + "step": 3388 + }, + { + "epoch": 2.4099555555555554, + "grad_norm": 0.8417660593986511, + "learning_rate": 2.2626813416240654e-06, + "loss": 0.2166, + "step": 3389 + }, + { + "epoch": 2.4106666666666667, + "grad_norm": 0.8384866714477539, + "learning_rate": 2.2574409981083224e-06, + "loss": 0.2556, + "step": 3390 + }, + { + "epoch": 2.4113777777777776, + "grad_norm": 0.7790694236755371, + "learning_rate": 2.2522059577172473e-06, + "loss": 0.2349, + "step": 3391 + }, + { + "epoch": 2.412088888888889, + "grad_norm": 0.724858283996582, + "learning_rate": 2.2469762240364847e-06, + "loss": 0.2136, + "step": 3392 + }, + { + "epoch": 2.4128, + "grad_norm": 0.9192396402359009, + "learning_rate": 2.2417518006480445e-06, + "loss": 0.2618, + "step": 3393 + }, + { + "epoch": 2.413511111111111, + "grad_norm": 0.8153179287910461, + "learning_rate": 2.236532691130299e-06, + "loss": 0.2863, + "step": 3394 + }, + { + "epoch": 2.414222222222222, + "grad_norm": 0.8437612056732178, + "learning_rate": 2.2313188990579847e-06, + "loss": 0.2451, + "step": 3395 + }, + { + "epoch": 2.4149333333333334, + "grad_norm": 0.8530423045158386, + "learning_rate": 2.2261104280021937e-06, + "loss": 0.2197, + "step": 3396 + }, + { + "epoch": 2.4156444444444443, + "grad_norm": 1.2109960317611694, + "learning_rate": 2.220907281530368e-06, + "loss": 0.2318, + "step": 3397 + }, + { + "epoch": 2.4163555555555556, + "grad_norm": 0.9143720865249634, + "learning_rate": 2.215709463206316e-06, + "loss": 0.2788, + "step": 3398 + }, + { + "epoch": 2.4170666666666665, + "grad_norm": 0.8061244487762451, + "learning_rate": 2.210516976590179e-06, + "loss": 0.2657, + "step": 3399 + }, + { + "epoch": 2.417777777777778, + "grad_norm": 1.0813885927200317, + "learning_rate": 2.205329825238467e-06, + "loss": 0.2792, + "step": 3400 + }, + { + "epoch": 2.4184888888888887, + "grad_norm": 1.0541248321533203, + "learning_rate": 2.200148012704013e-06, + "loss": 0.2284, + "step": 3401 + }, + { + "epoch": 2.4192, + "grad_norm": 0.7808001637458801, + "learning_rate": 2.1949715425360173e-06, + "loss": 0.2311, + "step": 3402 + }, + { + "epoch": 2.4199111111111113, + "grad_norm": 0.8609311580657959, + "learning_rate": 2.189800418279998e-06, + "loss": 0.2581, + "step": 3403 + }, + { + "epoch": 2.4206222222222222, + "grad_norm": 1.0920259952545166, + "learning_rate": 2.184634643477831e-06, + "loss": 0.2777, + "step": 3404 + }, + { + "epoch": 2.421333333333333, + "grad_norm": 1.139925241470337, + "learning_rate": 2.1794742216677144e-06, + "loss": 0.2344, + "step": 3405 + }, + { + "epoch": 2.4220444444444444, + "grad_norm": 0.7828865647315979, + "learning_rate": 2.174319156384186e-06, + "loss": 0.2313, + "step": 3406 + }, + { + "epoch": 2.4227555555555558, + "grad_norm": 0.8084216713905334, + "learning_rate": 2.169169451158114e-06, + "loss": 0.24, + "step": 3407 + }, + { + "epoch": 2.4234666666666667, + "grad_norm": 0.753839373588562, + "learning_rate": 2.164025109516692e-06, + "loss": 0.2448, + "step": 3408 + }, + { + "epoch": 2.4241777777777775, + "grad_norm": 0.9027138948440552, + "learning_rate": 2.1588861349834524e-06, + "loss": 0.2661, + "step": 3409 + }, + { + "epoch": 2.424888888888889, + "grad_norm": 1.318329095840454, + "learning_rate": 2.15375253107823e-06, + "loss": 0.2463, + "step": 3410 + }, + { + "epoch": 2.4256, + "grad_norm": 0.9204988479614258, + "learning_rate": 2.148624301317206e-06, + "loss": 0.2466, + "step": 3411 + }, + { + "epoch": 2.426311111111111, + "grad_norm": 0.8544647097587585, + "learning_rate": 2.1435014492128547e-06, + "loss": 0.2551, + "step": 3412 + }, + { + "epoch": 2.4270222222222224, + "grad_norm": 0.7830209136009216, + "learning_rate": 2.1383839782739934e-06, + "loss": 0.2745, + "step": 3413 + }, + { + "epoch": 2.4277333333333333, + "grad_norm": 0.8886354565620422, + "learning_rate": 2.1332718920057307e-06, + "loss": 0.2754, + "step": 3414 + }, + { + "epoch": 2.4284444444444446, + "grad_norm": 0.8160765171051025, + "learning_rate": 2.1281651939094996e-06, + "loss": 0.2618, + "step": 3415 + }, + { + "epoch": 2.4291555555555555, + "grad_norm": 0.926052987575531, + "learning_rate": 2.1230638874830413e-06, + "loss": 0.2771, + "step": 3416 + }, + { + "epoch": 2.429866666666667, + "grad_norm": 0.7958491444587708, + "learning_rate": 2.1179679762204007e-06, + "loss": 0.2452, + "step": 3417 + }, + { + "epoch": 2.4305777777777777, + "grad_norm": 0.7723157405853271, + "learning_rate": 2.1128774636119307e-06, + "loss": 0.2442, + "step": 3418 + }, + { + "epoch": 2.431288888888889, + "grad_norm": 0.8454445600509644, + "learning_rate": 2.1077923531442858e-06, + "loss": 0.2592, + "step": 3419 + }, + { + "epoch": 2.432, + "grad_norm": 0.7022626399993896, + "learning_rate": 2.102712648300418e-06, + "loss": 0.2263, + "step": 3420 + }, + { + "epoch": 2.4327111111111113, + "grad_norm": 0.9142389297485352, + "learning_rate": 2.097638352559579e-06, + "loss": 0.2293, + "step": 3421 + }, + { + "epoch": 2.433422222222222, + "grad_norm": 0.9684513807296753, + "learning_rate": 2.0925694693973162e-06, + "loss": 0.2506, + "step": 3422 + }, + { + "epoch": 2.4341333333333335, + "grad_norm": 0.9157447218894958, + "learning_rate": 2.0875060022854678e-06, + "loss": 0.2686, + "step": 3423 + }, + { + "epoch": 2.4348444444444444, + "grad_norm": 0.9209379553794861, + "learning_rate": 2.082447954692164e-06, + "loss": 0.2647, + "step": 3424 + }, + { + "epoch": 2.4355555555555557, + "grad_norm": 0.9552001357078552, + "learning_rate": 2.0773953300818204e-06, + "loss": 0.2467, + "step": 3425 + }, + { + "epoch": 2.4362666666666666, + "grad_norm": 0.9056341052055359, + "learning_rate": 2.0723481319151427e-06, + "loss": 0.2539, + "step": 3426 + }, + { + "epoch": 2.436977777777778, + "grad_norm": 1.1216858625411987, + "learning_rate": 2.0673063636491165e-06, + "loss": 0.2605, + "step": 3427 + }, + { + "epoch": 2.437688888888889, + "grad_norm": 0.8456347584724426, + "learning_rate": 2.062270028737008e-06, + "loss": 0.2135, + "step": 3428 + }, + { + "epoch": 2.4384, + "grad_norm": 0.8898791670799255, + "learning_rate": 2.0572391306283644e-06, + "loss": 0.2151, + "step": 3429 + }, + { + "epoch": 2.439111111111111, + "grad_norm": 0.9329548478126526, + "learning_rate": 2.052213672769007e-06, + "loss": 0.2497, + "step": 3430 + }, + { + "epoch": 2.4398222222222223, + "grad_norm": 0.9240226745605469, + "learning_rate": 2.047193658601031e-06, + "loss": 0.2318, + "step": 3431 + }, + { + "epoch": 2.440533333333333, + "grad_norm": 0.9872658848762512, + "learning_rate": 2.042179091562805e-06, + "loss": 0.2867, + "step": 3432 + }, + { + "epoch": 2.4412444444444446, + "grad_norm": 0.770595908164978, + "learning_rate": 2.037169975088964e-06, + "loss": 0.2451, + "step": 3433 + }, + { + "epoch": 2.4419555555555554, + "grad_norm": 0.8369803428649902, + "learning_rate": 2.032166312610411e-06, + "loss": 0.2269, + "step": 3434 + }, + { + "epoch": 2.4426666666666668, + "grad_norm": 1.29037606716156, + "learning_rate": 2.0271681075543147e-06, + "loss": 0.2345, + "step": 3435 + }, + { + "epoch": 2.4433777777777776, + "grad_norm": 0.8664661645889282, + "learning_rate": 2.0221753633441033e-06, + "loss": 0.228, + "step": 3436 + }, + { + "epoch": 2.444088888888889, + "grad_norm": 0.7831581234931946, + "learning_rate": 2.0171880833994663e-06, + "loss": 0.2523, + "step": 3437 + }, + { + "epoch": 2.4448, + "grad_norm": 0.9257645010948181, + "learning_rate": 2.012206271136353e-06, + "loss": 0.2568, + "step": 3438 + }, + { + "epoch": 2.445511111111111, + "grad_norm": 0.6720404028892517, + "learning_rate": 2.007229929966957e-06, + "loss": 0.2435, + "step": 3439 + }, + { + "epoch": 2.446222222222222, + "grad_norm": 0.9370737671852112, + "learning_rate": 2.002259063299744e-06, + "loss": 0.2752, + "step": 3440 + }, + { + "epoch": 2.4469333333333334, + "grad_norm": 0.7742470502853394, + "learning_rate": 1.9972936745394067e-06, + "loss": 0.2378, + "step": 3441 + }, + { + "epoch": 2.4476444444444443, + "grad_norm": 0.8704806566238403, + "learning_rate": 1.992333767086905e-06, + "loss": 0.2591, + "step": 3442 + }, + { + "epoch": 2.4483555555555556, + "grad_norm": 0.8959915637969971, + "learning_rate": 1.987379344339435e-06, + "loss": 0.2231, + "step": 3443 + }, + { + "epoch": 2.4490666666666665, + "grad_norm": 0.8344115614891052, + "learning_rate": 1.982430409690439e-06, + "loss": 0.2306, + "step": 3444 + }, + { + "epoch": 2.449777777777778, + "grad_norm": 0.8404794335365295, + "learning_rate": 1.9774869665295994e-06, + "loss": 0.2752, + "step": 3445 + }, + { + "epoch": 2.4504888888888887, + "grad_norm": 0.7656662464141846, + "learning_rate": 1.972549018242836e-06, + "loss": 0.2348, + "step": 3446 + }, + { + "epoch": 2.4512, + "grad_norm": 0.8397576808929443, + "learning_rate": 1.9676165682123118e-06, + "loss": 0.2717, + "step": 3447 + }, + { + "epoch": 2.451911111111111, + "grad_norm": 0.7498820424079895, + "learning_rate": 1.9626896198164093e-06, + "loss": 0.2654, + "step": 3448 + }, + { + "epoch": 2.4526222222222223, + "grad_norm": 0.7822222113609314, + "learning_rate": 1.9577681764297607e-06, + "loss": 0.2427, + "step": 3449 + }, + { + "epoch": 2.453333333333333, + "grad_norm": 0.8933442831039429, + "learning_rate": 1.9528522414232122e-06, + "loss": 0.2483, + "step": 3450 + }, + { + "epoch": 2.4540444444444445, + "grad_norm": 0.8212963342666626, + "learning_rate": 1.9479418181638508e-06, + "loss": 0.2178, + "step": 3451 + }, + { + "epoch": 2.454755555555556, + "grad_norm": 0.8187187314033508, + "learning_rate": 1.9430369100149727e-06, + "loss": 0.2941, + "step": 3452 + }, + { + "epoch": 2.4554666666666667, + "grad_norm": 0.8432840704917908, + "learning_rate": 1.9381375203361152e-06, + "loss": 0.2395, + "step": 3453 + }, + { + "epoch": 2.4561777777777776, + "grad_norm": 0.7637280821800232, + "learning_rate": 1.9332436524830167e-06, + "loss": 0.2742, + "step": 3454 + }, + { + "epoch": 2.456888888888889, + "grad_norm": 0.8147969245910645, + "learning_rate": 1.9283553098076514e-06, + "loss": 0.2664, + "step": 3455 + }, + { + "epoch": 2.4576000000000002, + "grad_norm": 1.1006672382354736, + "learning_rate": 1.9234724956581918e-06, + "loss": 0.2331, + "step": 3456 + }, + { + "epoch": 2.458311111111111, + "grad_norm": 1.0546988248825073, + "learning_rate": 1.9185952133790418e-06, + "loss": 0.2544, + "step": 3457 + }, + { + "epoch": 2.459022222222222, + "grad_norm": 0.9645570516586304, + "learning_rate": 1.9137234663107995e-06, + "loss": 0.2541, + "step": 3458 + }, + { + "epoch": 2.4597333333333333, + "grad_norm": 0.8298269510269165, + "learning_rate": 1.9088572577902787e-06, + "loss": 0.2374, + "step": 3459 + }, + { + "epoch": 2.4604444444444447, + "grad_norm": 0.8849007487297058, + "learning_rate": 1.9039965911505098e-06, + "loss": 0.2853, + "step": 3460 + }, + { + "epoch": 2.4611555555555555, + "grad_norm": 1.1191372871398926, + "learning_rate": 1.8991414697207055e-06, + "loss": 0.249, + "step": 3461 + }, + { + "epoch": 2.4618666666666664, + "grad_norm": 1.4892385005950928, + "learning_rate": 1.8942918968263036e-06, + "loss": 0.2568, + "step": 3462 + }, + { + "epoch": 2.4625777777777778, + "grad_norm": 0.9723178148269653, + "learning_rate": 1.8894478757889212e-06, + "loss": 0.2507, + "step": 3463 + }, + { + "epoch": 2.463288888888889, + "grad_norm": 0.8483824133872986, + "learning_rate": 1.8846094099263911e-06, + "loss": 0.2301, + "step": 3464 + }, + { + "epoch": 2.464, + "grad_norm": 0.8174514770507812, + "learning_rate": 1.8797765025527236e-06, + "loss": 0.252, + "step": 3465 + }, + { + "epoch": 2.4647111111111113, + "grad_norm": 1.2634565830230713, + "learning_rate": 1.8749491569781397e-06, + "loss": 0.214, + "step": 3466 + }, + { + "epoch": 2.465422222222222, + "grad_norm": 1.1432791948318481, + "learning_rate": 1.8701273765090332e-06, + "loss": 0.2274, + "step": 3467 + }, + { + "epoch": 2.4661333333333335, + "grad_norm": 0.9590831398963928, + "learning_rate": 1.8653111644480004e-06, + "loss": 0.2345, + "step": 3468 + }, + { + "epoch": 2.4668444444444444, + "grad_norm": 0.8378301858901978, + "learning_rate": 1.8605005240938146e-06, + "loss": 0.2445, + "step": 3469 + }, + { + "epoch": 2.4675555555555557, + "grad_norm": 0.8732430338859558, + "learning_rate": 1.8556954587414377e-06, + "loss": 0.2344, + "step": 3470 + }, + { + "epoch": 2.4682666666666666, + "grad_norm": 0.913066565990448, + "learning_rate": 1.850895971682013e-06, + "loss": 0.229, + "step": 3471 + }, + { + "epoch": 2.468977777777778, + "grad_norm": 0.8112624287605286, + "learning_rate": 1.8461020662028583e-06, + "loss": 0.2444, + "step": 3472 + }, + { + "epoch": 2.469688888888889, + "grad_norm": 0.9384649991989136, + "learning_rate": 1.8413137455874752e-06, + "loss": 0.283, + "step": 3473 + }, + { + "epoch": 2.4704, + "grad_norm": 0.8735242486000061, + "learning_rate": 1.8365310131155345e-06, + "loss": 0.2821, + "step": 3474 + }, + { + "epoch": 2.471111111111111, + "grad_norm": 0.7806518077850342, + "learning_rate": 1.8317538720628824e-06, + "loss": 0.2508, + "step": 3475 + }, + { + "epoch": 2.4718222222222224, + "grad_norm": 0.8513110280036926, + "learning_rate": 1.8269823257015351e-06, + "loss": 0.2643, + "step": 3476 + }, + { + "epoch": 2.4725333333333332, + "grad_norm": 0.7763661742210388, + "learning_rate": 1.8222163772996749e-06, + "loss": 0.2352, + "step": 3477 + }, + { + "epoch": 2.4732444444444446, + "grad_norm": 0.8701643943786621, + "learning_rate": 1.8174560301216527e-06, + "loss": 0.297, + "step": 3478 + }, + { + "epoch": 2.4739555555555555, + "grad_norm": 0.8602277636528015, + "learning_rate": 1.8127012874279803e-06, + "loss": 0.2394, + "step": 3479 + }, + { + "epoch": 2.474666666666667, + "grad_norm": 0.9128549098968506, + "learning_rate": 1.807952152475333e-06, + "loss": 0.2195, + "step": 3480 + }, + { + "epoch": 2.4753777777777777, + "grad_norm": 0.8902149200439453, + "learning_rate": 1.8032086285165439e-06, + "loss": 0.2664, + "step": 3481 + }, + { + "epoch": 2.476088888888889, + "grad_norm": 0.7353541254997253, + "learning_rate": 1.7984707188006034e-06, + "loss": 0.2587, + "step": 3482 + }, + { + "epoch": 2.4768, + "grad_norm": 0.8406313061714172, + "learning_rate": 1.7937384265726565e-06, + "loss": 0.2247, + "step": 3483 + }, + { + "epoch": 2.477511111111111, + "grad_norm": 0.8059192299842834, + "learning_rate": 1.7890117550739995e-06, + "loss": 0.2403, + "step": 3484 + }, + { + "epoch": 2.478222222222222, + "grad_norm": 0.7803523540496826, + "learning_rate": 1.7842907075420813e-06, + "loss": 0.2275, + "step": 3485 + }, + { + "epoch": 2.4789333333333334, + "grad_norm": 0.7957448959350586, + "learning_rate": 1.7795752872104965e-06, + "loss": 0.2311, + "step": 3486 + }, + { + "epoch": 2.4796444444444443, + "grad_norm": 0.7284836173057556, + "learning_rate": 1.7748654973089862e-06, + "loss": 0.2545, + "step": 3487 + }, + { + "epoch": 2.4803555555555556, + "grad_norm": 0.7368085980415344, + "learning_rate": 1.7701613410634367e-06, + "loss": 0.2175, + "step": 3488 + }, + { + "epoch": 2.4810666666666665, + "grad_norm": 0.8826286792755127, + "learning_rate": 1.7654628216958714e-06, + "loss": 0.2707, + "step": 3489 + }, + { + "epoch": 2.481777777777778, + "grad_norm": 0.862820565700531, + "learning_rate": 1.7607699424244583e-06, + "loss": 0.2539, + "step": 3490 + }, + { + "epoch": 2.4824888888888887, + "grad_norm": 0.8364204168319702, + "learning_rate": 1.7560827064635e-06, + "loss": 0.2827, + "step": 3491 + }, + { + "epoch": 2.4832, + "grad_norm": 0.8173037171363831, + "learning_rate": 1.7514011170234258e-06, + "loss": 0.2563, + "step": 3492 + }, + { + "epoch": 2.483911111111111, + "grad_norm": 0.8859419822692871, + "learning_rate": 1.7467251773108174e-06, + "loss": 0.2221, + "step": 3493 + }, + { + "epoch": 2.4846222222222223, + "grad_norm": 0.7302088141441345, + "learning_rate": 1.7420548905283619e-06, + "loss": 0.2629, + "step": 3494 + }, + { + "epoch": 2.485333333333333, + "grad_norm": 0.6927922964096069, + "learning_rate": 1.7373902598748948e-06, + "loss": 0.2886, + "step": 3495 + }, + { + "epoch": 2.4860444444444445, + "grad_norm": 0.6742798089981079, + "learning_rate": 1.7327312885453695e-06, + "loss": 0.1943, + "step": 3496 + }, + { + "epoch": 2.4867555555555554, + "grad_norm": 0.8567212224006653, + "learning_rate": 1.7280779797308612e-06, + "loss": 0.2843, + "step": 3497 + }, + { + "epoch": 2.4874666666666667, + "grad_norm": 0.8511165976524353, + "learning_rate": 1.7234303366185712e-06, + "loss": 0.228, + "step": 3498 + }, + { + "epoch": 2.4881777777777776, + "grad_norm": 0.6392432451248169, + "learning_rate": 1.7187883623918155e-06, + "loss": 0.2186, + "step": 3499 + }, + { + "epoch": 2.488888888888889, + "grad_norm": 0.7985187768936157, + "learning_rate": 1.7141520602300332e-06, + "loss": 0.2623, + "step": 3500 + }, + { + "epoch": 2.4896, + "grad_norm": 0.8171596527099609, + "learning_rate": 1.7095214333087683e-06, + "loss": 0.2506, + "step": 3501 + }, + { + "epoch": 2.490311111111111, + "grad_norm": 0.7618283033370972, + "learning_rate": 1.7048964847996928e-06, + "loss": 0.2175, + "step": 3502 + }, + { + "epoch": 2.491022222222222, + "grad_norm": 0.9066774845123291, + "learning_rate": 1.7002772178705717e-06, + "loss": 0.2837, + "step": 3503 + }, + { + "epoch": 2.4917333333333334, + "grad_norm": 0.8003077507019043, + "learning_rate": 1.6956636356852984e-06, + "loss": 0.2433, + "step": 3504 + }, + { + "epoch": 2.4924444444444447, + "grad_norm": 0.8937931060791016, + "learning_rate": 1.691055741403851e-06, + "loss": 0.2546, + "step": 3505 + }, + { + "epoch": 2.4931555555555556, + "grad_norm": 0.8001189231872559, + "learning_rate": 1.6864535381823333e-06, + "loss": 0.2681, + "step": 3506 + }, + { + "epoch": 2.4938666666666665, + "grad_norm": 0.8502639532089233, + "learning_rate": 1.6818570291729296e-06, + "loss": 0.2483, + "step": 3507 + }, + { + "epoch": 2.494577777777778, + "grad_norm": 0.7704052925109863, + "learning_rate": 1.6772662175239451e-06, + "loss": 0.2497, + "step": 3508 + }, + { + "epoch": 2.495288888888889, + "grad_norm": 0.6480867266654968, + "learning_rate": 1.6726811063797665e-06, + "loss": 0.2446, + "step": 3509 + }, + { + "epoch": 2.496, + "grad_norm": 0.8028436303138733, + "learning_rate": 1.668101698880883e-06, + "loss": 0.2636, + "step": 3510 + }, + { + "epoch": 2.496711111111111, + "grad_norm": 0.6923930644989014, + "learning_rate": 1.6635279981638763e-06, + "loss": 0.2248, + "step": 3511 + }, + { + "epoch": 2.497422222222222, + "grad_norm": 0.8528380990028381, + "learning_rate": 1.6589600073614175e-06, + "loss": 0.2438, + "step": 3512 + }, + { + "epoch": 2.4981333333333335, + "grad_norm": 0.7833235263824463, + "learning_rate": 1.6543977296022774e-06, + "loss": 0.2348, + "step": 3513 + }, + { + "epoch": 2.4988444444444444, + "grad_norm": 0.8300179839134216, + "learning_rate": 1.6498411680112925e-06, + "loss": 0.2979, + "step": 3514 + }, + { + "epoch": 2.4995555555555553, + "grad_norm": 0.8023446798324585, + "learning_rate": 1.6452903257094078e-06, + "loss": 0.2555, + "step": 3515 + }, + { + "epoch": 2.5002666666666666, + "grad_norm": 0.7292760610580444, + "learning_rate": 1.6407452058136298e-06, + "loss": 0.2582, + "step": 3516 + }, + { + "epoch": 2.500977777777778, + "grad_norm": 0.82747882604599, + "learning_rate": 1.636205811437066e-06, + "loss": 0.2377, + "step": 3517 + }, + { + "epoch": 2.501688888888889, + "grad_norm": 0.837181806564331, + "learning_rate": 1.6316721456888807e-06, + "loss": 0.2332, + "step": 3518 + }, + { + "epoch": 2.5023999999999997, + "grad_norm": 0.793440043926239, + "learning_rate": 1.6271442116743363e-06, + "loss": 0.2473, + "step": 3519 + }, + { + "epoch": 2.503111111111111, + "grad_norm": 0.7254613041877747, + "learning_rate": 1.6226220124947513e-06, + "loss": 0.2511, + "step": 3520 + }, + { + "epoch": 2.5038222222222224, + "grad_norm": 0.8964571952819824, + "learning_rate": 1.6181055512475252e-06, + "loss": 0.2497, + "step": 3521 + }, + { + "epoch": 2.5045333333333333, + "grad_norm": 0.8645471334457397, + "learning_rate": 1.6135948310261272e-06, + "loss": 0.2443, + "step": 3522 + }, + { + "epoch": 2.5052444444444446, + "grad_norm": 0.8229385614395142, + "learning_rate": 1.609089854920093e-06, + "loss": 0.2409, + "step": 3523 + }, + { + "epoch": 2.5059555555555555, + "grad_norm": 0.9627602100372314, + "learning_rate": 1.6045906260150212e-06, + "loss": 0.2825, + "step": 3524 + }, + { + "epoch": 2.506666666666667, + "grad_norm": 0.7668552398681641, + "learning_rate": 1.6000971473925797e-06, + "loss": 0.252, + "step": 3525 + }, + { + "epoch": 2.5073777777777777, + "grad_norm": 0.7879424095153809, + "learning_rate": 1.595609422130494e-06, + "loss": 0.264, + "step": 3526 + }, + { + "epoch": 2.508088888888889, + "grad_norm": 0.8180865049362183, + "learning_rate": 1.591127453302549e-06, + "loss": 0.2208, + "step": 3527 + }, + { + "epoch": 2.5088, + "grad_norm": 0.8570448160171509, + "learning_rate": 1.5866512439785876e-06, + "loss": 0.2594, + "step": 3528 + }, + { + "epoch": 2.5095111111111112, + "grad_norm": 0.7927308678627014, + "learning_rate": 1.5821807972245073e-06, + "loss": 0.2189, + "step": 3529 + }, + { + "epoch": 2.510222222222222, + "grad_norm": 0.7723230123519897, + "learning_rate": 1.5777161161022614e-06, + "loss": 0.2523, + "step": 3530 + }, + { + "epoch": 2.5109333333333335, + "grad_norm": 0.7767552733421326, + "learning_rate": 1.573257203669849e-06, + "loss": 0.2562, + "step": 3531 + }, + { + "epoch": 2.5116444444444443, + "grad_norm": 0.8277260661125183, + "learning_rate": 1.5688040629813229e-06, + "loss": 0.2119, + "step": 3532 + }, + { + "epoch": 2.5123555555555557, + "grad_norm": 0.8761545419692993, + "learning_rate": 1.5643566970867795e-06, + "loss": 0.2458, + "step": 3533 + }, + { + "epoch": 2.5130666666666666, + "grad_norm": 0.7448089718818665, + "learning_rate": 1.5599151090323627e-06, + "loss": 0.2408, + "step": 3534 + }, + { + "epoch": 2.513777777777778, + "grad_norm": 0.9663931131362915, + "learning_rate": 1.5554793018602555e-06, + "loss": 0.2644, + "step": 3535 + }, + { + "epoch": 2.5144888888888888, + "grad_norm": 0.8360951542854309, + "learning_rate": 1.5510492786086828e-06, + "loss": 0.2619, + "step": 3536 + }, + { + "epoch": 2.5152, + "grad_norm": 0.8592237234115601, + "learning_rate": 1.54662504231191e-06, + "loss": 0.2665, + "step": 3537 + }, + { + "epoch": 2.515911111111111, + "grad_norm": 0.7966651320457458, + "learning_rate": 1.5422065960002364e-06, + "loss": 0.2534, + "step": 3538 + }, + { + "epoch": 2.5166222222222223, + "grad_norm": 0.8036684989929199, + "learning_rate": 1.5377939426999967e-06, + "loss": 0.245, + "step": 3539 + }, + { + "epoch": 2.517333333333333, + "grad_norm": 0.8059629797935486, + "learning_rate": 1.5333870854335554e-06, + "loss": 0.216, + "step": 3540 + }, + { + "epoch": 2.5180444444444445, + "grad_norm": 0.7563759088516235, + "learning_rate": 1.5289860272193114e-06, + "loss": 0.235, + "step": 3541 + }, + { + "epoch": 2.5187555555555554, + "grad_norm": 0.8402915000915527, + "learning_rate": 1.5245907710716912e-06, + "loss": 0.2615, + "step": 3542 + }, + { + "epoch": 2.5194666666666667, + "grad_norm": 0.6156527400016785, + "learning_rate": 1.5202013200011378e-06, + "loss": 0.216, + "step": 3543 + }, + { + "epoch": 2.5201777777777776, + "grad_norm": 0.8102596998214722, + "learning_rate": 1.5158176770141342e-06, + "loss": 0.2502, + "step": 3544 + }, + { + "epoch": 2.520888888888889, + "grad_norm": 0.9468392729759216, + "learning_rate": 1.5114398451131696e-06, + "loss": 0.2403, + "step": 3545 + }, + { + "epoch": 2.5216, + "grad_norm": 0.7898325324058533, + "learning_rate": 1.5070678272967654e-06, + "loss": 0.2234, + "step": 3546 + }, + { + "epoch": 2.522311111111111, + "grad_norm": 0.7864283323287964, + "learning_rate": 1.502701626559454e-06, + "loss": 0.2245, + "step": 3547 + }, + { + "epoch": 2.523022222222222, + "grad_norm": 0.8425540328025818, + "learning_rate": 1.4983412458917846e-06, + "loss": 0.2657, + "step": 3548 + }, + { + "epoch": 2.5237333333333334, + "grad_norm": 0.8133649826049805, + "learning_rate": 1.4939866882803211e-06, + "loss": 0.2427, + "step": 3549 + }, + { + "epoch": 2.5244444444444447, + "grad_norm": 0.8068386912345886, + "learning_rate": 1.4896379567076369e-06, + "loss": 0.2211, + "step": 3550 + }, + { + "epoch": 2.5251555555555556, + "grad_norm": 0.7677752375602722, + "learning_rate": 1.4852950541523182e-06, + "loss": 0.2484, + "step": 3551 + }, + { + "epoch": 2.5258666666666665, + "grad_norm": 0.7878347635269165, + "learning_rate": 1.4809579835889564e-06, + "loss": 0.2652, + "step": 3552 + }, + { + "epoch": 2.526577777777778, + "grad_norm": 0.6647890210151672, + "learning_rate": 1.4766267479881502e-06, + "loss": 0.2424, + "step": 3553 + }, + { + "epoch": 2.527288888888889, + "grad_norm": 0.6844689846038818, + "learning_rate": 1.472301350316495e-06, + "loss": 0.2476, + "step": 3554 + }, + { + "epoch": 2.528, + "grad_norm": 0.7026804685592651, + "learning_rate": 1.4679817935366014e-06, + "loss": 0.2517, + "step": 3555 + }, + { + "epoch": 2.528711111111111, + "grad_norm": 0.9221867322921753, + "learning_rate": 1.4636680806070625e-06, + "loss": 0.2486, + "step": 3556 + }, + { + "epoch": 2.5294222222222222, + "grad_norm": 0.8527686595916748, + "learning_rate": 1.459360214482486e-06, + "loss": 0.2358, + "step": 3557 + }, + { + "epoch": 2.5301333333333336, + "grad_norm": 0.8664951324462891, + "learning_rate": 1.4550581981134571e-06, + "loss": 0.2416, + "step": 3558 + }, + { + "epoch": 2.5308444444444445, + "grad_norm": 0.8075878620147705, + "learning_rate": 1.4507620344465734e-06, + "loss": 0.2384, + "step": 3559 + }, + { + "epoch": 2.5315555555555553, + "grad_norm": 0.8874047994613647, + "learning_rate": 1.4464717264244043e-06, + "loss": 0.2455, + "step": 3560 + }, + { + "epoch": 2.5322666666666667, + "grad_norm": 0.8619409799575806, + "learning_rate": 1.4421872769855262e-06, + "loss": 0.245, + "step": 3561 + }, + { + "epoch": 2.532977777777778, + "grad_norm": 0.7095908522605896, + "learning_rate": 1.43790868906449e-06, + "loss": 0.2625, + "step": 3562 + }, + { + "epoch": 2.533688888888889, + "grad_norm": 0.7570310831069946, + "learning_rate": 1.4336359655918342e-06, + "loss": 0.2318, + "step": 3563 + }, + { + "epoch": 2.5343999999999998, + "grad_norm": 0.7521541118621826, + "learning_rate": 1.429369109494091e-06, + "loss": 0.2596, + "step": 3564 + }, + { + "epoch": 2.535111111111111, + "grad_norm": 1.2003728151321411, + "learning_rate": 1.4251081236937569e-06, + "loss": 0.2216, + "step": 3565 + }, + { + "epoch": 2.5358222222222224, + "grad_norm": 0.7628364562988281, + "learning_rate": 1.4208530111093244e-06, + "loss": 0.2402, + "step": 3566 + }, + { + "epoch": 2.5365333333333333, + "grad_norm": 0.7850592732429504, + "learning_rate": 1.4166037746552474e-06, + "loss": 0.2398, + "step": 3567 + }, + { + "epoch": 2.537244444444444, + "grad_norm": 0.8207187652587891, + "learning_rate": 1.4123604172419714e-06, + "loss": 0.2133, + "step": 3568 + }, + { + "epoch": 2.5379555555555555, + "grad_norm": 0.8211995959281921, + "learning_rate": 1.4081229417758979e-06, + "loss": 0.2379, + "step": 3569 + }, + { + "epoch": 2.538666666666667, + "grad_norm": 0.8729783892631531, + "learning_rate": 1.4038913511594166e-06, + "loss": 0.2486, + "step": 3570 + }, + { + "epoch": 2.5393777777777777, + "grad_norm": 0.8495983481407166, + "learning_rate": 1.3996656482908733e-06, + "loss": 0.2391, + "step": 3571 + }, + { + "epoch": 2.5400888888888886, + "grad_norm": 0.7731685042381287, + "learning_rate": 1.395445836064586e-06, + "loss": 0.2084, + "step": 3572 + }, + { + "epoch": 2.5408, + "grad_norm": 0.7984690070152283, + "learning_rate": 1.3912319173708412e-06, + "loss": 0.2448, + "step": 3573 + }, + { + "epoch": 2.5415111111111113, + "grad_norm": 0.7849193811416626, + "learning_rate": 1.3870238950958837e-06, + "loss": 0.2181, + "step": 3574 + }, + { + "epoch": 2.542222222222222, + "grad_norm": 0.713087797164917, + "learning_rate": 1.3828217721219216e-06, + "loss": 0.2217, + "step": 3575 + }, + { + "epoch": 2.5429333333333335, + "grad_norm": 0.7438204288482666, + "learning_rate": 1.378625551327124e-06, + "loss": 0.2631, + "step": 3576 + }, + { + "epoch": 2.5436444444444444, + "grad_norm": 0.8073191046714783, + "learning_rate": 1.374435235585614e-06, + "loss": 0.2611, + "step": 3577 + }, + { + "epoch": 2.5443555555555557, + "grad_norm": 0.8913996815681458, + "learning_rate": 1.3702508277674731e-06, + "loss": 0.2773, + "step": 3578 + }, + { + "epoch": 2.5450666666666666, + "grad_norm": 0.867111325263977, + "learning_rate": 1.3660723307387346e-06, + "loss": 0.3033, + "step": 3579 + }, + { + "epoch": 2.545777777777778, + "grad_norm": 0.8071076273918152, + "learning_rate": 1.3618997473613837e-06, + "loss": 0.2563, + "step": 3580 + }, + { + "epoch": 2.546488888888889, + "grad_norm": 0.7647906541824341, + "learning_rate": 1.3577330804933563e-06, + "loss": 0.2304, + "step": 3581 + }, + { + "epoch": 2.5472, + "grad_norm": 0.9915387034416199, + "learning_rate": 1.353572332988534e-06, + "loss": 0.2478, + "step": 3582 + }, + { + "epoch": 2.547911111111111, + "grad_norm": 0.8178264498710632, + "learning_rate": 1.3494175076967465e-06, + "loss": 0.2484, + "step": 3583 + }, + { + "epoch": 2.5486222222222223, + "grad_norm": 0.8954777717590332, + "learning_rate": 1.3452686074637632e-06, + "loss": 0.2588, + "step": 3584 + }, + { + "epoch": 2.5493333333333332, + "grad_norm": 0.8536892533302307, + "learning_rate": 1.3411256351312995e-06, + "loss": 0.2614, + "step": 3585 + }, + { + "epoch": 2.5500444444444446, + "grad_norm": 0.8681527376174927, + "learning_rate": 1.3369885935370086e-06, + "loss": 0.2428, + "step": 3586 + }, + { + "epoch": 2.5507555555555554, + "grad_norm": 0.812282145023346, + "learning_rate": 1.3328574855144815e-06, + "loss": 0.234, + "step": 3587 + }, + { + "epoch": 2.5514666666666668, + "grad_norm": 0.8284729719161987, + "learning_rate": 1.328732313893245e-06, + "loss": 0.2413, + "step": 3588 + }, + { + "epoch": 2.5521777777777777, + "grad_norm": 0.7114407420158386, + "learning_rate": 1.3246130814987602e-06, + "loss": 0.2377, + "step": 3589 + }, + { + "epoch": 2.552888888888889, + "grad_norm": 0.7381722331047058, + "learning_rate": 1.320499791152421e-06, + "loss": 0.2634, + "step": 3590 + }, + { + "epoch": 2.5536, + "grad_norm": 0.8435962200164795, + "learning_rate": 1.3163924456715493e-06, + "loss": 0.2439, + "step": 3591 + }, + { + "epoch": 2.554311111111111, + "grad_norm": 0.5660008788108826, + "learning_rate": 1.3122910478693984e-06, + "loss": 0.2018, + "step": 3592 + }, + { + "epoch": 2.555022222222222, + "grad_norm": 0.8502591252326965, + "learning_rate": 1.3081956005551443e-06, + "loss": 0.2641, + "step": 3593 + }, + { + "epoch": 2.5557333333333334, + "grad_norm": 0.811636745929718, + "learning_rate": 1.30410610653389e-06, + "loss": 0.2263, + "step": 3594 + }, + { + "epoch": 2.5564444444444443, + "grad_norm": 0.6605433821678162, + "learning_rate": 1.3000225686066625e-06, + "loss": 0.2174, + "step": 3595 + }, + { + "epoch": 2.5571555555555556, + "grad_norm": 0.8886907696723938, + "learning_rate": 1.295944989570398e-06, + "loss": 0.28, + "step": 3596 + }, + { + "epoch": 2.5578666666666665, + "grad_norm": 0.7983787059783936, + "learning_rate": 1.2918733722179687e-06, + "loss": 0.2524, + "step": 3597 + }, + { + "epoch": 2.558577777777778, + "grad_norm": 0.8261702656745911, + "learning_rate": 1.2878077193381511e-06, + "loss": 0.2615, + "step": 3598 + }, + { + "epoch": 2.5592888888888887, + "grad_norm": 0.7143946290016174, + "learning_rate": 1.2837480337156416e-06, + "loss": 0.2253, + "step": 3599 + }, + { + "epoch": 2.56, + "grad_norm": 0.7815272808074951, + "learning_rate": 1.279694318131046e-06, + "loss": 0.2327, + "step": 3600 + }, + { + "epoch": 2.560711111111111, + "grad_norm": 0.8215644955635071, + "learning_rate": 1.2756465753608837e-06, + "loss": 0.2173, + "step": 3601 + }, + { + "epoch": 2.5614222222222223, + "grad_norm": 0.8760786652565002, + "learning_rate": 1.2716048081775823e-06, + "loss": 0.2565, + "step": 3602 + }, + { + "epoch": 2.5621333333333336, + "grad_norm": 0.8162429332733154, + "learning_rate": 1.2675690193494772e-06, + "loss": 0.2491, + "step": 3603 + }, + { + "epoch": 2.5628444444444445, + "grad_norm": 0.7052783966064453, + "learning_rate": 1.2635392116408095e-06, + "loss": 0.2425, + "step": 3604 + }, + { + "epoch": 2.5635555555555554, + "grad_norm": 0.9410218596458435, + "learning_rate": 1.2595153878117172e-06, + "loss": 0.2589, + "step": 3605 + }, + { + "epoch": 2.5642666666666667, + "grad_norm": 0.686633825302124, + "learning_rate": 1.2554975506182533e-06, + "loss": 0.2046, + "step": 3606 + }, + { + "epoch": 2.564977777777778, + "grad_norm": 0.8434421420097351, + "learning_rate": 1.251485702812353e-06, + "loss": 0.2335, + "step": 3607 + }, + { + "epoch": 2.565688888888889, + "grad_norm": 0.7142518162727356, + "learning_rate": 1.247479847141867e-06, + "loss": 0.2127, + "step": 3608 + }, + { + "epoch": 2.5664, + "grad_norm": 0.7531276345252991, + "learning_rate": 1.2434799863505254e-06, + "loss": 0.2304, + "step": 3609 + }, + { + "epoch": 2.567111111111111, + "grad_norm": 0.753470242023468, + "learning_rate": 1.2394861231779677e-06, + "loss": 0.2291, + "step": 3610 + }, + { + "epoch": 2.5678222222222225, + "grad_norm": 0.7206466197967529, + "learning_rate": 1.2354982603597087e-06, + "loss": 0.2374, + "step": 3611 + }, + { + "epoch": 2.5685333333333333, + "grad_norm": 0.8956260681152344, + "learning_rate": 1.2315164006271718e-06, + "loss": 0.2364, + "step": 3612 + }, + { + "epoch": 2.569244444444444, + "grad_norm": 0.8329667448997498, + "learning_rate": 1.2275405467076507e-06, + "loss": 0.2412, + "step": 3613 + }, + { + "epoch": 2.5699555555555555, + "grad_norm": 0.7541993856430054, + "learning_rate": 1.2235707013243426e-06, + "loss": 0.2539, + "step": 3614 + }, + { + "epoch": 2.570666666666667, + "grad_norm": 0.7543962001800537, + "learning_rate": 1.2196068671963147e-06, + "loss": 0.3037, + "step": 3615 + }, + { + "epoch": 2.5713777777777778, + "grad_norm": 0.8327545523643494, + "learning_rate": 1.2156490470385207e-06, + "loss": 0.2398, + "step": 3616 + }, + { + "epoch": 2.5720888888888886, + "grad_norm": 0.7573009729385376, + "learning_rate": 1.211697243561807e-06, + "loss": 0.2295, + "step": 3617 + }, + { + "epoch": 2.5728, + "grad_norm": 0.7903913259506226, + "learning_rate": 1.2077514594728778e-06, + "loss": 0.2723, + "step": 3618 + }, + { + "epoch": 2.5735111111111113, + "grad_norm": 0.7421913743019104, + "learning_rate": 1.2038116974743374e-06, + "loss": 0.2509, + "step": 3619 + }, + { + "epoch": 2.574222222222222, + "grad_norm": 0.9952073693275452, + "learning_rate": 1.1998779602646438e-06, + "loss": 0.2234, + "step": 3620 + }, + { + "epoch": 2.574933333333333, + "grad_norm": 0.7922948598861694, + "learning_rate": 1.1959502505381459e-06, + "loss": 0.2643, + "step": 3621 + }, + { + "epoch": 2.5756444444444444, + "grad_norm": 0.8719518184661865, + "learning_rate": 1.1920285709850509e-06, + "loss": 0.2345, + "step": 3622 + }, + { + "epoch": 2.5763555555555557, + "grad_norm": 0.8297742009162903, + "learning_rate": 1.1881129242914503e-06, + "loss": 0.2361, + "step": 3623 + }, + { + "epoch": 2.5770666666666666, + "grad_norm": 0.8777033090591431, + "learning_rate": 1.184203313139286e-06, + "loss": 0.261, + "step": 3624 + }, + { + "epoch": 2.5777777777777775, + "grad_norm": 0.7928270101547241, + "learning_rate": 1.1802997402063788e-06, + "loss": 0.2198, + "step": 3625 + }, + { + "epoch": 2.578488888888889, + "grad_norm": 0.854528546333313, + "learning_rate": 1.1764022081664094e-06, + "loss": 0.2283, + "step": 3626 + }, + { + "epoch": 2.5792, + "grad_norm": 0.8084720373153687, + "learning_rate": 1.1725107196889207e-06, + "loss": 0.207, + "step": 3627 + }, + { + "epoch": 2.579911111111111, + "grad_norm": 0.8246445059776306, + "learning_rate": 1.1686252774393181e-06, + "loss": 0.2641, + "step": 3628 + }, + { + "epoch": 2.5806222222222224, + "grad_norm": 0.7998336553573608, + "learning_rate": 1.164745884078864e-06, + "loss": 0.2612, + "step": 3629 + }, + { + "epoch": 2.5813333333333333, + "grad_norm": 0.788468062877655, + "learning_rate": 1.1608725422646782e-06, + "loss": 0.2088, + "step": 3630 + }, + { + "epoch": 2.5820444444444446, + "grad_norm": 0.7958565354347229, + "learning_rate": 1.1570052546497356e-06, + "loss": 0.2358, + "step": 3631 + }, + { + "epoch": 2.5827555555555555, + "grad_norm": 0.8269078731536865, + "learning_rate": 1.1531440238828639e-06, + "loss": 0.252, + "step": 3632 + }, + { + "epoch": 2.583466666666667, + "grad_norm": 0.7740273475646973, + "learning_rate": 1.149288852608743e-06, + "loss": 0.2746, + "step": 3633 + }, + { + "epoch": 2.5841777777777777, + "grad_norm": 0.9736828207969666, + "learning_rate": 1.1454397434679022e-06, + "loss": 0.2633, + "step": 3634 + }, + { + "epoch": 2.584888888888889, + "grad_norm": 0.9935736060142517, + "learning_rate": 1.1415966990967187e-06, + "loss": 0.2411, + "step": 3635 + }, + { + "epoch": 2.5856, + "grad_norm": 1.079317569732666, + "learning_rate": 1.137759722127415e-06, + "loss": 0.2635, + "step": 3636 + }, + { + "epoch": 2.5863111111111112, + "grad_norm": 0.8414633274078369, + "learning_rate": 1.13392881518806e-06, + "loss": 0.2668, + "step": 3637 + }, + { + "epoch": 2.587022222222222, + "grad_norm": 0.7144757509231567, + "learning_rate": 1.1301039809025628e-06, + "loss": 0.2292, + "step": 3638 + }, + { + "epoch": 2.5877333333333334, + "grad_norm": 0.6632040739059448, + "learning_rate": 1.1262852218906727e-06, + "loss": 0.2252, + "step": 3639 + }, + { + "epoch": 2.5884444444444443, + "grad_norm": 0.7559524774551392, + "learning_rate": 1.1224725407679814e-06, + "loss": 0.2487, + "step": 3640 + }, + { + "epoch": 2.5891555555555557, + "grad_norm": 0.7129889130592346, + "learning_rate": 1.1186659401459144e-06, + "loss": 0.3008, + "step": 3641 + }, + { + "epoch": 2.5898666666666665, + "grad_norm": 0.7513238191604614, + "learning_rate": 1.1148654226317325e-06, + "loss": 0.2352, + "step": 3642 + }, + { + "epoch": 2.590577777777778, + "grad_norm": 0.6975008845329285, + "learning_rate": 1.1110709908285321e-06, + "loss": 0.2817, + "step": 3643 + }, + { + "epoch": 2.5912888888888888, + "grad_norm": 0.8723995089530945, + "learning_rate": 1.1072826473352394e-06, + "loss": 0.2469, + "step": 3644 + }, + { + "epoch": 2.592, + "grad_norm": 0.7657083868980408, + "learning_rate": 1.1035003947466117e-06, + "loss": 0.2218, + "step": 3645 + }, + { + "epoch": 2.592711111111111, + "grad_norm": 0.7740097641944885, + "learning_rate": 1.0997242356532335e-06, + "loss": 0.2149, + "step": 3646 + }, + { + "epoch": 2.5934222222222223, + "grad_norm": 0.8600714802742004, + "learning_rate": 1.0959541726415169e-06, + "loss": 0.2395, + "step": 3647 + }, + { + "epoch": 2.594133333333333, + "grad_norm": 0.8022419810295105, + "learning_rate": 1.0921902082936987e-06, + "loss": 0.2335, + "step": 3648 + }, + { + "epoch": 2.5948444444444445, + "grad_norm": 0.7621999382972717, + "learning_rate": 1.088432345187832e-06, + "loss": 0.2269, + "step": 3649 + }, + { + "epoch": 2.5955555555555554, + "grad_norm": 0.8658623695373535, + "learning_rate": 1.0846805858978038e-06, + "loss": 0.2977, + "step": 3650 + }, + { + "epoch": 2.5962666666666667, + "grad_norm": 0.7368628978729248, + "learning_rate": 1.0809349329933106e-06, + "loss": 0.2545, + "step": 3651 + }, + { + "epoch": 2.596977777777778, + "grad_norm": 0.7881020307540894, + "learning_rate": 1.0771953890398679e-06, + "loss": 0.2166, + "step": 3652 + }, + { + "epoch": 2.597688888888889, + "grad_norm": 0.7656248211860657, + "learning_rate": 1.0734619565988102e-06, + "loss": 0.2193, + "step": 3653 + }, + { + "epoch": 2.5984, + "grad_norm": 0.8152212500572205, + "learning_rate": 1.0697346382272822e-06, + "loss": 0.2037, + "step": 3654 + }, + { + "epoch": 2.599111111111111, + "grad_norm": 0.8432772755622864, + "learning_rate": 1.0660134364782438e-06, + "loss": 0.2443, + "step": 3655 + }, + { + "epoch": 2.5998222222222225, + "grad_norm": 0.7905818819999695, + "learning_rate": 1.0622983539004628e-06, + "loss": 0.2504, + "step": 3656 + }, + { + "epoch": 2.6005333333333334, + "grad_norm": 0.7895497679710388, + "learning_rate": 1.0585893930385217e-06, + "loss": 0.2319, + "step": 3657 + }, + { + "epoch": 2.6012444444444442, + "grad_norm": 0.6540895104408264, + "learning_rate": 1.054886556432798e-06, + "loss": 0.2126, + "step": 3658 + }, + { + "epoch": 2.6019555555555556, + "grad_norm": 0.9945231080055237, + "learning_rate": 1.0511898466194903e-06, + "loss": 0.2502, + "step": 3659 + }, + { + "epoch": 2.602666666666667, + "grad_norm": 0.807640790939331, + "learning_rate": 1.047499266130585e-06, + "loss": 0.2425, + "step": 3660 + }, + { + "epoch": 2.603377777777778, + "grad_norm": 0.6584818959236145, + "learning_rate": 1.0438148174938866e-06, + "loss": 0.2197, + "step": 3661 + }, + { + "epoch": 2.6040888888888887, + "grad_norm": 0.7614957690238953, + "learning_rate": 1.0401365032329812e-06, + "loss": 0.2392, + "step": 3662 + }, + { + "epoch": 2.6048, + "grad_norm": 0.7761163115501404, + "learning_rate": 1.0364643258672724e-06, + "loss": 0.2341, + "step": 3663 + }, + { + "epoch": 2.6055111111111113, + "grad_norm": 0.814751923084259, + "learning_rate": 1.0327982879119425e-06, + "loss": 0.2646, + "step": 3664 + }, + { + "epoch": 2.606222222222222, + "grad_norm": 0.8331725597381592, + "learning_rate": 1.0291383918779863e-06, + "loss": 0.2499, + "step": 3665 + }, + { + "epoch": 2.606933333333333, + "grad_norm": 0.7394776344299316, + "learning_rate": 1.0254846402721764e-06, + "loss": 0.253, + "step": 3666 + }, + { + "epoch": 2.6076444444444444, + "grad_norm": 0.8019039034843445, + "learning_rate": 1.021837035597083e-06, + "loss": 0.2513, + "step": 3667 + }, + { + "epoch": 2.6083555555555558, + "grad_norm": 0.6951479911804199, + "learning_rate": 1.0181955803510724e-06, + "loss": 0.236, + "step": 3668 + }, + { + "epoch": 2.6090666666666666, + "grad_norm": 0.8799194097518921, + "learning_rate": 1.0145602770282858e-06, + "loss": 0.2425, + "step": 3669 + }, + { + "epoch": 2.6097777777777775, + "grad_norm": 0.8319181799888611, + "learning_rate": 1.010931128118665e-06, + "loss": 0.2156, + "step": 3670 + }, + { + "epoch": 2.610488888888889, + "grad_norm": 0.7191848158836365, + "learning_rate": 1.0073081361079217e-06, + "loss": 0.2691, + "step": 3671 + }, + { + "epoch": 2.6112, + "grad_norm": 0.747683048248291, + "learning_rate": 1.0036913034775675e-06, + "loss": 0.237, + "step": 3672 + }, + { + "epoch": 2.611911111111111, + "grad_norm": 0.846871554851532, + "learning_rate": 1.0000806327048763e-06, + "loss": 0.2242, + "step": 3673 + }, + { + "epoch": 2.612622222222222, + "grad_norm": 0.8147781491279602, + "learning_rate": 9.964761262629196e-07, + "loss": 0.2803, + "step": 3674 + }, + { + "epoch": 2.6133333333333333, + "grad_norm": 1.1053522825241089, + "learning_rate": 9.92877786620533e-07, + "loss": 0.286, + "step": 3675 + }, + { + "epoch": 2.6140444444444446, + "grad_norm": 0.9101401567459106, + "learning_rate": 9.892856162423348e-07, + "loss": 0.2404, + "step": 3676 + }, + { + "epoch": 2.6147555555555555, + "grad_norm": 0.7220370769500732, + "learning_rate": 9.85699617588718e-07, + "loss": 0.2378, + "step": 3677 + }, + { + "epoch": 2.615466666666667, + "grad_norm": 0.8417822122573853, + "learning_rate": 9.821197931158455e-07, + "loss": 0.2441, + "step": 3678 + }, + { + "epoch": 2.6161777777777777, + "grad_norm": 0.795430600643158, + "learning_rate": 9.78546145275654e-07, + "loss": 0.2366, + "step": 3679 + }, + { + "epoch": 2.616888888888889, + "grad_norm": 0.6951660513877869, + "learning_rate": 9.749786765158464e-07, + "loss": 0.2416, + "step": 3680 + }, + { + "epoch": 2.6176, + "grad_norm": 0.7543653249740601, + "learning_rate": 9.714173892798984e-07, + "loss": 0.2507, + "step": 3681 + }, + { + "epoch": 2.6183111111111113, + "grad_norm": 0.7920611500740051, + "learning_rate": 9.678622860070474e-07, + "loss": 0.2294, + "step": 3682 + }, + { + "epoch": 2.619022222222222, + "grad_norm": 0.7584957480430603, + "learning_rate": 9.643133691322959e-07, + "loss": 0.2652, + "step": 3683 + }, + { + "epoch": 2.6197333333333335, + "grad_norm": 0.863461971282959, + "learning_rate": 9.607706410864083e-07, + "loss": 0.2355, + "step": 3684 + }, + { + "epoch": 2.6204444444444444, + "grad_norm": 0.7709413766860962, + "learning_rate": 9.572341042959177e-07, + "loss": 0.2294, + "step": 3685 + }, + { + "epoch": 2.6211555555555557, + "grad_norm": 0.7490400075912476, + "learning_rate": 9.537037611831047e-07, + "loss": 0.2514, + "step": 3686 + }, + { + "epoch": 2.6218666666666666, + "grad_norm": 0.7036316394805908, + "learning_rate": 9.501796141660146e-07, + "loss": 0.2448, + "step": 3687 + }, + { + "epoch": 2.622577777777778, + "grad_norm": 0.8008118271827698, + "learning_rate": 9.466616656584493e-07, + "loss": 0.2377, + "step": 3688 + }, + { + "epoch": 2.623288888888889, + "grad_norm": 0.8194980025291443, + "learning_rate": 9.431499180699621e-07, + "loss": 0.2345, + "step": 3689 + }, + { + "epoch": 2.624, + "grad_norm": 0.7536994814872742, + "learning_rate": 9.396443738058614e-07, + "loss": 0.2175, + "step": 3690 + }, + { + "epoch": 2.624711111111111, + "grad_norm": 0.780031144618988, + "learning_rate": 9.361450352672041e-07, + "loss": 0.2518, + "step": 3691 + }, + { + "epoch": 2.6254222222222223, + "grad_norm": 0.6689401865005493, + "learning_rate": 9.32651904850801e-07, + "loss": 0.2384, + "step": 3692 + }, + { + "epoch": 2.626133333333333, + "grad_norm": 0.865814208984375, + "learning_rate": 9.291649849492068e-07, + "loss": 0.2745, + "step": 3693 + }, + { + "epoch": 2.6268444444444445, + "grad_norm": 0.7438658475875854, + "learning_rate": 9.256842779507236e-07, + "loss": 0.219, + "step": 3694 + }, + { + "epoch": 2.6275555555555554, + "grad_norm": 0.8591089248657227, + "learning_rate": 9.222097862393975e-07, + "loss": 0.2379, + "step": 3695 + }, + { + "epoch": 2.6282666666666668, + "grad_norm": 0.8800861239433289, + "learning_rate": 9.187415121950194e-07, + "loss": 0.2204, + "step": 3696 + }, + { + "epoch": 2.6289777777777776, + "grad_norm": 0.8184394240379333, + "learning_rate": 9.152794581931201e-07, + "loss": 0.2362, + "step": 3697 + }, + { + "epoch": 2.629688888888889, + "grad_norm": 0.9007277488708496, + "learning_rate": 9.118236266049707e-07, + "loss": 0.2326, + "step": 3698 + }, + { + "epoch": 2.6304, + "grad_norm": 0.7556014060974121, + "learning_rate": 9.083740197975822e-07, + "loss": 0.2026, + "step": 3699 + }, + { + "epoch": 2.631111111111111, + "grad_norm": 1.1586464643478394, + "learning_rate": 9.049306401336922e-07, + "loss": 0.2464, + "step": 3700 + }, + { + "epoch": 2.631822222222222, + "grad_norm": 0.7712914347648621, + "learning_rate": 9.014934899717887e-07, + "loss": 0.2353, + "step": 3701 + }, + { + "epoch": 2.6325333333333334, + "grad_norm": 0.790988564491272, + "learning_rate": 8.980625716660829e-07, + "loss": 0.2387, + "step": 3702 + }, + { + "epoch": 2.6332444444444443, + "grad_norm": 0.8568844795227051, + "learning_rate": 8.946378875665185e-07, + "loss": 0.2527, + "step": 3703 + }, + { + "epoch": 2.6339555555555556, + "grad_norm": 0.7952578663825989, + "learning_rate": 8.912194400187712e-07, + "loss": 0.2301, + "step": 3704 + }, + { + "epoch": 2.634666666666667, + "grad_norm": 0.7855502367019653, + "learning_rate": 8.87807231364245e-07, + "loss": 0.2311, + "step": 3705 + }, + { + "epoch": 2.635377777777778, + "grad_norm": 0.8910679221153259, + "learning_rate": 8.84401263940069e-07, + "loss": 0.2615, + "step": 3706 + }, + { + "epoch": 2.6360888888888887, + "grad_norm": 0.7724398970603943, + "learning_rate": 8.810015400790994e-07, + "loss": 0.2273, + "step": 3707 + }, + { + "epoch": 2.6368, + "grad_norm": 0.7857895493507385, + "learning_rate": 8.776080621099159e-07, + "loss": 0.2488, + "step": 3708 + }, + { + "epoch": 2.6375111111111114, + "grad_norm": 0.803518533706665, + "learning_rate": 8.742208323568146e-07, + "loss": 0.2502, + "step": 3709 + }, + { + "epoch": 2.6382222222222222, + "grad_norm": 1.0529948472976685, + "learning_rate": 8.708398531398233e-07, + "loss": 0.2319, + "step": 3710 + }, + { + "epoch": 2.638933333333333, + "grad_norm": 1.0052566528320312, + "learning_rate": 8.674651267746769e-07, + "loss": 0.2306, + "step": 3711 + }, + { + "epoch": 2.6396444444444445, + "grad_norm": 0.7725250124931335, + "learning_rate": 8.640966555728369e-07, + "loss": 0.2043, + "step": 3712 + }, + { + "epoch": 2.640355555555556, + "grad_norm": 0.7907592058181763, + "learning_rate": 8.60734441841472e-07, + "loss": 0.2656, + "step": 3713 + }, + { + "epoch": 2.6410666666666667, + "grad_norm": 0.8058518767356873, + "learning_rate": 8.573784878834734e-07, + "loss": 0.2532, + "step": 3714 + }, + { + "epoch": 2.6417777777777776, + "grad_norm": 0.852972686290741, + "learning_rate": 8.540287959974369e-07, + "loss": 0.2891, + "step": 3715 + }, + { + "epoch": 2.642488888888889, + "grad_norm": 0.754054844379425, + "learning_rate": 8.506853684776773e-07, + "loss": 0.2284, + "step": 3716 + }, + { + "epoch": 2.6432, + "grad_norm": 1.004063606262207, + "learning_rate": 8.47348207614207e-07, + "loss": 0.2415, + "step": 3717 + }, + { + "epoch": 2.643911111111111, + "grad_norm": 0.9302878975868225, + "learning_rate": 8.440173156927612e-07, + "loss": 0.2622, + "step": 3718 + }, + { + "epoch": 2.644622222222222, + "grad_norm": 0.8116752505302429, + "learning_rate": 8.406926949947725e-07, + "loss": 0.2395, + "step": 3719 + }, + { + "epoch": 2.6453333333333333, + "grad_norm": 0.9603848457336426, + "learning_rate": 8.373743477973739e-07, + "loss": 0.2531, + "step": 3720 + }, + { + "epoch": 2.6460444444444446, + "grad_norm": 0.6816114783287048, + "learning_rate": 8.340622763734129e-07, + "loss": 0.2482, + "step": 3721 + }, + { + "epoch": 2.6467555555555555, + "grad_norm": 0.8684312105178833, + "learning_rate": 8.307564829914272e-07, + "loss": 0.2364, + "step": 3722 + }, + { + "epoch": 2.6474666666666664, + "grad_norm": 0.7987639307975769, + "learning_rate": 8.274569699156654e-07, + "loss": 0.2498, + "step": 3723 + }, + { + "epoch": 2.6481777777777777, + "grad_norm": 0.8600010275840759, + "learning_rate": 8.241637394060619e-07, + "loss": 0.2563, + "step": 3724 + }, + { + "epoch": 2.648888888888889, + "grad_norm": 1.0580215454101562, + "learning_rate": 8.208767937182627e-07, + "loss": 0.2584, + "step": 3725 + }, + { + "epoch": 2.6496, + "grad_norm": 1.1398588418960571, + "learning_rate": 8.175961351035943e-07, + "loss": 0.2606, + "step": 3726 + }, + { + "epoch": 2.650311111111111, + "grad_norm": 1.0001095533370972, + "learning_rate": 8.143217658090907e-07, + "loss": 0.2296, + "step": 3727 + }, + { + "epoch": 2.651022222222222, + "grad_norm": 0.9162297248840332, + "learning_rate": 8.110536880774655e-07, + "loss": 0.2478, + "step": 3728 + }, + { + "epoch": 2.6517333333333335, + "grad_norm": 1.048332691192627, + "learning_rate": 8.077919041471316e-07, + "loss": 0.2377, + "step": 3729 + }, + { + "epoch": 2.6524444444444444, + "grad_norm": 0.9458330273628235, + "learning_rate": 8.045364162521884e-07, + "loss": 0.2534, + "step": 3730 + }, + { + "epoch": 2.6531555555555557, + "grad_norm": 1.0031903982162476, + "learning_rate": 8.012872266224226e-07, + "loss": 0.2399, + "step": 3731 + }, + { + "epoch": 2.6538666666666666, + "grad_norm": 1.0651928186416626, + "learning_rate": 7.98044337483308e-07, + "loss": 0.2643, + "step": 3732 + }, + { + "epoch": 2.654577777777778, + "grad_norm": 0.932671308517456, + "learning_rate": 7.948077510560015e-07, + "loss": 0.2319, + "step": 3733 + }, + { + "epoch": 2.655288888888889, + "grad_norm": 0.9897739887237549, + "learning_rate": 7.915774695573452e-07, + "loss": 0.2351, + "step": 3734 + }, + { + "epoch": 2.656, + "grad_norm": 0.9137410521507263, + "learning_rate": 7.883534951998594e-07, + "loss": 0.243, + "step": 3735 + }, + { + "epoch": 2.656711111111111, + "grad_norm": 0.7333173751831055, + "learning_rate": 7.851358301917511e-07, + "loss": 0.2647, + "step": 3736 + }, + { + "epoch": 2.6574222222222224, + "grad_norm": 0.7916717529296875, + "learning_rate": 7.81924476736896e-07, + "loss": 0.2457, + "step": 3737 + }, + { + "epoch": 2.6581333333333332, + "grad_norm": 0.7790485620498657, + "learning_rate": 7.787194370348549e-07, + "loss": 0.2095, + "step": 3738 + }, + { + "epoch": 2.6588444444444446, + "grad_norm": 1.0272417068481445, + "learning_rate": 7.755207132808607e-07, + "loss": 0.2324, + "step": 3739 + }, + { + "epoch": 2.6595555555555555, + "grad_norm": 0.9117814302444458, + "learning_rate": 7.723283076658217e-07, + "loss": 0.2602, + "step": 3740 + }, + { + "epoch": 2.660266666666667, + "grad_norm": 0.7342990040779114, + "learning_rate": 7.691422223763168e-07, + "loss": 0.252, + "step": 3741 + }, + { + "epoch": 2.6609777777777777, + "grad_norm": 0.7725772261619568, + "learning_rate": 7.659624595945969e-07, + "loss": 0.2415, + "step": 3742 + }, + { + "epoch": 2.661688888888889, + "grad_norm": 0.9380192160606384, + "learning_rate": 7.627890214985845e-07, + "loss": 0.2418, + "step": 3743 + }, + { + "epoch": 2.6624, + "grad_norm": 0.8229965567588806, + "learning_rate": 7.596219102618652e-07, + "loss": 0.2389, + "step": 3744 + }, + { + "epoch": 2.663111111111111, + "grad_norm": 0.7865661382675171, + "learning_rate": 7.564611280536971e-07, + "loss": 0.2476, + "step": 3745 + }, + { + "epoch": 2.663822222222222, + "grad_norm": 0.9502144455909729, + "learning_rate": 7.533066770389985e-07, + "loss": 0.2538, + "step": 3746 + }, + { + "epoch": 2.6645333333333334, + "grad_norm": 0.8162227869033813, + "learning_rate": 7.501585593783534e-07, + "loss": 0.2554, + "step": 3747 + }, + { + "epoch": 2.6652444444444443, + "grad_norm": 0.6620343923568726, + "learning_rate": 7.470167772280091e-07, + "loss": 0.2139, + "step": 3748 + }, + { + "epoch": 2.6659555555555556, + "grad_norm": 0.9968510270118713, + "learning_rate": 7.438813327398698e-07, + "loss": 0.2227, + "step": 3749 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 0.8502991199493408, + "learning_rate": 7.40752228061502e-07, + "loss": 0.2467, + "step": 3750 + }, + { + "epoch": 2.667377777777778, + "grad_norm": 0.9511616826057434, + "learning_rate": 7.376294653361293e-07, + "loss": 0.2344, + "step": 3751 + }, + { + "epoch": 2.6680888888888887, + "grad_norm": 0.8088825345039368, + "learning_rate": 7.345130467026318e-07, + "loss": 0.2551, + "step": 3752 + }, + { + "epoch": 2.6688, + "grad_norm": 0.6750593185424805, + "learning_rate": 7.314029742955397e-07, + "loss": 0.23, + "step": 3753 + }, + { + "epoch": 2.669511111111111, + "grad_norm": 0.7476187944412231, + "learning_rate": 7.282992502450447e-07, + "loss": 0.2419, + "step": 3754 + }, + { + "epoch": 2.6702222222222223, + "grad_norm": 0.8072186708450317, + "learning_rate": 7.252018766769852e-07, + "loss": 0.2308, + "step": 3755 + }, + { + "epoch": 2.670933333333333, + "grad_norm": 1.0617496967315674, + "learning_rate": 7.221108557128509e-07, + "loss": 0.2243, + "step": 3756 + }, + { + "epoch": 2.6716444444444445, + "grad_norm": 0.7823320627212524, + "learning_rate": 7.190261894697781e-07, + "loss": 0.2209, + "step": 3757 + }, + { + "epoch": 2.672355555555556, + "grad_norm": 1.2282345294952393, + "learning_rate": 7.159478800605546e-07, + "loss": 0.2179, + "step": 3758 + }, + { + "epoch": 2.6730666666666667, + "grad_norm": 1.1320315599441528, + "learning_rate": 7.128759295936128e-07, + "loss": 0.2455, + "step": 3759 + }, + { + "epoch": 2.6737777777777776, + "grad_norm": 0.8011634349822998, + "learning_rate": 7.098103401730272e-07, + "loss": 0.2424, + "step": 3760 + }, + { + "epoch": 2.674488888888889, + "grad_norm": 0.985451340675354, + "learning_rate": 7.067511138985205e-07, + "loss": 0.2636, + "step": 3761 + }, + { + "epoch": 2.6752000000000002, + "grad_norm": 0.7659029364585876, + "learning_rate": 7.03698252865449e-07, + "loss": 0.2728, + "step": 3762 + }, + { + "epoch": 2.675911111111111, + "grad_norm": 0.7149627208709717, + "learning_rate": 7.00651759164821e-07, + "loss": 0.2339, + "step": 3763 + }, + { + "epoch": 2.676622222222222, + "grad_norm": 0.7408421635627747, + "learning_rate": 6.976116348832684e-07, + "loss": 0.2686, + "step": 3764 + }, + { + "epoch": 2.6773333333333333, + "grad_norm": 1.070568323135376, + "learning_rate": 6.94577882103078e-07, + "loss": 0.266, + "step": 3765 + }, + { + "epoch": 2.6780444444444447, + "grad_norm": 0.9475810527801514, + "learning_rate": 6.915505029021552e-07, + "loss": 0.258, + "step": 3766 + }, + { + "epoch": 2.6787555555555556, + "grad_norm": 0.7391075491905212, + "learning_rate": 6.885294993540548e-07, + "loss": 0.272, + "step": 3767 + }, + { + "epoch": 2.6794666666666664, + "grad_norm": 0.7997547388076782, + "learning_rate": 6.855148735279527e-07, + "loss": 0.276, + "step": 3768 + }, + { + "epoch": 2.6801777777777778, + "grad_norm": 0.7207791805267334, + "learning_rate": 6.825066274886683e-07, + "loss": 0.2392, + "step": 3769 + }, + { + "epoch": 2.680888888888889, + "grad_norm": 0.680671215057373, + "learning_rate": 6.795047632966379e-07, + "loss": 0.231, + "step": 3770 + }, + { + "epoch": 2.6816, + "grad_norm": 0.6879740357398987, + "learning_rate": 6.765092830079367e-07, + "loss": 0.2321, + "step": 3771 + }, + { + "epoch": 2.682311111111111, + "grad_norm": 0.6808366179466248, + "learning_rate": 6.735201886742671e-07, + "loss": 0.2338, + "step": 3772 + }, + { + "epoch": 2.683022222222222, + "grad_norm": 0.7136128544807434, + "learning_rate": 6.7053748234295e-07, + "loss": 0.219, + "step": 3773 + }, + { + "epoch": 2.6837333333333335, + "grad_norm": 0.9002569913864136, + "learning_rate": 6.675611660569403e-07, + "loss": 0.2247, + "step": 3774 + }, + { + "epoch": 2.6844444444444444, + "grad_norm": 0.8971635103225708, + "learning_rate": 6.645912418548062e-07, + "loss": 0.2459, + "step": 3775 + }, + { + "epoch": 2.6851555555555553, + "grad_norm": 0.802325963973999, + "learning_rate": 6.616277117707493e-07, + "loss": 0.2181, + "step": 3776 + }, + { + "epoch": 2.6858666666666666, + "grad_norm": 0.7804865837097168, + "learning_rate": 6.586705778345792e-07, + "loss": 0.2567, + "step": 3777 + }, + { + "epoch": 2.686577777777778, + "grad_norm": 0.8815149068832397, + "learning_rate": 6.55719842071737e-07, + "loss": 0.2497, + "step": 3778 + }, + { + "epoch": 2.687288888888889, + "grad_norm": 0.6935712099075317, + "learning_rate": 6.527755065032703e-07, + "loss": 0.2129, + "step": 3779 + }, + { + "epoch": 2.6879999999999997, + "grad_norm": 0.7736771702766418, + "learning_rate": 6.498375731458529e-07, + "loss": 0.2657, + "step": 3780 + }, + { + "epoch": 2.688711111111111, + "grad_norm": 0.7569565176963806, + "learning_rate": 6.469060440117647e-07, + "loss": 0.208, + "step": 3781 + }, + { + "epoch": 2.6894222222222224, + "grad_norm": 0.6426849365234375, + "learning_rate": 6.439809211089043e-07, + "loss": 0.2286, + "step": 3782 + }, + { + "epoch": 2.6901333333333333, + "grad_norm": 0.9683665037155151, + "learning_rate": 6.410622064407812e-07, + "loss": 0.245, + "step": 3783 + }, + { + "epoch": 2.6908444444444446, + "grad_norm": 0.8809741139411926, + "learning_rate": 6.381499020065163e-07, + "loss": 0.2662, + "step": 3784 + }, + { + "epoch": 2.6915555555555555, + "grad_norm": 0.8865557909011841, + "learning_rate": 6.352440098008394e-07, + "loss": 0.2171, + "step": 3785 + }, + { + "epoch": 2.692266666666667, + "grad_norm": 0.8526678085327148, + "learning_rate": 6.323445318140886e-07, + "loss": 0.2731, + "step": 3786 + }, + { + "epoch": 2.6929777777777777, + "grad_norm": 1.2072006464004517, + "learning_rate": 6.294514700322074e-07, + "loss": 0.2464, + "step": 3787 + }, + { + "epoch": 2.693688888888889, + "grad_norm": 0.7175335884094238, + "learning_rate": 6.265648264367452e-07, + "loss": 0.2362, + "step": 3788 + }, + { + "epoch": 2.6944, + "grad_norm": 0.876399040222168, + "learning_rate": 6.236846030048605e-07, + "loss": 0.2421, + "step": 3789 + }, + { + "epoch": 2.6951111111111112, + "grad_norm": 0.967387855052948, + "learning_rate": 6.20810801709305e-07, + "loss": 0.2493, + "step": 3790 + }, + { + "epoch": 2.695822222222222, + "grad_norm": 0.9304255843162537, + "learning_rate": 6.17943424518439e-07, + "loss": 0.266, + "step": 3791 + }, + { + "epoch": 2.6965333333333334, + "grad_norm": 0.7948147058486938, + "learning_rate": 6.15082473396218e-07, + "loss": 0.2038, + "step": 3792 + }, + { + "epoch": 2.6972444444444443, + "grad_norm": 0.9227646589279175, + "learning_rate": 6.122279503022e-07, + "loss": 0.2449, + "step": 3793 + }, + { + "epoch": 2.6979555555555557, + "grad_norm": 1.0841495990753174, + "learning_rate": 6.093798571915389e-07, + "loss": 0.2566, + "step": 3794 + }, + { + "epoch": 2.6986666666666665, + "grad_norm": 0.9105656743049622, + "learning_rate": 6.06538196014983e-07, + "loss": 0.2152, + "step": 3795 + }, + { + "epoch": 2.699377777777778, + "grad_norm": 0.8337554335594177, + "learning_rate": 6.037029687188767e-07, + "loss": 0.2225, + "step": 3796 + }, + { + "epoch": 2.7000888888888888, + "grad_norm": 0.9058566093444824, + "learning_rate": 6.008741772451577e-07, + "loss": 0.2552, + "step": 3797 + }, + { + "epoch": 2.7008, + "grad_norm": 0.6731008887290955, + "learning_rate": 5.980518235313549e-07, + "loss": 0.2507, + "step": 3798 + }, + { + "epoch": 2.701511111111111, + "grad_norm": 0.8459587097167969, + "learning_rate": 5.952359095105875e-07, + "loss": 0.2243, + "step": 3799 + }, + { + "epoch": 2.7022222222222223, + "grad_norm": 0.6987087726593018, + "learning_rate": 5.924264371115652e-07, + "loss": 0.2639, + "step": 3800 + }, + { + "epoch": 2.702933333333333, + "grad_norm": 0.8812319040298462, + "learning_rate": 5.89623408258585e-07, + "loss": 0.2268, + "step": 3801 + }, + { + "epoch": 2.7036444444444445, + "grad_norm": 0.7991546988487244, + "learning_rate": 5.868268248715292e-07, + "loss": 0.2505, + "step": 3802 + }, + { + "epoch": 2.7043555555555554, + "grad_norm": 0.8590669631958008, + "learning_rate": 5.840366888658699e-07, + "loss": 0.2442, + "step": 3803 + }, + { + "epoch": 2.7050666666666667, + "grad_norm": 1.0711246728897095, + "learning_rate": 5.812530021526541e-07, + "loss": 0.2449, + "step": 3804 + }, + { + "epoch": 2.7057777777777776, + "grad_norm": 0.7640824913978577, + "learning_rate": 5.784757666385222e-07, + "loss": 0.2661, + "step": 3805 + }, + { + "epoch": 2.706488888888889, + "grad_norm": 0.7913153171539307, + "learning_rate": 5.7570498422569e-07, + "loss": 0.2392, + "step": 3806 + }, + { + "epoch": 2.7072000000000003, + "grad_norm": 0.9095047116279602, + "learning_rate": 5.729406568119533e-07, + "loss": 0.2253, + "step": 3807 + }, + { + "epoch": 2.707911111111111, + "grad_norm": 0.979907751083374, + "learning_rate": 5.701827862906894e-07, + "loss": 0.2482, + "step": 3808 + }, + { + "epoch": 2.708622222222222, + "grad_norm": 0.7347231507301331, + "learning_rate": 5.674313745508497e-07, + "loss": 0.2323, + "step": 3809 + }, + { + "epoch": 2.7093333333333334, + "grad_norm": 0.7760056853294373, + "learning_rate": 5.646864234769644e-07, + "loss": 0.2242, + "step": 3810 + }, + { + "epoch": 2.7100444444444447, + "grad_norm": 0.7026920914649963, + "learning_rate": 5.619479349491375e-07, + "loss": 0.2524, + "step": 3811 + }, + { + "epoch": 2.7107555555555556, + "grad_norm": 0.8522655963897705, + "learning_rate": 5.592159108430472e-07, + "loss": 0.2222, + "step": 3812 + }, + { + "epoch": 2.7114666666666665, + "grad_norm": 0.7915143966674805, + "learning_rate": 5.564903530299425e-07, + "loss": 0.2508, + "step": 3813 + }, + { + "epoch": 2.712177777777778, + "grad_norm": 0.7544441223144531, + "learning_rate": 5.537712633766479e-07, + "loss": 0.2308, + "step": 3814 + }, + { + "epoch": 2.712888888888889, + "grad_norm": 1.0607975721359253, + "learning_rate": 5.510586437455478e-07, + "loss": 0.2362, + "step": 3815 + }, + { + "epoch": 2.7136, + "grad_norm": 0.7248743772506714, + "learning_rate": 5.483524959946097e-07, + "loss": 0.2104, + "step": 3816 + }, + { + "epoch": 2.714311111111111, + "grad_norm": 1.08781099319458, + "learning_rate": 5.456528219773516e-07, + "loss": 0.2356, + "step": 3817 + }, + { + "epoch": 2.7150222222222222, + "grad_norm": 0.859074056148529, + "learning_rate": 5.429596235428746e-07, + "loss": 0.2463, + "step": 3818 + }, + { + "epoch": 2.7157333333333336, + "grad_norm": 0.8499848246574402, + "learning_rate": 5.402729025358289e-07, + "loss": 0.2305, + "step": 3819 + }, + { + "epoch": 2.7164444444444444, + "grad_norm": 0.8064936995506287, + "learning_rate": 5.375926607964399e-07, + "loss": 0.2442, + "step": 3820 + }, + { + "epoch": 2.7171555555555553, + "grad_norm": 0.8330156207084656, + "learning_rate": 5.34918900160486e-07, + "loss": 0.2261, + "step": 3821 + }, + { + "epoch": 2.7178666666666667, + "grad_norm": 0.8274140954017639, + "learning_rate": 5.322516224593143e-07, + "loss": 0.2266, + "step": 3822 + }, + { + "epoch": 2.718577777777778, + "grad_norm": 1.1732409000396729, + "learning_rate": 5.295908295198282e-07, + "loss": 0.2572, + "step": 3823 + }, + { + "epoch": 2.719288888888889, + "grad_norm": 1.0670359134674072, + "learning_rate": 5.269365231644851e-07, + "loss": 0.2428, + "step": 3824 + }, + { + "epoch": 2.7199999999999998, + "grad_norm": 1.1796815395355225, + "learning_rate": 5.2428870521131e-07, + "loss": 0.2503, + "step": 3825 + }, + { + "epoch": 2.720711111111111, + "grad_norm": 0.6921119093894958, + "learning_rate": 5.216473774738706e-07, + "loss": 0.2226, + "step": 3826 + }, + { + "epoch": 2.7214222222222224, + "grad_norm": 0.7383131980895996, + "learning_rate": 5.190125417613035e-07, + "loss": 0.2769, + "step": 3827 + }, + { + "epoch": 2.7221333333333333, + "grad_norm": 0.7471231818199158, + "learning_rate": 5.163841998782837e-07, + "loss": 0.2555, + "step": 3828 + }, + { + "epoch": 2.722844444444444, + "grad_norm": 0.9638481140136719, + "learning_rate": 5.137623536250535e-07, + "loss": 0.258, + "step": 3829 + }, + { + "epoch": 2.7235555555555555, + "grad_norm": 0.6957417726516724, + "learning_rate": 5.111470047973932e-07, + "loss": 0.2512, + "step": 3830 + }, + { + "epoch": 2.724266666666667, + "grad_norm": 0.7481257915496826, + "learning_rate": 5.085381551866419e-07, + "loss": 0.2161, + "step": 3831 + }, + { + "epoch": 2.7249777777777777, + "grad_norm": 0.7723830342292786, + "learning_rate": 5.059358065796816e-07, + "loss": 0.2771, + "step": 3832 + }, + { + "epoch": 2.725688888888889, + "grad_norm": 0.6775333881378174, + "learning_rate": 5.03339960758944e-07, + "loss": 0.2238, + "step": 3833 + }, + { + "epoch": 2.7264, + "grad_norm": 0.7074286937713623, + "learning_rate": 5.007506195024059e-07, + "loss": 0.2218, + "step": 3834 + }, + { + "epoch": 2.7271111111111113, + "grad_norm": 0.6157869100570679, + "learning_rate": 4.981677845835897e-07, + "loss": 0.2247, + "step": 3835 + }, + { + "epoch": 2.727822222222222, + "grad_norm": 1.1283228397369385, + "learning_rate": 4.955914577715615e-07, + "loss": 0.217, + "step": 3836 + }, + { + "epoch": 2.7285333333333335, + "grad_norm": 0.8209608197212219, + "learning_rate": 4.930216408309296e-07, + "loss": 0.2587, + "step": 3837 + }, + { + "epoch": 2.7292444444444444, + "grad_norm": 0.8684893846511841, + "learning_rate": 4.904583355218429e-07, + "loss": 0.245, + "step": 3838 + }, + { + "epoch": 2.7299555555555557, + "grad_norm": 0.7787725329399109, + "learning_rate": 4.87901543599989e-07, + "loss": 0.2465, + "step": 3839 + }, + { + "epoch": 2.7306666666666666, + "grad_norm": 0.8419796824455261, + "learning_rate": 4.853512668166005e-07, + "loss": 0.2536, + "step": 3840 + }, + { + "epoch": 2.731377777777778, + "grad_norm": 0.738047182559967, + "learning_rate": 4.828075069184379e-07, + "loss": 0.2209, + "step": 3841 + }, + { + "epoch": 2.732088888888889, + "grad_norm": 0.7076853513717651, + "learning_rate": 4.802702656478053e-07, + "loss": 0.2398, + "step": 3842 + }, + { + "epoch": 2.7328, + "grad_norm": 0.6901286840438843, + "learning_rate": 4.777395447425404e-07, + "loss": 0.2083, + "step": 3843 + }, + { + "epoch": 2.733511111111111, + "grad_norm": 0.8334278464317322, + "learning_rate": 4.752153459360143e-07, + "loss": 0.2964, + "step": 3844 + }, + { + "epoch": 2.7342222222222223, + "grad_norm": 0.7959723472595215, + "learning_rate": 4.726976709571296e-07, + "loss": 0.247, + "step": 3845 + }, + { + "epoch": 2.734933333333333, + "grad_norm": 0.7733443975448608, + "learning_rate": 4.701865215303236e-07, + "loss": 0.2394, + "step": 3846 + }, + { + "epoch": 2.7356444444444445, + "grad_norm": 0.7509182691574097, + "learning_rate": 4.676818993755605e-07, + "loss": 0.2333, + "step": 3847 + }, + { + "epoch": 2.7363555555555554, + "grad_norm": 0.8381515145301819, + "learning_rate": 4.6518380620833694e-07, + "loss": 0.3058, + "step": 3848 + }, + { + "epoch": 2.7370666666666668, + "grad_norm": 1.0751551389694214, + "learning_rate": 4.6269224373967634e-07, + "loss": 0.2475, + "step": 3849 + }, + { + "epoch": 2.7377777777777776, + "grad_norm": 0.7437405586242676, + "learning_rate": 4.602072136761282e-07, + "loss": 0.2243, + "step": 3850 + }, + { + "epoch": 2.738488888888889, + "grad_norm": 1.070305585861206, + "learning_rate": 4.577287177197698e-07, + "loss": 0.2307, + "step": 3851 + }, + { + "epoch": 2.7392, + "grad_norm": 0.9328885674476624, + "learning_rate": 4.5525675756819987e-07, + "loss": 0.2694, + "step": 3852 + }, + { + "epoch": 2.739911111111111, + "grad_norm": 0.812667191028595, + "learning_rate": 4.5279133491454406e-07, + "loss": 0.2314, + "step": 3853 + }, + { + "epoch": 2.740622222222222, + "grad_norm": 0.7248435616493225, + "learning_rate": 4.503324514474483e-07, + "loss": 0.2325, + "step": 3854 + }, + { + "epoch": 2.7413333333333334, + "grad_norm": 0.7020560503005981, + "learning_rate": 4.478801088510798e-07, + "loss": 0.2416, + "step": 3855 + }, + { + "epoch": 2.7420444444444443, + "grad_norm": 0.7571731209754944, + "learning_rate": 4.4543430880512604e-07, + "loss": 0.236, + "step": 3856 + }, + { + "epoch": 2.7427555555555556, + "grad_norm": 0.6868278980255127, + "learning_rate": 4.429950529847926e-07, + "loss": 0.2273, + "step": 3857 + }, + { + "epoch": 2.7434666666666665, + "grad_norm": 1.0635989904403687, + "learning_rate": 4.4056234306080415e-07, + "loss": 0.2466, + "step": 3858 + }, + { + "epoch": 2.744177777777778, + "grad_norm": 0.9972834587097168, + "learning_rate": 4.3813618069940113e-07, + "loss": 0.248, + "step": 3859 + }, + { + "epoch": 2.744888888888889, + "grad_norm": 0.8927868008613586, + "learning_rate": 4.357165675623376e-07, + "loss": 0.2352, + "step": 3860 + }, + { + "epoch": 2.7456, + "grad_norm": 0.8256017565727234, + "learning_rate": 4.333035053068857e-07, + "loss": 0.2423, + "step": 3861 + }, + { + "epoch": 2.746311111111111, + "grad_norm": 0.7411656379699707, + "learning_rate": 4.3089699558582776e-07, + "loss": 0.2589, + "step": 3862 + }, + { + "epoch": 2.7470222222222223, + "grad_norm": 0.9446854591369629, + "learning_rate": 4.284970400474575e-07, + "loss": 0.2255, + "step": 3863 + }, + { + "epoch": 2.7477333333333336, + "grad_norm": 0.7672861218452454, + "learning_rate": 4.261036403355823e-07, + "loss": 0.2174, + "step": 3864 + }, + { + "epoch": 2.7484444444444445, + "grad_norm": 0.7886629104614258, + "learning_rate": 4.2371679808951737e-07, + "loss": 0.2517, + "step": 3865 + }, + { + "epoch": 2.7491555555555554, + "grad_norm": 0.821966826915741, + "learning_rate": 4.2133651494408513e-07, + "loss": 0.2408, + "step": 3866 + }, + { + "epoch": 2.7498666666666667, + "grad_norm": 0.8815449476242065, + "learning_rate": 4.189627925296202e-07, + "loss": 0.2635, + "step": 3867 + }, + { + "epoch": 2.750577777777778, + "grad_norm": 0.7052047848701477, + "learning_rate": 4.165956324719556e-07, + "loss": 0.2164, + "step": 3868 + }, + { + "epoch": 2.751288888888889, + "grad_norm": 0.7831767797470093, + "learning_rate": 4.1423503639244077e-07, + "loss": 0.2756, + "step": 3869 + }, + { + "epoch": 2.752, + "grad_norm": 0.7593711614608765, + "learning_rate": 4.1188100590791704e-07, + "loss": 0.2472, + "step": 3870 + }, + { + "epoch": 2.752711111111111, + "grad_norm": 0.7397927045822144, + "learning_rate": 4.095335426307401e-07, + "loss": 0.2631, + "step": 3871 + }, + { + "epoch": 2.7534222222222224, + "grad_norm": 0.704389750957489, + "learning_rate": 4.0719264816875713e-07, + "loss": 0.2633, + "step": 3872 + }, + { + "epoch": 2.7541333333333333, + "grad_norm": 0.7228230834007263, + "learning_rate": 4.0485832412532456e-07, + "loss": 0.2384, + "step": 3873 + }, + { + "epoch": 2.754844444444444, + "grad_norm": 0.7623085379600525, + "learning_rate": 4.0253057209929556e-07, + "loss": 0.238, + "step": 3874 + }, + { + "epoch": 2.7555555555555555, + "grad_norm": 0.8749637007713318, + "learning_rate": 4.002093936850171e-07, + "loss": 0.2812, + "step": 3875 + }, + { + "epoch": 2.756266666666667, + "grad_norm": 0.8028215169906616, + "learning_rate": 3.9789479047234293e-07, + "loss": 0.2324, + "step": 3876 + }, + { + "epoch": 2.7569777777777777, + "grad_norm": 0.6977918148040771, + "learning_rate": 3.955867640466138e-07, + "loss": 0.2387, + "step": 3877 + }, + { + "epoch": 2.7576888888888886, + "grad_norm": 0.7715423107147217, + "learning_rate": 3.9328531598867517e-07, + "loss": 0.208, + "step": 3878 + }, + { + "epoch": 2.7584, + "grad_norm": 0.6760863065719604, + "learning_rate": 3.909904478748572e-07, + "loss": 0.2123, + "step": 3879 + }, + { + "epoch": 2.7591111111111113, + "grad_norm": 0.8038487434387207, + "learning_rate": 3.887021612769937e-07, + "loss": 0.2326, + "step": 3880 + }, + { + "epoch": 2.759822222222222, + "grad_norm": 0.7836837768554688, + "learning_rate": 3.864204577623976e-07, + "loss": 0.2272, + "step": 3881 + }, + { + "epoch": 2.760533333333333, + "grad_norm": 0.7355610728263855, + "learning_rate": 3.841453388938876e-07, + "loss": 0.251, + "step": 3882 + }, + { + "epoch": 2.7612444444444444, + "grad_norm": 0.7760096788406372, + "learning_rate": 3.818768062297573e-07, + "loss": 0.2548, + "step": 3883 + }, + { + "epoch": 2.7619555555555557, + "grad_norm": 0.9365607500076294, + "learning_rate": 3.7961486132380487e-07, + "loss": 0.2451, + "step": 3884 + }, + { + "epoch": 2.7626666666666666, + "grad_norm": 0.76507169008255, + "learning_rate": 3.7735950572530213e-07, + "loss": 0.2251, + "step": 3885 + }, + { + "epoch": 2.763377777777778, + "grad_norm": 0.7453303337097168, + "learning_rate": 3.7511074097901557e-07, + "loss": 0.2328, + "step": 3886 + }, + { + "epoch": 2.764088888888889, + "grad_norm": 0.8934824466705322, + "learning_rate": 3.7286856862519647e-07, + "loss": 0.2542, + "step": 3887 + }, + { + "epoch": 2.7648, + "grad_norm": 0.760960578918457, + "learning_rate": 3.7063299019957867e-07, + "loss": 0.2437, + "step": 3888 + }, + { + "epoch": 2.765511111111111, + "grad_norm": 0.6528473496437073, + "learning_rate": 3.6840400723338274e-07, + "loss": 0.2124, + "step": 3889 + }, + { + "epoch": 2.7662222222222224, + "grad_norm": 0.7588576078414917, + "learning_rate": 3.661816212533076e-07, + "loss": 0.2368, + "step": 3890 + }, + { + "epoch": 2.7669333333333332, + "grad_norm": 0.7926591038703918, + "learning_rate": 3.6396583378153773e-07, + "loss": 0.2378, + "step": 3891 + }, + { + "epoch": 2.7676444444444446, + "grad_norm": 0.8089482188224792, + "learning_rate": 3.617566463357336e-07, + "loss": 0.2316, + "step": 3892 + }, + { + "epoch": 2.7683555555555555, + "grad_norm": 0.7289047837257385, + "learning_rate": 3.595540604290437e-07, + "loss": 0.2098, + "step": 3893 + }, + { + "epoch": 2.769066666666667, + "grad_norm": 0.7840014100074768, + "learning_rate": 3.5735807757008354e-07, + "loss": 0.2281, + "step": 3894 + }, + { + "epoch": 2.7697777777777777, + "grad_norm": 0.7807803750038147, + "learning_rate": 3.551686992629533e-07, + "loss": 0.2407, + "step": 3895 + }, + { + "epoch": 2.770488888888889, + "grad_norm": 0.7170054912567139, + "learning_rate": 3.529859270072289e-07, + "loss": 0.2305, + "step": 3896 + }, + { + "epoch": 2.7712, + "grad_norm": 0.7762348055839539, + "learning_rate": 3.508097622979589e-07, + "loss": 0.247, + "step": 3897 + }, + { + "epoch": 2.771911111111111, + "grad_norm": 0.9453489184379578, + "learning_rate": 3.4864020662566775e-07, + "loss": 0.2163, + "step": 3898 + }, + { + "epoch": 2.772622222222222, + "grad_norm": 0.8619599342346191, + "learning_rate": 3.464772614763534e-07, + "loss": 0.2305, + "step": 3899 + }, + { + "epoch": 2.7733333333333334, + "grad_norm": 0.7582694888114929, + "learning_rate": 3.443209283314863e-07, + "loss": 0.2626, + "step": 3900 + }, + { + "epoch": 2.7740444444444443, + "grad_norm": 0.782243549823761, + "learning_rate": 3.4217120866800733e-07, + "loss": 0.2577, + "step": 3901 + }, + { + "epoch": 2.7747555555555556, + "grad_norm": 0.823579728603363, + "learning_rate": 3.4002810395832753e-07, + "loss": 0.2255, + "step": 3902 + }, + { + "epoch": 2.7754666666666665, + "grad_norm": 0.723082959651947, + "learning_rate": 3.3789161567032604e-07, + "loss": 0.2359, + "step": 3903 + }, + { + "epoch": 2.776177777777778, + "grad_norm": 0.6456478834152222, + "learning_rate": 3.357617452673545e-07, + "loss": 0.218, + "step": 3904 + }, + { + "epoch": 2.7768888888888887, + "grad_norm": 0.6962716579437256, + "learning_rate": 3.33638494208226e-07, + "loss": 0.2437, + "step": 3905 + }, + { + "epoch": 2.7776, + "grad_norm": 0.7217891812324524, + "learning_rate": 3.3152186394722506e-07, + "loss": 0.2325, + "step": 3906 + }, + { + "epoch": 2.778311111111111, + "grad_norm": 0.9268063306808472, + "learning_rate": 3.2941185593409755e-07, + "loss": 0.2374, + "step": 3907 + }, + { + "epoch": 2.7790222222222223, + "grad_norm": 0.7258709669113159, + "learning_rate": 3.27308471614054e-07, + "loss": 0.2113, + "step": 3908 + }, + { + "epoch": 2.779733333333333, + "grad_norm": 0.7328567504882812, + "learning_rate": 3.252117124277721e-07, + "loss": 0.194, + "step": 3909 + }, + { + "epoch": 2.7804444444444445, + "grad_norm": 0.692202627658844, + "learning_rate": 3.2312157981138626e-07, + "loss": 0.2444, + "step": 3910 + }, + { + "epoch": 2.7811555555555554, + "grad_norm": 0.7437503337860107, + "learning_rate": 3.210380751964959e-07, + "loss": 0.2435, + "step": 3911 + }, + { + "epoch": 2.7818666666666667, + "grad_norm": 0.8119533061981201, + "learning_rate": 3.189612000101594e-07, + "loss": 0.2918, + "step": 3912 + }, + { + "epoch": 2.782577777777778, + "grad_norm": 0.720231831073761, + "learning_rate": 3.1689095567489335e-07, + "loss": 0.2854, + "step": 3913 + }, + { + "epoch": 2.783288888888889, + "grad_norm": 0.7610108256340027, + "learning_rate": 3.148273436086757e-07, + "loss": 0.2624, + "step": 3914 + }, + { + "epoch": 2.784, + "grad_norm": 0.7235201001167297, + "learning_rate": 3.127703652249392e-07, + "loss": 0.2263, + "step": 3915 + }, + { + "epoch": 2.784711111111111, + "grad_norm": 0.8396919369697571, + "learning_rate": 3.107200219325746e-07, + "loss": 0.2638, + "step": 3916 + }, + { + "epoch": 2.7854222222222225, + "grad_norm": 0.691450834274292, + "learning_rate": 3.0867631513592513e-07, + "loss": 0.2258, + "step": 3917 + }, + { + "epoch": 2.7861333333333334, + "grad_norm": 0.8500252366065979, + "learning_rate": 3.0663924623479337e-07, + "loss": 0.2174, + "step": 3918 + }, + { + "epoch": 2.7868444444444442, + "grad_norm": 0.6777299046516418, + "learning_rate": 3.046088166244276e-07, + "loss": 0.2467, + "step": 3919 + }, + { + "epoch": 2.7875555555555556, + "grad_norm": 0.6786453127861023, + "learning_rate": 3.0258502769553996e-07, + "loss": 0.2299, + "step": 3920 + }, + { + "epoch": 2.788266666666667, + "grad_norm": 0.7209764122962952, + "learning_rate": 3.005678808342816e-07, + "loss": 0.2315, + "step": 3921 + }, + { + "epoch": 2.7889777777777778, + "grad_norm": 0.8441829085350037, + "learning_rate": 2.985573774222661e-07, + "loss": 0.226, + "step": 3922 + }, + { + "epoch": 2.7896888888888887, + "grad_norm": 0.8074929118156433, + "learning_rate": 2.9655351883654646e-07, + "loss": 0.2577, + "step": 3923 + }, + { + "epoch": 2.7904, + "grad_norm": 0.7209919095039368, + "learning_rate": 2.945563064496326e-07, + "loss": 0.2701, + "step": 3924 + }, + { + "epoch": 2.7911111111111113, + "grad_norm": 0.7438526749610901, + "learning_rate": 2.925657416294758e-07, + "loss": 0.2225, + "step": 3925 + }, + { + "epoch": 2.791822222222222, + "grad_norm": 0.7262961864471436, + "learning_rate": 2.905818257394799e-07, + "loss": 0.2379, + "step": 3926 + }, + { + "epoch": 2.792533333333333, + "grad_norm": 0.7506723403930664, + "learning_rate": 2.8860456013849126e-07, + "loss": 0.2174, + "step": 3927 + }, + { + "epoch": 2.7932444444444444, + "grad_norm": 0.8771202564239502, + "learning_rate": 2.8663394618079875e-07, + "loss": 0.2674, + "step": 3928 + }, + { + "epoch": 2.7939555555555557, + "grad_norm": 0.7342710494995117, + "learning_rate": 2.846699852161439e-07, + "loss": 0.2408, + "step": 3929 + }, + { + "epoch": 2.7946666666666666, + "grad_norm": 0.8700303435325623, + "learning_rate": 2.827126785897005e-07, + "loss": 0.2497, + "step": 3930 + }, + { + "epoch": 2.7953777777777775, + "grad_norm": 0.7819193601608276, + "learning_rate": 2.8076202764209526e-07, + "loss": 0.2413, + "step": 3931 + }, + { + "epoch": 2.796088888888889, + "grad_norm": 0.7479329109191895, + "learning_rate": 2.78818033709386e-07, + "loss": 0.2476, + "step": 3932 + }, + { + "epoch": 2.7968, + "grad_norm": 0.7391296029090881, + "learning_rate": 2.7688069812308095e-07, + "loss": 0.2454, + "step": 3933 + }, + { + "epoch": 2.797511111111111, + "grad_norm": 0.7644091844558716, + "learning_rate": 2.7495002221011757e-07, + "loss": 0.2621, + "step": 3934 + }, + { + "epoch": 2.7982222222222224, + "grad_norm": 0.8075675368309021, + "learning_rate": 2.730260072928825e-07, + "loss": 0.2819, + "step": 3935 + }, + { + "epoch": 2.7989333333333333, + "grad_norm": 0.7469271421432495, + "learning_rate": 2.7110865468919057e-07, + "loss": 0.2429, + "step": 3936 + }, + { + "epoch": 2.7996444444444446, + "grad_norm": 0.9127926230430603, + "learning_rate": 2.6919796571229894e-07, + "loss": 0.2267, + "step": 3937 + }, + { + "epoch": 2.8003555555555555, + "grad_norm": 0.7835930585861206, + "learning_rate": 2.672939416708986e-07, + "loss": 0.2322, + "step": 3938 + }, + { + "epoch": 2.801066666666667, + "grad_norm": 0.9110485315322876, + "learning_rate": 2.653965838691164e-07, + "loss": 0.2452, + "step": 3939 + }, + { + "epoch": 2.8017777777777777, + "grad_norm": 0.6781100034713745, + "learning_rate": 2.635058936065138e-07, + "loss": 0.2214, + "step": 3940 + }, + { + "epoch": 2.802488888888889, + "grad_norm": 0.8788836598396301, + "learning_rate": 2.6162187217808497e-07, + "loss": 0.2914, + "step": 3941 + }, + { + "epoch": 2.8032, + "grad_norm": 0.7158138751983643, + "learning_rate": 2.5974452087425437e-07, + "loss": 0.2595, + "step": 3942 + }, + { + "epoch": 2.8039111111111112, + "grad_norm": 1.0095617771148682, + "learning_rate": 2.57873840980879e-07, + "loss": 0.2122, + "step": 3943 + }, + { + "epoch": 2.804622222222222, + "grad_norm": 0.7524649500846863, + "learning_rate": 2.5600983377925046e-07, + "loss": 0.2675, + "step": 3944 + }, + { + "epoch": 2.8053333333333335, + "grad_norm": 0.6945964694023132, + "learning_rate": 2.541525005460821e-07, + "loss": 0.2204, + "step": 3945 + }, + { + "epoch": 2.8060444444444443, + "grad_norm": 0.7184180021286011, + "learning_rate": 2.523018425535251e-07, + "loss": 0.2372, + "step": 3946 + }, + { + "epoch": 2.8067555555555557, + "grad_norm": 0.6773695349693298, + "learning_rate": 2.5045786106915103e-07, + "loss": 0.2344, + "step": 3947 + }, + { + "epoch": 2.8074666666666666, + "grad_norm": 0.704878032207489, + "learning_rate": 2.486205573559608e-07, + "loss": 0.2045, + "step": 3948 + }, + { + "epoch": 2.808177777777778, + "grad_norm": 0.6579682230949402, + "learning_rate": 2.4678993267238436e-07, + "loss": 0.2031, + "step": 3949 + }, + { + "epoch": 2.8088888888888888, + "grad_norm": 0.8020352721214294, + "learning_rate": 2.4496598827227213e-07, + "loss": 0.2533, + "step": 3950 + }, + { + "epoch": 2.8096, + "grad_norm": 0.7732877731323242, + "learning_rate": 2.4314872540490365e-07, + "loss": 0.2752, + "step": 3951 + }, + { + "epoch": 2.810311111111111, + "grad_norm": 0.7987539768218994, + "learning_rate": 2.413381453149799e-07, + "loss": 0.2375, + "step": 3952 + }, + { + "epoch": 2.8110222222222223, + "grad_norm": 0.7538567781448364, + "learning_rate": 2.3953424924262336e-07, + "loss": 0.2134, + "step": 3953 + }, + { + "epoch": 2.811733333333333, + "grad_norm": 0.7594051957130432, + "learning_rate": 2.3773703842338125e-07, + "loss": 0.262, + "step": 3954 + }, + { + "epoch": 2.8124444444444445, + "grad_norm": 0.7859817147254944, + "learning_rate": 2.3594651408822e-07, + "loss": 0.2412, + "step": 3955 + }, + { + "epoch": 2.8131555555555554, + "grad_norm": 0.9182428121566772, + "learning_rate": 2.3416267746352528e-07, + "loss": 0.2361, + "step": 3956 + }, + { + "epoch": 2.8138666666666667, + "grad_norm": 0.7234243750572205, + "learning_rate": 2.323855297711053e-07, + "loss": 0.2508, + "step": 3957 + }, + { + "epoch": 2.8145777777777776, + "grad_norm": 0.8114535808563232, + "learning_rate": 2.3061507222818303e-07, + "loss": 0.2559, + "step": 3958 + }, + { + "epoch": 2.815288888888889, + "grad_norm": 0.7948520183563232, + "learning_rate": 2.2885130604740292e-07, + "loss": 0.2189, + "step": 3959 + }, + { + "epoch": 2.816, + "grad_norm": 0.7426165342330933, + "learning_rate": 2.2709423243682416e-07, + "loss": 0.2136, + "step": 3960 + }, + { + "epoch": 2.816711111111111, + "grad_norm": 0.6879301071166992, + "learning_rate": 2.2534385259992186e-07, + "loss": 0.2279, + "step": 3961 + }, + { + "epoch": 2.8174222222222225, + "grad_norm": 0.8902060985565186, + "learning_rate": 2.23600167735587e-07, + "loss": 0.2801, + "step": 3962 + }, + { + "epoch": 2.8181333333333334, + "grad_norm": 0.733044445514679, + "learning_rate": 2.2186317903812426e-07, + "loss": 0.2268, + "step": 3963 + }, + { + "epoch": 2.8188444444444443, + "grad_norm": 0.7688709497451782, + "learning_rate": 2.2013288769725194e-07, + "loss": 0.2123, + "step": 3964 + }, + { + "epoch": 2.8195555555555556, + "grad_norm": 0.843580961227417, + "learning_rate": 2.1840929489810315e-07, + "loss": 0.2717, + "step": 3965 + }, + { + "epoch": 2.820266666666667, + "grad_norm": 0.8445359468460083, + "learning_rate": 2.166924018212202e-07, + "loss": 0.25, + "step": 3966 + }, + { + "epoch": 2.820977777777778, + "grad_norm": 0.7432076334953308, + "learning_rate": 2.1498220964255912e-07, + "loss": 0.2189, + "step": 3967 + }, + { + "epoch": 2.8216888888888887, + "grad_norm": 0.7160336375236511, + "learning_rate": 2.132787195334829e-07, + "loss": 0.2184, + "step": 3968 + }, + { + "epoch": 2.8224, + "grad_norm": 0.6729885339736938, + "learning_rate": 2.1158193266076822e-07, + "loss": 0.2233, + "step": 3969 + }, + { + "epoch": 2.8231111111111113, + "grad_norm": 0.7733002305030823, + "learning_rate": 2.0989185018659431e-07, + "loss": 0.2694, + "step": 3970 + }, + { + "epoch": 2.8238222222222222, + "grad_norm": 0.7478377819061279, + "learning_rate": 2.082084732685574e-07, + "loss": 0.2522, + "step": 3971 + }, + { + "epoch": 2.824533333333333, + "grad_norm": 0.7440409064292908, + "learning_rate": 2.0653180305965194e-07, + "loss": 0.2479, + "step": 3972 + }, + { + "epoch": 2.8252444444444444, + "grad_norm": 0.7322220802307129, + "learning_rate": 2.0486184070828585e-07, + "loss": 0.2308, + "step": 3973 + }, + { + "epoch": 2.8259555555555558, + "grad_norm": 0.790088415145874, + "learning_rate": 2.0319858735826648e-07, + "loss": 0.2622, + "step": 3974 + }, + { + "epoch": 2.8266666666666667, + "grad_norm": 0.7627602219581604, + "learning_rate": 2.015420441488114e-07, + "loss": 0.2161, + "step": 3975 + }, + { + "epoch": 2.8273777777777775, + "grad_norm": 1.1273821592330933, + "learning_rate": 1.9989221221453746e-07, + "loss": 0.2543, + "step": 3976 + }, + { + "epoch": 2.828088888888889, + "grad_norm": 0.6661327481269836, + "learning_rate": 1.9824909268546854e-07, + "loss": 0.2359, + "step": 3977 + }, + { + "epoch": 2.8288, + "grad_norm": 0.7240387201309204, + "learning_rate": 1.966126866870277e-07, + "loss": 0.2365, + "step": 3978 + }, + { + "epoch": 2.829511111111111, + "grad_norm": 0.8152077794075012, + "learning_rate": 1.9498299534004395e-07, + "loss": 0.2562, + "step": 3979 + }, + { + "epoch": 2.830222222222222, + "grad_norm": 0.64036625623703, + "learning_rate": 1.9336001976074326e-07, + "loss": 0.2191, + "step": 3980 + }, + { + "epoch": 2.8309333333333333, + "grad_norm": 1.1448575258255005, + "learning_rate": 1.91743761060752e-07, + "loss": 0.2502, + "step": 3981 + }, + { + "epoch": 2.8316444444444446, + "grad_norm": 0.8239279389381409, + "learning_rate": 1.9013422034710016e-07, + "loss": 0.2721, + "step": 3982 + }, + { + "epoch": 2.8323555555555555, + "grad_norm": 0.7425834536552429, + "learning_rate": 1.8853139872220927e-07, + "loss": 0.2307, + "step": 3983 + }, + { + "epoch": 2.8330666666666664, + "grad_norm": 0.7114003896713257, + "learning_rate": 1.869352972839067e-07, + "loss": 0.2365, + "step": 3984 + }, + { + "epoch": 2.8337777777777777, + "grad_norm": 0.625544548034668, + "learning_rate": 1.8534591712541018e-07, + "loss": 0.2255, + "step": 3985 + }, + { + "epoch": 2.834488888888889, + "grad_norm": 0.8608654737472534, + "learning_rate": 1.837632593353389e-07, + "loss": 0.2593, + "step": 3986 + }, + { + "epoch": 2.8352, + "grad_norm": 0.829176127910614, + "learning_rate": 1.8218732499770354e-07, + "loss": 0.2848, + "step": 3987 + }, + { + "epoch": 2.8359111111111113, + "grad_norm": 0.7615945339202881, + "learning_rate": 1.8061811519191287e-07, + "loss": 0.2408, + "step": 3988 + }, + { + "epoch": 2.836622222222222, + "grad_norm": 0.9380969405174255, + "learning_rate": 1.7905563099276824e-07, + "loss": 0.2691, + "step": 3989 + }, + { + "epoch": 2.8373333333333335, + "grad_norm": 0.7316835522651672, + "learning_rate": 1.7749987347046471e-07, + "loss": 0.2068, + "step": 3990 + }, + { + "epoch": 2.8380444444444444, + "grad_norm": 0.6602096557617188, + "learning_rate": 1.759508436905888e-07, + "loss": 0.2386, + "step": 3991 + }, + { + "epoch": 2.8387555555555557, + "grad_norm": 0.7777907252311707, + "learning_rate": 1.7440854271412288e-07, + "loss": 0.2226, + "step": 3992 + }, + { + "epoch": 2.8394666666666666, + "grad_norm": 0.805792510509491, + "learning_rate": 1.7287297159743533e-07, + "loss": 0.2716, + "step": 3993 + }, + { + "epoch": 2.840177777777778, + "grad_norm": 0.7894183993339539, + "learning_rate": 1.7134413139228812e-07, + "loss": 0.2416, + "step": 3994 + }, + { + "epoch": 2.840888888888889, + "grad_norm": 0.7818019986152649, + "learning_rate": 1.6982202314583475e-07, + "loss": 0.2615, + "step": 3995 + }, + { + "epoch": 2.8416, + "grad_norm": 1.0053645372390747, + "learning_rate": 1.6830664790061124e-07, + "loss": 0.2876, + "step": 3996 + }, + { + "epoch": 2.842311111111111, + "grad_norm": 0.8720224499702454, + "learning_rate": 1.6679800669455072e-07, + "loss": 0.2126, + "step": 3997 + }, + { + "epoch": 2.8430222222222223, + "grad_norm": 0.7764236330986023, + "learning_rate": 1.6529610056096768e-07, + "loss": 0.2402, + "step": 3998 + }, + { + "epoch": 2.8437333333333332, + "grad_norm": 0.7535574436187744, + "learning_rate": 1.6380093052856482e-07, + "loss": 0.2412, + "step": 3999 + }, + { + "epoch": 2.8444444444444446, + "grad_norm": 0.7100918889045715, + "learning_rate": 1.6231249762143187e-07, + "loss": 0.2499, + "step": 4000 + }, + { + "epoch": 2.8451555555555554, + "grad_norm": 0.7520055174827576, + "learning_rate": 1.6083080285904328e-07, + "loss": 0.2136, + "step": 4001 + }, + { + "epoch": 2.8458666666666668, + "grad_norm": 0.7084394097328186, + "learning_rate": 1.5935584725626062e-07, + "loss": 0.2551, + "step": 4002 + }, + { + "epoch": 2.8465777777777777, + "grad_norm": 0.7877987027168274, + "learning_rate": 1.578876318233258e-07, + "loss": 0.2569, + "step": 4003 + }, + { + "epoch": 2.847288888888889, + "grad_norm": 0.7148776650428772, + "learning_rate": 1.5642615756586765e-07, + "loss": 0.2517, + "step": 4004 + }, + { + "epoch": 2.848, + "grad_norm": 0.8360061645507812, + "learning_rate": 1.5497142548489552e-07, + "loss": 0.2805, + "step": 4005 + }, + { + "epoch": 2.848711111111111, + "grad_norm": 0.6980102062225342, + "learning_rate": 1.5352343657680234e-07, + "loss": 0.2684, + "step": 4006 + }, + { + "epoch": 2.849422222222222, + "grad_norm": 0.755066990852356, + "learning_rate": 1.520821918333615e-07, + "loss": 0.235, + "step": 4007 + }, + { + "epoch": 2.8501333333333334, + "grad_norm": 1.0296367406845093, + "learning_rate": 1.506476922417266e-07, + "loss": 0.2313, + "step": 4008 + }, + { + "epoch": 2.8508444444444443, + "grad_norm": 0.7743528485298157, + "learning_rate": 1.4921993878443287e-07, + "loss": 0.2036, + "step": 4009 + }, + { + "epoch": 2.8515555555555556, + "grad_norm": 0.7376415729522705, + "learning_rate": 1.4779893243939358e-07, + "loss": 0.2417, + "step": 4010 + }, + { + "epoch": 2.8522666666666665, + "grad_norm": 0.6917979121208191, + "learning_rate": 1.4638467417990133e-07, + "loss": 0.2389, + "step": 4011 + }, + { + "epoch": 2.852977777777778, + "grad_norm": 0.6671186685562134, + "learning_rate": 1.4497716497462676e-07, + "loss": 0.223, + "step": 4012 + }, + { + "epoch": 2.8536888888888887, + "grad_norm": 1.0391956567764282, + "learning_rate": 1.4357640578761655e-07, + "loss": 0.2311, + "step": 4013 + }, + { + "epoch": 2.8544, + "grad_norm": 0.6257952451705933, + "learning_rate": 1.4218239757829656e-07, + "loss": 0.2285, + "step": 4014 + }, + { + "epoch": 2.8551111111111114, + "grad_norm": 1.067718505859375, + "learning_rate": 1.4079514130146632e-07, + "loss": 0.216, + "step": 4015 + }, + { + "epoch": 2.8558222222222223, + "grad_norm": 0.8114745020866394, + "learning_rate": 1.3941463790730248e-07, + "loss": 0.2193, + "step": 4016 + }, + { + "epoch": 2.856533333333333, + "grad_norm": 0.8659932613372803, + "learning_rate": 1.380408883413553e-07, + "loss": 0.2582, + "step": 4017 + }, + { + "epoch": 2.8572444444444445, + "grad_norm": 0.7676635980606079, + "learning_rate": 1.3667389354454997e-07, + "loss": 0.2479, + "step": 4018 + }, + { + "epoch": 2.857955555555556, + "grad_norm": 0.714787483215332, + "learning_rate": 1.3531365445318301e-07, + "loss": 0.2155, + "step": 4019 + }, + { + "epoch": 2.8586666666666667, + "grad_norm": 0.753187358379364, + "learning_rate": 1.3396017199892808e-07, + "loss": 0.2484, + "step": 4020 + }, + { + "epoch": 2.8593777777777776, + "grad_norm": 0.8648319244384766, + "learning_rate": 1.3261344710882584e-07, + "loss": 0.2547, + "step": 4021 + }, + { + "epoch": 2.860088888888889, + "grad_norm": 0.9470051527023315, + "learning_rate": 1.312734807052929e-07, + "loss": 0.248, + "step": 4022 + }, + { + "epoch": 2.8608000000000002, + "grad_norm": 0.8101348280906677, + "learning_rate": 1.2994027370611173e-07, + "loss": 0.2738, + "step": 4023 + }, + { + "epoch": 2.861511111111111, + "grad_norm": 0.6414603590965271, + "learning_rate": 1.2861382702444304e-07, + "loss": 0.2139, + "step": 4024 + }, + { + "epoch": 2.862222222222222, + "grad_norm": 0.7470347881317139, + "learning_rate": 1.272941415688067e-07, + "loss": 0.247, + "step": 4025 + }, + { + "epoch": 2.8629333333333333, + "grad_norm": 0.8031718730926514, + "learning_rate": 1.2598121824310305e-07, + "loss": 0.2291, + "step": 4026 + }, + { + "epoch": 2.8636444444444447, + "grad_norm": 0.7824362516403198, + "learning_rate": 1.246750579465894e-07, + "loss": 0.2561, + "step": 4027 + }, + { + "epoch": 2.8643555555555555, + "grad_norm": 0.8182615637779236, + "learning_rate": 1.2337566157390124e-07, + "loss": 0.2533, + "step": 4028 + }, + { + "epoch": 2.8650666666666664, + "grad_norm": 0.705797553062439, + "learning_rate": 1.2208303001503219e-07, + "loss": 0.2331, + "step": 4029 + }, + { + "epoch": 2.8657777777777778, + "grad_norm": 0.6509291529655457, + "learning_rate": 1.2079716415534958e-07, + "loss": 0.2161, + "step": 4030 + }, + { + "epoch": 2.866488888888889, + "grad_norm": 0.7072452902793884, + "learning_rate": 1.195180648755845e-07, + "loss": 0.2133, + "step": 4031 + }, + { + "epoch": 2.8672, + "grad_norm": 0.719694197177887, + "learning_rate": 1.1824573305182829e-07, + "loss": 0.2423, + "step": 4032 + }, + { + "epoch": 2.867911111111111, + "grad_norm": 0.6339098811149597, + "learning_rate": 1.1698016955554725e-07, + "loss": 0.2014, + "step": 4033 + }, + { + "epoch": 2.868622222222222, + "grad_norm": 0.785672664642334, + "learning_rate": 1.1572137525356019e-07, + "loss": 0.2306, + "step": 4034 + }, + { + "epoch": 2.8693333333333335, + "grad_norm": 0.7842385768890381, + "learning_rate": 1.1446935100806078e-07, + "loss": 0.2088, + "step": 4035 + }, + { + "epoch": 2.8700444444444444, + "grad_norm": 0.7443892359733582, + "learning_rate": 1.1322409767659526e-07, + "loss": 0.2265, + "step": 4036 + }, + { + "epoch": 2.8707555555555553, + "grad_norm": 0.7398084402084351, + "learning_rate": 1.1198561611208247e-07, + "loss": 0.2405, + "step": 4037 + }, + { + "epoch": 2.8714666666666666, + "grad_norm": 0.7655425667762756, + "learning_rate": 1.1075390716279167e-07, + "loss": 0.2711, + "step": 4038 + }, + { + "epoch": 2.872177777777778, + "grad_norm": 0.7643004059791565, + "learning_rate": 1.0952897167236576e-07, + "loss": 0.284, + "step": 4039 + }, + { + "epoch": 2.872888888888889, + "grad_norm": 0.7536296248435974, + "learning_rate": 1.0831081047979585e-07, + "loss": 0.251, + "step": 4040 + }, + { + "epoch": 2.8736, + "grad_norm": 0.8775938749313354, + "learning_rate": 1.0709942441944454e-07, + "loss": 0.2337, + "step": 4041 + }, + { + "epoch": 2.874311111111111, + "grad_norm": 0.768210232257843, + "learning_rate": 1.0589481432102588e-07, + "loss": 0.2313, + "step": 4042 + }, + { + "epoch": 2.8750222222222224, + "grad_norm": 0.7359346151351929, + "learning_rate": 1.0469698100961545e-07, + "loss": 0.2257, + "step": 4043 + }, + { + "epoch": 2.8757333333333333, + "grad_norm": 0.673621416091919, + "learning_rate": 1.0350592530564919e-07, + "loss": 0.2278, + "step": 4044 + }, + { + "epoch": 2.8764444444444446, + "grad_norm": 0.7693796753883362, + "learning_rate": 1.0232164802491673e-07, + "loss": 0.2444, + "step": 4045 + }, + { + "epoch": 2.8771555555555555, + "grad_norm": 0.8311963677406311, + "learning_rate": 1.0114414997856814e-07, + "loss": 0.2626, + "step": 4046 + }, + { + "epoch": 2.877866666666667, + "grad_norm": 0.7847995758056641, + "learning_rate": 9.997343197310937e-08, + "loss": 0.255, + "step": 4047 + }, + { + "epoch": 2.8785777777777777, + "grad_norm": 0.712587833404541, + "learning_rate": 9.880949481040347e-08, + "loss": 0.2434, + "step": 4048 + }, + { + "epoch": 2.879288888888889, + "grad_norm": 0.7481316328048706, + "learning_rate": 9.765233928766493e-08, + "loss": 0.2349, + "step": 4049 + }, + { + "epoch": 2.88, + "grad_norm": 0.77059006690979, + "learning_rate": 9.650196619747088e-08, + "loss": 0.2482, + "step": 4050 + }, + { + "epoch": 2.8807111111111112, + "grad_norm": 0.8342442512512207, + "learning_rate": 9.535837632774436e-08, + "loss": 0.2655, + "step": 4051 + }, + { + "epoch": 2.881422222222222, + "grad_norm": 0.7040227055549622, + "learning_rate": 9.422157046176772e-08, + "loss": 0.1946, + "step": 4052 + }, + { + "epoch": 2.8821333333333334, + "grad_norm": 0.846810519695282, + "learning_rate": 9.309154937817589e-08, + "loss": 0.2462, + "step": 4053 + }, + { + "epoch": 2.8828444444444443, + "grad_norm": 0.7049356698989868, + "learning_rate": 9.19683138509564e-08, + "loss": 0.2279, + "step": 4054 + }, + { + "epoch": 2.8835555555555556, + "grad_norm": 0.9012750387191772, + "learning_rate": 9.085186464944828e-08, + "loss": 0.2464, + "step": 4055 + }, + { + "epoch": 2.8842666666666665, + "grad_norm": 0.8118312954902649, + "learning_rate": 8.974220253834209e-08, + "loss": 0.2437, + "step": 4056 + }, + { + "epoch": 2.884977777777778, + "grad_norm": 0.7107411026954651, + "learning_rate": 8.863932827768318e-08, + "loss": 0.2382, + "step": 4057 + }, + { + "epoch": 2.8856888888888887, + "grad_norm": 0.7071009874343872, + "learning_rate": 8.754324262286284e-08, + "loss": 0.254, + "step": 4058 + }, + { + "epoch": 2.8864, + "grad_norm": 0.8001231551170349, + "learning_rate": 8.645394632462722e-08, + "loss": 0.2369, + "step": 4059 + }, + { + "epoch": 2.887111111111111, + "grad_norm": 0.8696957230567932, + "learning_rate": 8.537144012906728e-08, + "loss": 0.2482, + "step": 4060 + }, + { + "epoch": 2.8878222222222223, + "grad_norm": 0.7909144759178162, + "learning_rate": 8.429572477762771e-08, + "loss": 0.2379, + "step": 4061 + }, + { + "epoch": 2.888533333333333, + "grad_norm": 0.7825571298599243, + "learning_rate": 8.322680100710023e-08, + "loss": 0.2369, + "step": 4062 + }, + { + "epoch": 2.8892444444444445, + "grad_norm": 0.7649456262588501, + "learning_rate": 8.216466954962365e-08, + "loss": 0.2662, + "step": 4063 + }, + { + "epoch": 2.889955555555556, + "grad_norm": 0.7349497079849243, + "learning_rate": 8.110933113268604e-08, + "loss": 0.2437, + "step": 4064 + }, + { + "epoch": 2.8906666666666667, + "grad_norm": 0.7409260272979736, + "learning_rate": 8.006078647912252e-08, + "loss": 0.2091, + "step": 4065 + }, + { + "epoch": 2.8913777777777776, + "grad_norm": 0.6903036832809448, + "learning_rate": 7.901903630711416e-08, + "loss": 0.2548, + "step": 4066 + }, + { + "epoch": 2.892088888888889, + "grad_norm": 0.8919834494590759, + "learning_rate": 7.798408133018908e-08, + "loss": 0.2066, + "step": 4067 + }, + { + "epoch": 2.8928000000000003, + "grad_norm": 0.8072580099105835, + "learning_rate": 7.695592225722137e-08, + "loss": 0.2804, + "step": 4068 + }, + { + "epoch": 2.893511111111111, + "grad_norm": 0.7057785391807556, + "learning_rate": 7.59345597924277e-08, + "loss": 0.2458, + "step": 4069 + }, + { + "epoch": 2.894222222222222, + "grad_norm": 0.6738962531089783, + "learning_rate": 7.491999463537403e-08, + "loss": 0.2512, + "step": 4070 + }, + { + "epoch": 2.8949333333333334, + "grad_norm": 0.7081482410430908, + "learning_rate": 7.391222748096783e-08, + "loss": 0.2314, + "step": 4071 + }, + { + "epoch": 2.8956444444444447, + "grad_norm": 0.8884003162384033, + "learning_rate": 7.291125901946027e-08, + "loss": 0.2771, + "step": 4072 + }, + { + "epoch": 2.8963555555555556, + "grad_norm": 0.7267641425132751, + "learning_rate": 7.19170899364463e-08, + "loss": 0.243, + "step": 4073 + }, + { + "epoch": 2.8970666666666665, + "grad_norm": 0.7329164743423462, + "learning_rate": 7.092972091286454e-08, + "loss": 0.2862, + "step": 4074 + }, + { + "epoch": 2.897777777777778, + "grad_norm": 0.7838193774223328, + "learning_rate": 6.994915262499513e-08, + "loss": 0.2821, + "step": 4075 + }, + { + "epoch": 2.898488888888889, + "grad_norm": 0.654632568359375, + "learning_rate": 6.897538574445972e-08, + "loss": 0.2276, + "step": 4076 + }, + { + "epoch": 2.8992, + "grad_norm": 0.7694046497344971, + "learning_rate": 6.800842093822369e-08, + "loss": 0.2832, + "step": 4077 + }, + { + "epoch": 2.899911111111111, + "grad_norm": 0.7550276517868042, + "learning_rate": 6.704825886858946e-08, + "loss": 0.2324, + "step": 4078 + }, + { + "epoch": 2.900622222222222, + "grad_norm": 0.6601117849349976, + "learning_rate": 6.609490019320653e-08, + "loss": 0.2194, + "step": 4079 + }, + { + "epoch": 2.9013333333333335, + "grad_norm": 0.6832478642463684, + "learning_rate": 6.5148345565057e-08, + "loss": 0.2459, + "step": 4080 + }, + { + "epoch": 2.9020444444444444, + "grad_norm": 0.807975709438324, + "learning_rate": 6.420859563246673e-08, + "loss": 0.2464, + "step": 4081 + }, + { + "epoch": 2.9027555555555553, + "grad_norm": 0.7746208310127258, + "learning_rate": 6.327565103910193e-08, + "loss": 0.2328, + "step": 4082 + }, + { + "epoch": 2.9034666666666666, + "grad_norm": 0.7866553068161011, + "learning_rate": 6.234951242396592e-08, + "loss": 0.2403, + "step": 4083 + }, + { + "epoch": 2.904177777777778, + "grad_norm": 0.9155770540237427, + "learning_rate": 6.143018042139903e-08, + "loss": 0.2237, + "step": 4084 + }, + { + "epoch": 2.904888888888889, + "grad_norm": 0.7667056322097778, + "learning_rate": 6.051765566108092e-08, + "loss": 0.2422, + "step": 4085 + }, + { + "epoch": 2.9055999999999997, + "grad_norm": 0.7932671904563904, + "learning_rate": 5.96119387680294e-08, + "loss": 0.2589, + "step": 4086 + }, + { + "epoch": 2.906311111111111, + "grad_norm": 0.7859063148498535, + "learning_rate": 5.871303036259934e-08, + "loss": 0.2686, + "step": 4087 + }, + { + "epoch": 2.9070222222222224, + "grad_norm": 0.7164979577064514, + "learning_rate": 5.782093106048159e-08, + "loss": 0.2583, + "step": 4088 + }, + { + "epoch": 2.9077333333333333, + "grad_norm": 0.7516294121742249, + "learning_rate": 5.6935641472701807e-08, + "loss": 0.2575, + "step": 4089 + }, + { + "epoch": 2.9084444444444446, + "grad_norm": 0.8133431673049927, + "learning_rate": 5.605716220562385e-08, + "loss": 0.2542, + "step": 4090 + }, + { + "epoch": 2.9091555555555555, + "grad_norm": 0.6361157894134521, + "learning_rate": 5.5185493860946406e-08, + "loss": 0.2489, + "step": 4091 + }, + { + "epoch": 2.909866666666667, + "grad_norm": 0.7280561923980713, + "learning_rate": 5.4320637035704114e-08, + "loss": 0.2388, + "step": 4092 + }, + { + "epoch": 2.9105777777777777, + "grad_norm": 0.6971002817153931, + "learning_rate": 5.346259232226203e-08, + "loss": 0.2307, + "step": 4093 + }, + { + "epoch": 2.911288888888889, + "grad_norm": 0.7002100348472595, + "learning_rate": 5.2611360308323364e-08, + "loss": 0.2074, + "step": 4094 + }, + { + "epoch": 2.912, + "grad_norm": 0.8195298314094543, + "learning_rate": 5.176694157692508e-08, + "loss": 0.2705, + "step": 4095 + }, + { + "epoch": 2.9127111111111113, + "grad_norm": 0.6844266057014465, + "learning_rate": 5.092933670643452e-08, + "loss": 0.2456, + "step": 4096 + }, + { + "epoch": 2.913422222222222, + "grad_norm": 0.6407532095909119, + "learning_rate": 5.0098546270556106e-08, + "loss": 0.2355, + "step": 4097 + }, + { + "epoch": 2.9141333333333335, + "grad_norm": 0.7205145359039307, + "learning_rate": 4.9274570838322436e-08, + "loss": 0.2508, + "step": 4098 + }, + { + "epoch": 2.9148444444444443, + "grad_norm": 0.778473436832428, + "learning_rate": 4.845741097410206e-08, + "loss": 0.2831, + "step": 4099 + }, + { + "epoch": 2.9155555555555557, + "grad_norm": 0.8694626092910767, + "learning_rate": 4.764706723759172e-08, + "loss": 0.2574, + "step": 4100 + }, + { + "epoch": 2.9162666666666666, + "grad_norm": 0.709980309009552, + "learning_rate": 4.684354018382409e-08, + "loss": 0.2446, + "step": 4101 + }, + { + "epoch": 2.916977777777778, + "grad_norm": 1.1480015516281128, + "learning_rate": 4.604683036316004e-08, + "loss": 0.222, + "step": 4102 + }, + { + "epoch": 2.9176888888888888, + "grad_norm": 0.796094536781311, + "learning_rate": 4.525693832128975e-08, + "loss": 0.2736, + "step": 4103 + }, + { + "epoch": 2.9184, + "grad_norm": 0.7743335962295532, + "learning_rate": 4.4473864599235975e-08, + "loss": 0.2603, + "step": 4104 + }, + { + "epoch": 2.919111111111111, + "grad_norm": 0.7136473655700684, + "learning_rate": 4.36976097333508e-08, + "loss": 0.211, + "step": 4105 + }, + { + "epoch": 2.9198222222222223, + "grad_norm": 0.696791410446167, + "learning_rate": 4.29281742553167e-08, + "loss": 0.2724, + "step": 4106 + }, + { + "epoch": 2.920533333333333, + "grad_norm": 0.7885302305221558, + "learning_rate": 4.216555869214434e-08, + "loss": 0.2227, + "step": 4107 + }, + { + "epoch": 2.9212444444444445, + "grad_norm": 0.8319805860519409, + "learning_rate": 4.1409763566172544e-08, + "loss": 0.2702, + "step": 4108 + }, + { + "epoch": 2.9219555555555554, + "grad_norm": 0.7918619513511658, + "learning_rate": 4.066078939506945e-08, + "loss": 0.2471, + "step": 4109 + }, + { + "epoch": 2.9226666666666667, + "grad_norm": 0.7412719130516052, + "learning_rate": 3.991863669183138e-08, + "loss": 0.229, + "step": 4110 + }, + { + "epoch": 2.9233777777777776, + "grad_norm": 0.9271965622901917, + "learning_rate": 3.918330596478392e-08, + "loss": 0.2421, + "step": 4111 + }, + { + "epoch": 2.924088888888889, + "grad_norm": 0.822387158870697, + "learning_rate": 3.845479771757532e-08, + "loss": 0.2437, + "step": 4112 + }, + { + "epoch": 2.9248, + "grad_norm": 0.7074776291847229, + "learning_rate": 3.773311244918643e-08, + "loss": 0.2132, + "step": 4113 + }, + { + "epoch": 2.925511111111111, + "grad_norm": 0.7303098440170288, + "learning_rate": 3.701825065392184e-08, + "loss": 0.2249, + "step": 4114 + }, + { + "epoch": 2.926222222222222, + "grad_norm": 0.7142307758331299, + "learning_rate": 3.6310212821413225e-08, + "loss": 0.2487, + "step": 4115 + }, + { + "epoch": 2.9269333333333334, + "grad_norm": 0.8108505010604858, + "learning_rate": 3.560899943661822e-08, + "loss": 0.2706, + "step": 4116 + }, + { + "epoch": 2.9276444444444447, + "grad_norm": 0.6779433488845825, + "learning_rate": 3.49146109798193e-08, + "loss": 0.2426, + "step": 4117 + }, + { + "epoch": 2.9283555555555556, + "grad_norm": 0.8308231830596924, + "learning_rate": 3.422704792662601e-08, + "loss": 0.2496, + "step": 4118 + }, + { + "epoch": 2.9290666666666665, + "grad_norm": 0.7029505968093872, + "learning_rate": 3.354631074797166e-08, + "loss": 0.2718, + "step": 4119 + }, + { + "epoch": 2.929777777777778, + "grad_norm": 0.9990431666374207, + "learning_rate": 3.2872399910115484e-08, + "loss": 0.2767, + "step": 4120 + }, + { + "epoch": 2.930488888888889, + "grad_norm": 0.7050060033798218, + "learning_rate": 3.220531587463937e-08, + "loss": 0.2549, + "step": 4121 + }, + { + "epoch": 2.9312, + "grad_norm": 0.6807776689529419, + "learning_rate": 3.154505909845229e-08, + "loss": 0.263, + "step": 4122 + }, + { + "epoch": 2.931911111111111, + "grad_norm": 0.6746832132339478, + "learning_rate": 3.0891630033782484e-08, + "loss": 0.2637, + "step": 4123 + }, + { + "epoch": 2.9326222222222222, + "grad_norm": 0.8378908634185791, + "learning_rate": 3.024502912818528e-08, + "loss": 0.2486, + "step": 4124 + }, + { + "epoch": 2.9333333333333336, + "grad_norm": 0.8298405408859253, + "learning_rate": 2.960525682453974e-08, + "loss": 0.2141, + "step": 4125 + }, + { + "epoch": 2.9340444444444445, + "grad_norm": 0.8034265637397766, + "learning_rate": 2.897231356104424e-08, + "loss": 0.2952, + "step": 4126 + }, + { + "epoch": 2.9347555555555553, + "grad_norm": 0.81182861328125, + "learning_rate": 2.8346199771221995e-08, + "loss": 0.2331, + "step": 4127 + }, + { + "epoch": 2.9354666666666667, + "grad_norm": 0.7426519393920898, + "learning_rate": 2.7726915883919958e-08, + "loss": 0.2295, + "step": 4128 + }, + { + "epoch": 2.936177777777778, + "grad_norm": 0.8620747327804565, + "learning_rate": 2.711446232330328e-08, + "loss": 0.2484, + "step": 4129 + }, + { + "epoch": 2.936888888888889, + "grad_norm": 0.8129112124443054, + "learning_rate": 2.6508839508861963e-08, + "loss": 0.241, + "step": 4130 + }, + { + "epoch": 2.9375999999999998, + "grad_norm": 0.7819024920463562, + "learning_rate": 2.5910047855405296e-08, + "loss": 0.2475, + "step": 4131 + }, + { + "epoch": 2.938311111111111, + "grad_norm": 0.7196345925331116, + "learning_rate": 2.5318087773066325e-08, + "loss": 0.2554, + "step": 4132 + }, + { + "epoch": 2.9390222222222224, + "grad_norm": 0.8363116979598999, + "learning_rate": 2.4732959667296276e-08, + "loss": 0.2637, + "step": 4133 + }, + { + "epoch": 2.9397333333333333, + "grad_norm": 0.6578731536865234, + "learning_rate": 2.4154663938867894e-08, + "loss": 0.1833, + "step": 4134 + }, + { + "epoch": 2.940444444444444, + "grad_norm": 0.8045686483383179, + "learning_rate": 2.3583200983874344e-08, + "loss": 0.2338, + "step": 4135 + }, + { + "epoch": 2.9411555555555555, + "grad_norm": 0.7801622748374939, + "learning_rate": 2.3018571193729188e-08, + "loss": 0.234, + "step": 4136 + }, + { + "epoch": 2.941866666666667, + "grad_norm": 0.8286011815071106, + "learning_rate": 2.2460774955165298e-08, + "loss": 0.2207, + "step": 4137 + }, + { + "epoch": 2.9425777777777777, + "grad_norm": 1.0101238489151, + "learning_rate": 2.190981265023373e-08, + "loss": 0.3053, + "step": 4138 + }, + { + "epoch": 2.9432888888888886, + "grad_norm": 0.8327244520187378, + "learning_rate": 2.1365684656308172e-08, + "loss": 0.2494, + "step": 4139 + }, + { + "epoch": 2.944, + "grad_norm": 0.7851852178573608, + "learning_rate": 2.082839134607828e-08, + "loss": 0.2426, + "step": 4140 + }, + { + "epoch": 2.9447111111111113, + "grad_norm": 0.6689561605453491, + "learning_rate": 2.029793308755301e-08, + "loss": 0.2159, + "step": 4141 + }, + { + "epoch": 2.945422222222222, + "grad_norm": 0.7238081097602844, + "learning_rate": 1.9774310244059512e-08, + "loss": 0.2191, + "step": 4142 + }, + { + "epoch": 2.9461333333333335, + "grad_norm": 0.750270664691925, + "learning_rate": 1.925752317424534e-08, + "loss": 0.2527, + "step": 4143 + }, + { + "epoch": 2.9468444444444444, + "grad_norm": 0.6648085117340088, + "learning_rate": 1.874757223207291e-08, + "loss": 0.2142, + "step": 4144 + }, + { + "epoch": 2.9475555555555557, + "grad_norm": 0.7633259892463684, + "learning_rate": 1.824445776682504e-08, + "loss": 0.2201, + "step": 4145 + }, + { + "epoch": 2.9482666666666666, + "grad_norm": 0.7433714270591736, + "learning_rate": 1.7748180123100535e-08, + "loss": 0.2235, + "step": 4146 + }, + { + "epoch": 2.948977777777778, + "grad_norm": 0.8591189384460449, + "learning_rate": 1.725873964081415e-08, + "loss": 0.2327, + "step": 4147 + }, + { + "epoch": 2.949688888888889, + "grad_norm": 0.6950580477714539, + "learning_rate": 1.677613665520106e-08, + "loss": 0.1922, + "step": 4148 + }, + { + "epoch": 2.9504, + "grad_norm": 0.752804696559906, + "learning_rate": 1.6300371496810184e-08, + "loss": 0.281, + "step": 4149 + }, + { + "epoch": 2.951111111111111, + "grad_norm": 0.7234923243522644, + "learning_rate": 1.583144449150975e-08, + "loss": 0.2358, + "step": 4150 + }, + { + "epoch": 2.9518222222222223, + "grad_norm": 0.7438666224479675, + "learning_rate": 1.5369355960481725e-08, + "loss": 0.2665, + "step": 4151 + }, + { + "epoch": 2.9525333333333332, + "grad_norm": 0.9132182598114014, + "learning_rate": 1.4914106220225156e-08, + "loss": 0.2446, + "step": 4152 + }, + { + "epoch": 2.9532444444444446, + "grad_norm": 0.7982379198074341, + "learning_rate": 1.4465695582553951e-08, + "loss": 0.2342, + "step": 4153 + }, + { + "epoch": 2.9539555555555554, + "grad_norm": 0.6945755481719971, + "learning_rate": 1.402412435460132e-08, + "loss": 0.2385, + "step": 4154 + }, + { + "epoch": 2.9546666666666668, + "grad_norm": 0.7345110774040222, + "learning_rate": 1.3589392838811999e-08, + "loss": 0.2424, + "step": 4155 + }, + { + "epoch": 2.9553777777777777, + "grad_norm": 0.7568763494491577, + "learning_rate": 1.3161501332947802e-08, + "loss": 0.2377, + "step": 4156 + }, + { + "epoch": 2.956088888888889, + "grad_norm": 0.8786019086837769, + "learning_rate": 1.2740450130084292e-08, + "loss": 0.2737, + "step": 4157 + }, + { + "epoch": 2.9568, + "grad_norm": 0.786629855632782, + "learning_rate": 1.2326239518614114e-08, + "loss": 0.2372, + "step": 4158 + }, + { + "epoch": 2.957511111111111, + "grad_norm": 0.7766958475112915, + "learning_rate": 1.1918869782242549e-08, + "loss": 0.2267, + "step": 4159 + }, + { + "epoch": 2.958222222222222, + "grad_norm": 0.7313870191574097, + "learning_rate": 1.1518341199989735e-08, + "loss": 0.2236, + "step": 4160 + }, + { + "epoch": 2.9589333333333334, + "grad_norm": 0.8405714631080627, + "learning_rate": 1.1124654046191785e-08, + "loss": 0.2701, + "step": 4161 + }, + { + "epoch": 2.9596444444444443, + "grad_norm": 0.6290045976638794, + "learning_rate": 1.0737808590495225e-08, + "loss": 0.2362, + "step": 4162 + }, + { + "epoch": 2.9603555555555556, + "grad_norm": 0.8103829622268677, + "learning_rate": 1.035780509786588e-08, + "loss": 0.2315, + "step": 4163 + }, + { + "epoch": 2.9610666666666665, + "grad_norm": 0.7599387764930725, + "learning_rate": 9.984643828576669e-09, + "loss": 0.2448, + "step": 4164 + }, + { + "epoch": 2.961777777777778, + "grad_norm": 0.7237648367881775, + "learning_rate": 9.61832503822091e-09, + "loss": 0.2646, + "step": 4165 + }, + { + "epoch": 2.9624888888888887, + "grad_norm": 0.7256041169166565, + "learning_rate": 9.25884897770013e-09, + "loss": 0.2501, + "step": 4166 + }, + { + "epoch": 2.9632, + "grad_norm": 0.7248656749725342, + "learning_rate": 8.906215893231818e-09, + "loss": 0.2339, + "step": 4167 + }, + { + "epoch": 2.963911111111111, + "grad_norm": 0.9443759322166443, + "learning_rate": 8.560426026343881e-09, + "loss": 0.2342, + "step": 4168 + }, + { + "epoch": 2.9646222222222223, + "grad_norm": 0.8492706418037415, + "learning_rate": 8.221479613881312e-09, + "loss": 0.2447, + "step": 4169 + }, + { + "epoch": 2.9653333333333336, + "grad_norm": 0.6213926076889038, + "learning_rate": 7.889376887997291e-09, + "loss": 0.2182, + "step": 4170 + }, + { + "epoch": 2.9660444444444445, + "grad_norm": 0.695268988609314, + "learning_rate": 7.564118076159866e-09, + "loss": 0.2217, + "step": 4171 + }, + { + "epoch": 2.9667555555555554, + "grad_norm": 0.7146511077880859, + "learning_rate": 7.245703401149717e-09, + "loss": 0.2564, + "step": 4172 + }, + { + "epoch": 2.9674666666666667, + "grad_norm": 0.9007747769355774, + "learning_rate": 6.9341330810590575e-09, + "loss": 0.2591, + "step": 4173 + }, + { + "epoch": 2.968177777777778, + "grad_norm": 0.7567370533943176, + "learning_rate": 6.629407329292736e-09, + "loss": 0.287, + "step": 4174 + }, + { + "epoch": 2.968888888888889, + "grad_norm": 0.6872206926345825, + "learning_rate": 6.33152635456602e-09, + "loss": 0.2399, + "step": 4175 + }, + { + "epoch": 2.9696, + "grad_norm": 0.6804379224777222, + "learning_rate": 6.0404903609068146e-09, + "loss": 0.2449, + "step": 4176 + }, + { + "epoch": 2.970311111111111, + "grad_norm": 0.7002463340759277, + "learning_rate": 5.756299547656774e-09, + "loss": 0.2287, + "step": 4177 + }, + { + "epoch": 2.9710222222222225, + "grad_norm": 0.7901144027709961, + "learning_rate": 5.47895410946575e-09, + "loss": 0.24, + "step": 4178 + }, + { + "epoch": 2.9717333333333333, + "grad_norm": 0.6939697265625, + "learning_rate": 5.208454236296234e-09, + "loss": 0.2385, + "step": 4179 + }, + { + "epoch": 2.9724444444444442, + "grad_norm": 0.8262694478034973, + "learning_rate": 4.9448001134233536e-09, + "loss": 0.2611, + "step": 4180 + }, + { + "epoch": 2.9731555555555556, + "grad_norm": 0.7699682116508484, + "learning_rate": 4.687991921431545e-09, + "loss": 0.2436, + "step": 4181 + }, + { + "epoch": 2.973866666666667, + "grad_norm": 0.6815473437309265, + "learning_rate": 4.438029836216773e-09, + "loss": 0.2479, + "step": 4182 + }, + { + "epoch": 2.9745777777777778, + "grad_norm": 0.7455981969833374, + "learning_rate": 4.19491402898542e-09, + "loss": 0.2343, + "step": 4183 + }, + { + "epoch": 2.9752888888888886, + "grad_norm": 0.7788275480270386, + "learning_rate": 3.958644666257616e-09, + "loss": 0.2121, + "step": 4184 + }, + { + "epoch": 2.976, + "grad_norm": 0.773189902305603, + "learning_rate": 3.72922190985836e-09, + "loss": 0.2593, + "step": 4185 + }, + { + "epoch": 2.9767111111111113, + "grad_norm": 0.9033201932907104, + "learning_rate": 3.5066459169297294e-09, + "loss": 0.2176, + "step": 4186 + }, + { + "epoch": 2.977422222222222, + "grad_norm": 0.6463629603385925, + "learning_rate": 3.2909168399197776e-09, + "loss": 0.2074, + "step": 4187 + }, + { + "epoch": 2.978133333333333, + "grad_norm": 0.8064982295036316, + "learning_rate": 3.082034826586977e-09, + "loss": 0.2517, + "step": 4188 + }, + { + "epoch": 2.9788444444444444, + "grad_norm": 0.761600911617279, + "learning_rate": 2.8800000200035483e-09, + "loss": 0.2655, + "step": 4189 + }, + { + "epoch": 2.9795555555555557, + "grad_norm": 0.652769148349762, + "learning_rate": 2.684812558547689e-09, + "loss": 0.2271, + "step": 4190 + }, + { + "epoch": 2.9802666666666666, + "grad_norm": 0.7855852246284485, + "learning_rate": 2.496472575910236e-09, + "loss": 0.2743, + "step": 4191 + }, + { + "epoch": 2.9809777777777775, + "grad_norm": 0.7462372183799744, + "learning_rate": 2.3149802010913323e-09, + "loss": 0.2587, + "step": 4192 + }, + { + "epoch": 2.981688888888889, + "grad_norm": 0.7185855507850647, + "learning_rate": 2.1403355584015406e-09, + "loss": 0.2036, + "step": 4193 + }, + { + "epoch": 2.9824, + "grad_norm": 0.7441174387931824, + "learning_rate": 1.9725387674585095e-09, + "loss": 0.2348, + "step": 4194 + }, + { + "epoch": 2.983111111111111, + "grad_norm": 0.8350330591201782, + "learning_rate": 1.8115899431936367e-09, + "loss": 0.2628, + "step": 4195 + }, + { + "epoch": 2.9838222222222224, + "grad_norm": 0.8122992515563965, + "learning_rate": 1.6574891958442973e-09, + "loss": 0.2564, + "step": 4196 + }, + { + "epoch": 2.9845333333333333, + "grad_norm": 0.7405588030815125, + "learning_rate": 1.5102366309605044e-09, + "loss": 0.2623, + "step": 4197 + }, + { + "epoch": 2.9852444444444446, + "grad_norm": 0.6682016849517822, + "learning_rate": 1.3698323493993582e-09, + "loss": 0.2276, + "step": 4198 + }, + { + "epoch": 2.9859555555555555, + "grad_norm": 0.8251067996025085, + "learning_rate": 1.2362764473283773e-09, + "loss": 0.2603, + "step": 4199 + }, + { + "epoch": 2.986666666666667, + "grad_norm": 0.7611521482467651, + "learning_rate": 1.1095690162243878e-09, + "loss": 0.2477, + "step": 4200 + }, + { + "epoch": 2.9873777777777777, + "grad_norm": 0.7084150910377502, + "learning_rate": 9.89710142872413e-10, + "loss": 0.2192, + "step": 4201 + }, + { + "epoch": 2.988088888888889, + "grad_norm": 0.7694860696792603, + "learning_rate": 8.766999093690054e-10, + "loss": 0.2387, + "step": 4202 + }, + { + "epoch": 2.9888, + "grad_norm": 0.8647977709770203, + "learning_rate": 7.705383931178034e-10, + "loss": 0.2359, + "step": 4203 + }, + { + "epoch": 2.9895111111111112, + "grad_norm": 0.853565514087677, + "learning_rate": 6.71225666831754e-10, + "loss": 0.2815, + "step": 4204 + }, + { + "epoch": 2.990222222222222, + "grad_norm": 0.7230651378631592, + "learning_rate": 5.78761798534222e-10, + "loss": 0.2209, + "step": 4205 + }, + { + "epoch": 2.9909333333333334, + "grad_norm": 0.7751672267913818, + "learning_rate": 4.931468515556593e-10, + "loss": 0.2399, + "step": 4206 + }, + { + "epoch": 2.9916444444444443, + "grad_norm": 0.9631101489067078, + "learning_rate": 4.143808845369357e-10, + "loss": 0.2928, + "step": 4207 + }, + { + "epoch": 2.9923555555555557, + "grad_norm": 0.7501070499420166, + "learning_rate": 3.4246395142822906e-10, + "loss": 0.2654, + "step": 4208 + }, + { + "epoch": 2.9930666666666665, + "grad_norm": 0.8723669052124023, + "learning_rate": 2.773961014856941e-10, + "loss": 0.2671, + "step": 4209 + }, + { + "epoch": 2.993777777777778, + "grad_norm": 0.7349015474319458, + "learning_rate": 2.1917737927812377e-10, + "loss": 0.21, + "step": 4210 + }, + { + "epoch": 2.9944888888888888, + "grad_norm": 0.7025666832923889, + "learning_rate": 1.6780782468139856e-10, + "loss": 0.2349, + "step": 4211 + }, + { + "epoch": 2.9952, + "grad_norm": 0.8385546803474426, + "learning_rate": 1.2328747287848609e-10, + "loss": 0.2952, + "step": 4212 + }, + { + "epoch": 2.995911111111111, + "grad_norm": 0.7203818559646606, + "learning_rate": 8.561635436499238e-11, + "loss": 0.222, + "step": 4213 + }, + { + "epoch": 2.9966222222222223, + "grad_norm": 0.7009567022323608, + "learning_rate": 5.4794494941390333e-11, + "loss": 0.2357, + "step": 4214 + }, + { + "epoch": 2.997333333333333, + "grad_norm": 0.7408928275108337, + "learning_rate": 3.082191571968096e-11, + "loss": 0.2421, + "step": 4215 + }, + { + "epoch": 2.9980444444444445, + "grad_norm": 0.7557035088539124, + "learning_rate": 1.3698633117842365e-11, + "loss": 0.2457, + "step": 4216 + }, + { + "epoch": 2.9987555555555554, + "grad_norm": 0.7734407186508179, + "learning_rate": 3.424658865380792e-12, + "loss": 0.2347, + "step": 4217 + }, + { + "epoch": 2.9994666666666667, + "grad_norm": 0.6536984443664551, + "learning_rate": 0.0, + "loss": 0.261, + "step": 4218 + } + ], + "logging_steps": 1, + "max_steps": 4218, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.3431341283834593e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}