Training in progress, epoch 14, checkpoint

checkpoint-2625/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.5.0

checkpoint-2625/adapter_config.json

@@ -0,0 +1,25 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen1.5-4B",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "k_proj",
+    "v_proj",
+    "down_proj",
+    "gate_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-2625/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9f7033efbff9e8298513508044e2a0dcb7df7a349cee80aa52d988137f50103
+size 112130416

checkpoint-2625/added_tokens.json

@@ -0,0 +1,5 @@
+{
+  "<|endoftext|>": 151643,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644
+}

checkpoint-2625/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff189f6a59aa751eaa0aa5b416300a430e70d0f669737e510ef4e4981301dd07
+size 224537202

checkpoint-2625/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0fd461d3cd2e6ddfa03082a4350a020e3286e945c970edf736743d084697cbd0
+size 15024

checkpoint-2625/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c54d148f3b9e98c35720356ed60d0fde44fc53e5bd842580ceb77c2e01f0ea2b
+size 15024

checkpoint-2625/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1de25f0ce9e4b1e80d639f4ef532e318e8c7401602ef4f15abe15fcd1553648
+size 15024

checkpoint-2625/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95b11af5c79a110afb1f74285c6bbeb7c41275f6af122b1f2f913fa156e47b2c
+size 15024

checkpoint-2625/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9257ce91eefc916b33cd07133b5744709c58ad88e3240c4d2ccb084cb5da3a34
+size 1064

checkpoint-2625/special_tokens_map.json

@@ -0,0 +1,14 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|endoftext|>"
+}

checkpoint-2625/tokenizer_config.json

@@ -0,0 +1,43 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-2625/trainer_state.json

@@ -0,0 +1,425 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 14.0,
+  "eval_steps": 500,
+  "global_step": 2625,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.9283596873283386,
+      "learning_rate": 3e-05,
+      "loss": 1.963,
+      "step": 100
+    },
+    {
+      "epoch": 0.9973333333333333,
+      "eval_accuracy": 0.5695384615384615,
+      "eval_loss": 1.6438876390457153,
+      "eval_runtime": 5.0534,
+      "eval_samples_per_second": 98.943,
+      "eval_steps_per_second": 12.467,
+      "step": 187
+    },
+    {
+      "epoch": 0.9973333333333333,
+      "eval_exact_match": 16.2,
+      "eval_f1": 25.888571428571442,
+      "step": 187
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 1.0407154560089111,
+      "learning_rate": 3e-05,
+      "loss": 1.6668,
+      "step": 200
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1281720399856567,
+      "learning_rate": 3e-05,
+      "loss": 1.6099,
+      "step": 300
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5733333333333334,
+      "eval_loss": 1.618270754814148,
+      "eval_runtime": 5.1983,
+      "eval_samples_per_second": 96.186,
+      "eval_steps_per_second": 12.119,
+      "step": 375
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 16.6,
+      "eval_f1": 26.287619047619057,
+      "step": 375
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 1.2508225440979004,
+      "learning_rate": 3e-05,
+      "loss": 1.5866,
+      "step": 400
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 1.3197834491729736,
+      "learning_rate": 3e-05,
+      "loss": 1.524,
+      "step": 500
+    },
+    {
+      "epoch": 2.997333333333333,
+      "eval_accuracy": 0.5743589743589743,
+      "eval_loss": 1.6164497137069702,
+      "eval_runtime": 5.0495,
+      "eval_samples_per_second": 99.019,
+      "eval_steps_per_second": 12.476,
+      "step": 562
+    },
+    {
+      "epoch": 2.997333333333333,
+      "eval_exact_match": 17.8,
+      "eval_f1": 27.35428571428571,
+      "step": 562
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.6320394277572632,
+      "learning_rate": 3e-05,
+      "loss": 1.4752,
+      "step": 600
+    },
+    {
+      "epoch": 3.7333333333333334,
+      "grad_norm": 1.7101725339889526,
+      "learning_rate": 3e-05,
+      "loss": 1.3938,
+      "step": 700
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5728717948717948,
+      "eval_loss": 1.6376241445541382,
+      "eval_runtime": 4.9662,
+      "eval_samples_per_second": 100.68,
+      "eval_steps_per_second": 12.686,
+      "step": 750
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 18.0,
+      "eval_f1": 27.498571428571427,
+      "step": 750
+    },
+    {
+      "epoch": 4.266666666666667,
+      "grad_norm": 2.1727664470672607,
+      "learning_rate": 3e-05,
+      "loss": 1.342,
+      "step": 800
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 2.341384172439575,
+      "learning_rate": 3e-05,
+      "loss": 1.2685,
+      "step": 900
+    },
+    {
+      "epoch": 4.997333333333334,
+      "eval_accuracy": 0.5699487179487179,
+      "eval_loss": 1.684618353843689,
+      "eval_runtime": 5.5457,
+      "eval_samples_per_second": 90.16,
+      "eval_steps_per_second": 11.36,
+      "step": 937
+    },
+    {
+      "epoch": 4.997333333333334,
+      "eval_exact_match": 19.2,
+      "eval_f1": 28.190952380952375,
+      "step": 937
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 2.5624172687530518,
+      "learning_rate": 3e-05,
+      "loss": 1.1847,
+      "step": 1000
+    },
+    {
+      "epoch": 5.866666666666667,
+      "grad_norm": 2.8159019947052,
+      "learning_rate": 3e-05,
+      "loss": 1.1591,
+      "step": 1100
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5672820512820513,
+      "eval_loss": 1.7546838521957397,
+      "eval_runtime": 5.0186,
+      "eval_samples_per_second": 99.629,
+      "eval_steps_per_second": 12.553,
+      "step": 1125
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 16.6,
+      "eval_f1": 25.054761904761914,
+      "step": 1125
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 3.056166410446167,
+      "learning_rate": 3e-05,
+      "loss": 1.0486,
+      "step": 1200
+    },
+    {
+      "epoch": 6.933333333333334,
+      "grad_norm": 2.873281240463257,
+      "learning_rate": 3e-05,
+      "loss": 1.0444,
+      "step": 1300
+    },
+    {
+      "epoch": 6.997333333333334,
+      "eval_accuracy": 0.5643076923076923,
+      "eval_loss": 1.8394943475723267,
+      "eval_runtime": 5.5931,
+      "eval_samples_per_second": 89.396,
+      "eval_steps_per_second": 11.264,
+      "step": 1312
+    },
+    {
+      "epoch": 6.997333333333334,
+      "eval_exact_match": 15.8,
+      "eval_f1": 23.925714285714296,
+      "step": 1312
+    },
+    {
+      "epoch": 7.466666666666667,
+      "grad_norm": 3.069322109222412,
+      "learning_rate": 3e-05,
+      "loss": 0.9374,
+      "step": 1400
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.2732129096984863,
+      "learning_rate": 3e-05,
+      "loss": 0.9535,
+      "step": 1500
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5612820512820513,
+      "eval_loss": 1.9008216857910156,
+      "eval_runtime": 5.1357,
+      "eval_samples_per_second": 97.358,
+      "eval_steps_per_second": 12.267,
+      "step": 1500
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 15.2,
+      "eval_f1": 23.12063492063493,
+      "step": 1500
+    },
+    {
+      "epoch": 8.533333333333333,
+      "grad_norm": 3.382610321044922,
+      "learning_rate": 3e-05,
+      "loss": 0.8235,
+      "step": 1600
+    },
+    {
+      "epoch": 8.997333333333334,
+      "eval_accuracy": 0.5592307692307692,
+      "eval_loss": 2.026776075363159,
+      "eval_runtime": 5.5056,
+      "eval_samples_per_second": 90.817,
+      "eval_steps_per_second": 11.443,
+      "step": 1687
+    },
+    {
+      "epoch": 8.997333333333334,
+      "eval_exact_match": 13.8,
+      "eval_f1": 21.42285714285715,
+      "step": 1687
+    },
+    {
+      "epoch": 9.066666666666666,
+      "grad_norm": 4.363819122314453,
+      "learning_rate": 3e-05,
+      "loss": 0.8547,
+      "step": 1700
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 3.685805320739746,
+      "learning_rate": 3e-05,
+      "loss": 0.7635,
+      "step": 1800
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.5567692307692308,
+      "eval_loss": 2.0937435626983643,
+      "eval_runtime": 5.235,
+      "eval_samples_per_second": 95.512,
+      "eval_steps_per_second": 12.034,
+      "step": 1875
+    },
+    {
+      "epoch": 10.0,
+      "eval_exact_match": 15.2,
+      "eval_f1": 22.031904761904762,
+      "step": 1875
+    },
+    {
+      "epoch": 10.133333333333333,
+      "grad_norm": 3.5318357944488525,
+      "learning_rate": 3e-05,
+      "loss": 0.7554,
+      "step": 1900
+    },
+    {
+      "epoch": 10.666666666666666,
+      "grad_norm": 3.6454901695251465,
+      "learning_rate": 3e-05,
+      "loss": 0.6978,
+      "step": 2000
+    },
+    {
+      "epoch": 10.997333333333334,
+      "eval_accuracy": 0.556974358974359,
+      "eval_loss": 2.1557741165161133,
+      "eval_runtime": 5.1556,
+      "eval_samples_per_second": 96.981,
+      "eval_steps_per_second": 12.22,
+      "step": 2062
+    },
+    {
+      "epoch": 10.997333333333334,
+      "eval_exact_match": 15.0,
+      "eval_f1": 23.010000000000012,
+      "step": 2062
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 3.658707618713379,
+      "learning_rate": 3e-05,
+      "loss": 0.6969,
+      "step": 2100
+    },
+    {
+      "epoch": 11.733333333333333,
+      "grad_norm": 3.7350716590881348,
+      "learning_rate": 3e-05,
+      "loss": 0.6615,
+      "step": 2200
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.5551794871794872,
+      "eval_loss": 2.239973545074463,
+      "eval_runtime": 4.9026,
+      "eval_samples_per_second": 101.986,
+      "eval_steps_per_second": 12.85,
+      "step": 2250
+    },
+    {
+      "epoch": 12.0,
+      "eval_exact_match": 15.6,
+      "eval_f1": 23.333492063492066,
+      "step": 2250
+    },
+    {
+      "epoch": 12.266666666666667,
+      "grad_norm": 4.4434709548950195,
+      "learning_rate": 3e-05,
+      "loss": 0.632,
+      "step": 2300
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 3.4198718070983887,
+      "learning_rate": 3e-05,
+      "loss": 0.6262,
+      "step": 2400
+    },
+    {
+      "epoch": 12.997333333333334,
+      "eval_accuracy": 0.5556410256410257,
+      "eval_loss": 2.268662452697754,
+      "eval_runtime": 5.327,
+      "eval_samples_per_second": 93.862,
+      "eval_steps_per_second": 11.827,
+      "step": 2437
+    },
+    {
+      "epoch": 12.997333333333334,
+      "eval_exact_match": 14.4,
+      "eval_f1": 22.583492063492077,
+      "step": 2437
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 3.158496618270874,
+      "learning_rate": 3e-05,
+      "loss": 0.5973,
+      "step": 2500
+    },
+    {
+      "epoch": 13.866666666666667,
+      "grad_norm": 3.5720713138580322,
+      "learning_rate": 3e-05,
+      "loss": 0.5958,
+      "step": 2600
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.5536923076923077,
+      "eval_loss": 2.358220100402832,
+      "eval_runtime": 5.236,
+      "eval_samples_per_second": 95.493,
+      "eval_steps_per_second": 12.032,
+      "step": 2625
+    },
+    {
+      "epoch": 14.0,
+      "eval_exact_match": 14.6,
+      "eval_f1": 21.79904761904763,
+      "step": 2625
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 9350,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.807998138063258e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-2625/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a175de730fa6917b3a611bb9e53766640b7a5e4a426f825f2a602cf98661e47
+size 5304