Hugging Face's logo

tyzhu
/
lmind_nq_train6000_eval6489_v1_doc_qa_v3_5e-5_lora2

Safetensors
Generated from Trainer
Eval Results
Model card Files Community
tyzhu commited on
Commit
15c5f6a
•
1 Parent(s): 8a55c4c

Training in progress, epoch 12, checkpoint

Browse files
Files changed (14) hide show
  1. checkpoint-6348/adapter_config.json +1 -1
  2. checkpoint-6348/adapter_model.bin +3 -0
  3. checkpoint-6348/added_tokens.json +3 -3
  4. checkpoint-6348/optimizer.pt +2 -2
  5. checkpoint-6348/rng_state_0.pth +1 -1
  6. checkpoint-6348/rng_state_1.pth +1 -1
  7. checkpoint-6348/rng_state_2.pth +1 -1
  8. checkpoint-6348/rng_state_3.pth +1 -1
  9. checkpoint-6348/special_tokens_map.json +4 -12
  10. checkpoint-6348/tokenizer.json +0 -0
  11. checkpoint-6348/tokenizer.model +3 -0
  12. checkpoint-6348/tokenizer_config.json +17 -20
  13. checkpoint-6348/trainer_state.json +211 -288
  14. checkpoint-6348/training_args.bin +2 -2
checkpoint-6348/adapter_config.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "auto_mapping": null,
3
- "base_model_name_or_path": "Qwen/Qwen1.5-4B",
4
  "bias": "none",
5
  "fan_in_fan_out": false,
6
  "inference_mode": true,
 
1
  {
2
  "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
4
  "bias": "none",
5
  "fan_in_fan_out": false,
6
  "inference_mode": true,
checkpoint-6348/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4430ffe72d173016b6ef8ca53b98ea59073a3e93deff4c869925458199630c44
3
+ size 143269386
checkpoint-6348/added_tokens.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "<|endoftext|>": 151643,
3
- "<|im_end|>": 151645,
4
- "<|im_start|>": 151644
5
  }
 
1
  {
2
+ "</s>": 2,
3
+ "<s>": 1,
4
+ "<unk>": 0
5
  }
checkpoint-6348/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c87ab103a50a2e0ef97803097ce1cc32667d29412adb586d3c75832e577eb859
3
- size 224537202
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:114937d6f7b3124e517742bd6f593c554d33acf995128f3ef7fd545036172fb5
3
+ size 286585234
checkpoint-6348/rng_state_0.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1ac9ae97a8b610f64cb11979a53a8c4a1a13184f86de07268dd95f936b60df8e
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7907b4bd7b504596c0aa4c3658f077704867c9e574c435794ef2b92f8ca9d4e
3
  size 15024
checkpoint-6348/rng_state_1.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4e355bffa6a6c4d2ec34decfacde624c20de8d027450a2d3c865521389f4220e
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6b56fb5a73e460e89b0e5fcdc45cae7083a031e7d28e0655447fcfef491b4fd
3
  size 15024
checkpoint-6348/rng_state_2.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:70814c35f58f2a0e4cdd1ccdeed60559d6732daa4c57f824c4eac1e7ae81f160
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ec04a7f17f410ccb60e0811a647a07d62fd164992672caf9778fca854c6d23b
3
  size 15024
checkpoint-6348/rng_state_3.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7e4c18762b24ec12a5e3dd94be754332039c488be34bca43c99f7477a4f98ff0
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01de704b5ff9b33951aa27fa07ea45dd0d04b69e2cc056d2c18ebbd2403bc3ea
3
  size 15024
checkpoint-6348/special_tokens_map.json CHANGED
@@ -1,14 +1,6 @@
1
  {
2
- "additional_special_tokens": [
3
- "<|im_start|>",
4
- "<|im_end|>"
5
- ],
6
- "eos_token": {
7
- "content": "<|endoftext|>",
8
- "lstrip": false,
9
- "normalized": false,
10
- "rstrip": false,
11
- "single_word": false
12
- },
13
- "pad_token": "<|endoftext|>"
14
  }
 
1
  {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "</s>",
5
+ "unk_token": "<unk>"
 
 
 
 
 
 
 
 
6
  }
checkpoint-6348/tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
checkpoint-6348/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
checkpoint-6348/tokenizer_config.json CHANGED
@@ -1,24 +1,23 @@
1
  {
2
- "add_prefix_space": false,
3
  "added_tokens_decoder": {
4
- "151643": {
5
- "content": "<|endoftext|>",
6
  "lstrip": false,
7
  "normalized": false,
8
  "rstrip": false,
9
  "single_word": false,
10
  "special": true
11
  },
12
- "151644": {
13
- "content": "<|im_start|>",
14
  "lstrip": false,
15
  "normalized": false,
16
  "rstrip": false,
17
  "single_word": false,
18
  "special": true
19
  },
20
- "151645": {
21
- "content": "<|im_end|>",
22
  "lstrip": false,
23
  "normalized": false,
24
  "rstrip": false,
@@ -26,18 +25,16 @@
26
  "special": true
27
  }
28
  },
29
- "additional_special_tokens": [
30
- "<|im_start|>",
31
- "<|im_end|>"
32
- ],
33
- "bos_token": null,
34
- "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 %}",
35
  "clean_up_tokenization_spaces": false,
36
- "eos_token": "<|endoftext|>",
37
- "errors": "replace",
38
- "model_max_length": 32768,
39
- "pad_token": "<|endoftext|>",
40
- "split_special_tokens": false,
41
- "tokenizer_class": "Qwen2Tokenizer",
42
- "unk_token": null
 
 
43
  }
 
1
  {
 
2
  "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<unk>",
5
  "lstrip": false,
6
  "normalized": false,
7
  "rstrip": false,
8
  "single_word": false,
9
  "special": true
10
  },
11
+ "1": {
12
+ "content": "<s>",
13
  "lstrip": false,
14
  "normalized": false,
15
  "rstrip": false,
16
  "single_word": false,
17
  "special": true
18
  },
19
+ "2": {
20
+ "content": "</s>",
21
  "lstrip": false,
22
  "normalized": false,
23
  "rstrip": false,
 
25
  "special": true
26
  }
27
  },
28
+ "additional_special_tokens": [],
29
+ "bos_token": "<s>",
 
 
 
 
30
  "clean_up_tokenization_spaces": false,
31
+ "eos_token": "</s>",
32
+ "legacy": false,
33
+ "model_max_length": 1000000000000000019884624838656,
34
+ "pad_token": "</s>",
35
+ "padding_side": "left",
36
+ "sp_model_kwargs": {},
37
+ "tokenizer_class": "LlamaTokenizer",
38
+ "unk_token": "<unk>",
39
+ "use_default_system_prompt": true
40
  }
checkpoint-6348/trainer_state.json CHANGED
@@ -9,646 +9,569 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 0.1890359168241966,
13
- "grad_norm": 0.5107442140579224,
14
  "learning_rate": 5e-05,
15
- "loss": 2.0482,
16
  "step": 100
17
  },
18
  {
19
- "epoch": 0.3780718336483932,
20
- "grad_norm": 0.5563424229621887,
21
  "learning_rate": 5e-05,
22
- "loss": 1.8835,
23
  "step": 200
24
  },
25
  {
26
- "epoch": 0.5671077504725898,
27
- "grad_norm": 0.4735371172428131,
28
  "learning_rate": 5e-05,
29
- "loss": 1.8512,
30
  "step": 300
31
  },
32
  {
33
- "epoch": 0.7561436672967864,
34
- "grad_norm": 0.4543495178222656,
35
  "learning_rate": 5e-05,
36
- "loss": 1.8489,
37
  "step": 400
38
  },
39
  {
40
- "epoch": 0.945179584120983,
41
- "grad_norm": 0.5007112622261047,
42
  "learning_rate": 5e-05,
43
- "loss": 1.8439,
44
  "step": 500
45
  },
46
  {
47
  "epoch": 1.0,
48
- "eval_accuracy": 0.5733333333333334,
49
- "eval_loss": 1.6194567680358887,
50
- "eval_runtime": 5.4154,
51
- "eval_samples_per_second": 92.33,
52
- "eval_steps_per_second": 11.634,
53
  "step": 529
54
  },
55
  {
56
  "epoch": 1.0,
57
- "eval_exact_match": 17.0,
58
- "eval_f1": 26.880952380952387,
59
  "step": 529
60
  },
61
  {
62
- "epoch": 1.1342155009451795,
63
- "grad_norm": 0.5181360244750977,
64
  "learning_rate": 5e-05,
65
- "loss": 1.8217,
66
  "step": 600
67
  },
68
  {
69
- "epoch": 1.3232514177693762,
70
- "grad_norm": 1.1348918676376343,
71
  "learning_rate": 5e-05,
72
- "loss": 1.7866,
73
  "step": 700
74
  },
75
  {
76
- "epoch": 1.5122873345935728,
77
- "grad_norm": 0.661758303642273,
78
  "learning_rate": 5e-05,
79
- "loss": 1.7934,
80
  "step": 800
81
  },
82
  {
83
- "epoch": 1.7013232514177694,
84
- "grad_norm": 0.5634772777557373,
85
  "learning_rate": 5e-05,
86
- "loss": 1.7731,
87
  "step": 900
88
  },
89
  {
90
- "epoch": 1.8903591682419658,
91
- "grad_norm": 0.6698930263519287,
92
  "learning_rate": 5e-05,
93
- "loss": 1.7969,
94
  "step": 1000
95
  },
96
  {
97
  "epoch": 2.0,
98
- "eval_accuracy": 0.5751794871794872,
99
- "eval_loss": 1.603736162185669,
100
- "eval_runtime": 5.0986,
101
- "eval_samples_per_second": 98.066,
102
- "eval_steps_per_second": 12.356,
103
  "step": 1058
104
  },
105
  {
106
  "epoch": 2.0,
107
- "eval_exact_match": 17.6,
108
- "eval_f1": 27.218095238095238,
109
  "step": 1058
110
  },
111
  {
112
- "epoch": 2.0793950850661624,
113
- "grad_norm": 0.6265193223953247,
114
  "learning_rate": 5e-05,
115
- "loss": 1.7416,
116
  "step": 1100
117
  },
118
  {
119
- "epoch": 2.268431001890359,
120
- "grad_norm": 0.812364935874939,
121
  "learning_rate": 5e-05,
122
- "loss": 1.6957,
123
  "step": 1200
124
  },
125
  {
126
- "epoch": 2.4574669187145557,
127
- "grad_norm": 1.0972926616668701,
128
  "learning_rate": 5e-05,
129
- "loss": 1.7033,
130
  "step": 1300
131
  },
132
  {
133
- "epoch": 2.6465028355387523,
134
- "grad_norm": 1.0831658840179443,
135
  "learning_rate": 5e-05,
136
- "loss": 1.6935,
137
  "step": 1400
138
  },
139
  {
140
- "epoch": 2.835538752362949,
141
- "grad_norm": 1.0022494792938232,
142
  "learning_rate": 5e-05,
143
- "loss": 1.6799,
144
  "step": 1500
145
  },
146
  {
147
  "epoch": 3.0,
148
- "eval_accuracy": 0.573025641025641,
149
- "eval_loss": 1.6428505182266235,
150
- "eval_runtime": 5.1488,
151
- "eval_samples_per_second": 97.111,
152
- "eval_steps_per_second": 12.236,
153
  "step": 1587
154
  },
155
  {
156
  "epoch": 3.0,
157
- "eval_exact_match": 20.2,
158
- "eval_f1": 27.93809523809523,
159
  "step": 1587
160
  },
161
  {
162
- "epoch": 3.0245746691871456,
163
- "grad_norm": 0.8240587711334229,
164
  "learning_rate": 5e-05,
165
- "loss": 1.6618,
166
  "step": 1600
167
  },
168
  {
169
- "epoch": 3.213610586011342,
170
- "grad_norm": 1.2994645833969116,
171
  "learning_rate": 5e-05,
172
- "loss": 1.5736,
173
  "step": 1700
174
  },
175
  {
176
- "epoch": 3.402646502835539,
177
- "grad_norm": 1.0891159772872925,
178
  "learning_rate": 5e-05,
179
- "loss": 1.5798,
180
  "step": 1800
181
  },
182
  {
183
- "epoch": 3.5916824196597354,
184
- "grad_norm": 1.3826475143432617,
185
  "learning_rate": 5e-05,
186
- "loss": 1.5957,
187
  "step": 1900
188
  },
189
  {
190
- "epoch": 3.780718336483932,
191
- "grad_norm": 0.9901586174964905,
192
  "learning_rate": 5e-05,
193
- "loss": 1.6026,
194
  "step": 2000
195
  },
196
  {
197
- "epoch": 3.9697542533081287,
198
- "grad_norm": 1.0607813596725464,
199
  "learning_rate": 5e-05,
200
- "loss": 1.6106,
201
  "step": 2100
202
  },
203
  {
204
  "epoch": 4.0,
205
- "eval_accuracy": 0.5708717948717948,
206
- "eval_loss": 1.7143596410751343,
207
- "eval_runtime": 5.3942,
208
- "eval_samples_per_second": 92.693,
209
- "eval_steps_per_second": 11.679,
210
  "step": 2116
211
  },
212
  {
213
  "epoch": 4.0,
214
- "eval_exact_match": 19.0,
215
- "eval_f1": 27.71428571428571,
216
  "step": 2116
217
  },
218
  {
219
- "epoch": 4.158790170132325,
220
- "grad_norm": 1.3189178705215454,
221
  "learning_rate": 5e-05,
222
- "loss": 1.5029,
223
  "step": 2200
224
  },
225
  {
226
- "epoch": 4.3478260869565215,
227
- "grad_norm": 1.062267541885376,
228
  "learning_rate": 5e-05,
229
- "loss": 1.51,
230
  "step": 2300
231
  },
232
  {
233
- "epoch": 4.536862003780718,
234
- "grad_norm": 1.3810843229293823,
235
  "learning_rate": 5e-05,
236
- "loss": 1.4867,
237
  "step": 2400
238
  },
239
  {
240
- "epoch": 4.725897920604915,
241
- "grad_norm": 1.1991653442382812,
242
  "learning_rate": 5e-05,
243
- "loss": 1.501,
244
  "step": 2500
245
  },
246
  {
247
- "epoch": 4.914933837429111,
248
- "grad_norm": 1.310654878616333,
249
  "learning_rate": 5e-05,
250
- "loss": 1.5034,
251
  "step": 2600
252
  },
253
  {
254
  "epoch": 5.0,
255
- "eval_accuracy": 0.5681025641025641,
256
- "eval_loss": 1.8328346014022827,
257
- "eval_runtime": 5.7763,
258
- "eval_samples_per_second": 86.561,
259
- "eval_steps_per_second": 10.907,
260
  "step": 2645
261
  },
262
  {
263
  "epoch": 5.0,
264
- "eval_exact_match": 18.8,
265
- "eval_f1": 26.6015873015873,
266
  "step": 2645
267
  },
268
  {
269
- "epoch": 5.103969754253308,
270
- "grad_norm": 1.1406631469726562,
271
  "learning_rate": 5e-05,
272
- "loss": 1.4605,
273
  "step": 2700
274
  },
275
  {
276
- "epoch": 5.293005671077505,
277
- "grad_norm": 1.447221279144287,
278
  "learning_rate": 5e-05,
279
- "loss": 1.4204,
280
  "step": 2800
281
  },
282
  {
283
- "epoch": 5.482041587901701,
284
- "grad_norm": 1.461799144744873,
285
  "learning_rate": 5e-05,
286
- "loss": 1.4233,
287
  "step": 2900
288
  },
289
  {
290
- "epoch": 5.671077504725898,
291
- "grad_norm": 1.4184092283248901,
292
  "learning_rate": 5e-05,
293
- "loss": 1.4265,
294
  "step": 3000
295
  },
296
  {
297
- "epoch": 5.8601134215500945,
298
- "grad_norm": 0.9850455522537231,
299
  "learning_rate": 5e-05,
300
- "loss": 1.4139,
301
  "step": 3100
302
  },
303
  {
304
  "epoch": 6.0,
305
- "eval_accuracy": 0.564974358974359,
306
- "eval_loss": 1.931396484375,
307
- "eval_runtime": 5.3358,
308
- "eval_samples_per_second": 93.706,
309
- "eval_steps_per_second": 11.807,
310
  "step": 3174
311
  },
312
  {
313
  "epoch": 6.0,
314
- "eval_exact_match": 17.6,
315
- "eval_f1": 25.464285714285715,
316
  "step": 3174
317
  },
318
  {
319
- "epoch": 6.049149338374291,
320
- "grad_norm": 1.2274116277694702,
321
  "learning_rate": 5e-05,
322
- "loss": 1.4159,
323
  "step": 3200
324
  },
325
  {
326
- "epoch": 6.238185255198488,
327
- "grad_norm": 1.2107988595962524,
328
  "learning_rate": 5e-05,
329
- "loss": 1.3579,
330
  "step": 3300
331
  },
332
  {
333
- "epoch": 6.427221172022684,
334
- "grad_norm": 1.2965952157974243,
335
  "learning_rate": 5e-05,
336
- "loss": 1.3485,
337
  "step": 3400
338
  },
339
  {
340
- "epoch": 6.616257088846881,
341
- "grad_norm": 1.3035151958465576,
342
  "learning_rate": 5e-05,
343
- "loss": 1.36,
344
  "step": 3500
345
  },
346
  {
347
- "epoch": 6.805293005671078,
348
- "grad_norm": 1.1840296983718872,
349
  "learning_rate": 5e-05,
350
- "loss": 1.3599,
351
  "step": 3600
352
  },
353
  {
354
- "epoch": 6.994328922495274,
355
- "grad_norm": 1.780287265777588,
356
  "learning_rate": 5e-05,
357
- "loss": 1.355,
358
  "step": 3700
359
  },
360
  {
361
  "epoch": 7.0,
362
- "eval_accuracy": 0.565076923076923,
363
- "eval_loss": 1.9659509658813477,
364
- "eval_runtime": 5.3127,
365
- "eval_samples_per_second": 94.114,
366
- "eval_steps_per_second": 11.858,
367
  "step": 3703
368
  },
369
  {
370
  "epoch": 7.0,
371
- "eval_exact_match": 18.0,
372
- "eval_f1": 26.546666666666663,
373
  "step": 3703
374
  },
375
  {
376
- "epoch": 7.183364839319471,
377
- "grad_norm": 1.1608421802520752,
378
  "learning_rate": 5e-05,
379
- "loss": 1.298,
380
  "step": 3800
381
  },
382
  {
383
- "epoch": 7.3724007561436675,
384
- "grad_norm": 1.083270788192749,
385
  "learning_rate": 5e-05,
386
- "loss": 1.3088,
387
  "step": 3900
388
  },
389
  {
390
- "epoch": 7.561436672967864,
391
- "grad_norm": 1.3489899635314941,
392
  "learning_rate": 5e-05,
393
- "loss": 1.299,
394
  "step": 4000
395
  },
396
  {
397
- "epoch": 7.750472589792061,
398
- "grad_norm": 1.1663447618484497,
399
  "learning_rate": 5e-05,
400
- "loss": 1.2841,
401
  "step": 4100
402
  },
403
  {
404
- "epoch": 7.939508506616257,
405
- "grad_norm": 1.9308583736419678,
406
  "learning_rate": 5e-05,
407
- "loss": 1.314,
408
  "step": 4200
409
  },
410
  {
411
  "epoch": 8.0,
412
- "eval_accuracy": 0.5616923076923077,
413
- "eval_loss": 2.1164443492889404,
414
- "eval_runtime": 5.2165,
415
- "eval_samples_per_second": 95.85,
416
- "eval_steps_per_second": 12.077,
417
  "step": 4232
418
  },
419
  {
420
  "epoch": 8.0,
421
- "eval_exact_match": 17.0,
422
- "eval_f1": 24.659841269841277,
423
  "step": 4232
424
  },
425
  {
426
- "epoch": 8.128544423440454,
427
- "grad_norm": 1.109632134437561,
428
  "learning_rate": 5e-05,
429
- "loss": 1.2558,
430
  "step": 4300
431
  },
432
  {
433
- "epoch": 8.31758034026465,
434
- "grad_norm": 1.1132491827011108,
435
  "learning_rate": 5e-05,
436
- "loss": 1.2264,
437
  "step": 4400
438
  },
439
  {
440
- "epoch": 8.506616257088847,
441
- "grad_norm": 1.1017369031906128,
442
  "learning_rate": 5e-05,
443
- "loss": 1.2649,
444
  "step": 4500
445
  },
446
  {
447
- "epoch": 8.695652173913043,
448
- "grad_norm": 1.0874963998794556,
449
  "learning_rate": 5e-05,
450
- "loss": 1.2512,
451
  "step": 4600
452
  },
453
  {
454
- "epoch": 8.88468809073724,
455
- "grad_norm": 1.2617064714431763,
456
  "learning_rate": 5e-05,
457
- "loss": 1.261,
458
  "step": 4700
459
  },
460
  {
461
  "epoch": 9.0,
462
- "eval_accuracy": 0.5622051282051282,
463
- "eval_loss": 2.1635327339172363,
464
- "eval_runtime": 5.0832,
465
- "eval_samples_per_second": 98.363,
466
- "eval_steps_per_second": 12.394,
467
  "step": 4761
468
  },
469
  {
470
  "epoch": 9.0,
471
- "eval_exact_match": 17.8,
472
- "eval_f1": 25.451428571428576,
473
  "step": 4761
474
  },
475
  {
476
- "epoch": 9.073724007561436,
477
- "grad_norm": 1.2258027791976929,
478
  "learning_rate": 5e-05,
479
- "loss": 1.2298,
480
  "step": 4800
481
  },
482
  {
483
- "epoch": 9.262759924385634,
484
- "grad_norm": 1.8261148929595947,
485
  "learning_rate": 5e-05,
486
- "loss": 1.1748,
487
  "step": 4900
488
  },
489
  {
490
- "epoch": 9.45179584120983,
491
- "grad_norm": 1.1221072673797607,
492
  "learning_rate": 5e-05,
493
- "loss": 1.19,
494
  "step": 5000
495
  },
496
  {
497
- "epoch": 9.640831758034027,
498
- "grad_norm": 1.3320170640945435,
499
  "learning_rate": 5e-05,
500
- "loss": 1.1932,
501
  "step": 5100
502
  },
503
  {
504
- "epoch": 9.829867674858223,
505
- "grad_norm": 1.194077730178833,
506
  "learning_rate": 5e-05,
507
- "loss": 1.217,
508
  "step": 5200
509
  },
510
  {
511
  "epoch": 10.0,
512
- "eval_accuracy": 0.5612820512820513,
513
- "eval_loss": 2.244041919708252,
514
- "eval_runtime": 5.1606,
515
- "eval_samples_per_second": 96.889,
516
- "eval_steps_per_second": 12.208,
517
  "step": 5290
518
  },
519
  {
520
  "epoch": 10.0,
521
- "eval_exact_match": 18.4,
522
- "eval_f1": 26.067662337662348,
523
  "step": 5290
524
  },
525
  {
526
- "epoch": 10.01890359168242,
527
- "grad_norm": 1.1352814435958862,
528
  "learning_rate": 5e-05,
529
- "loss": 1.2093,
530
  "step": 5300
531
  },
532
  {
533
- "epoch": 10.207939508506616,
534
- "grad_norm": 1.2741838693618774,
535
  "learning_rate": 5e-05,
536
- "loss": 1.1467,
537
  "step": 5400
538
  },
539
  {
540
- "epoch": 10.396975425330814,
541
- "grad_norm": 1.6901249885559082,
542
  "learning_rate": 5e-05,
543
- "loss": 1.1288,
544
  "step": 5500
545
  },
546
  {
547
- "epoch": 10.58601134215501,
548
- "grad_norm": 1.1944465637207031,
549
  "learning_rate": 5e-05,
550
- "loss": 1.1399,
551
  "step": 5600
552
  },
553
  {
554
- "epoch": 10.775047258979207,
555
- "grad_norm": 1.2234572172164917,
556
  "learning_rate": 5e-05,
557
- "loss": 1.1678,
558
  "step": 5700
559
  },
560
  {
561
- "epoch": 10.964083175803403,
562
- "grad_norm": 1.2781612873077393,
563
  "learning_rate": 5e-05,
564
- "loss": 1.146,
565
  "step": 5800
566
  },
567
  {
568
  "epoch": 11.0,
569
- "eval_accuracy": 0.5618461538461539,
570
- "eval_loss": 2.296630859375,
571
- "eval_runtime": 5.1078,
572
- "eval_samples_per_second": 97.889,
573
- "eval_steps_per_second": 12.334,
574
  "step": 5819
575
  },
576
  {
577
  "epoch": 11.0,
578
- "eval_exact_match": 16.8,
579
- "eval_f1": 23.63305916305917,
580
  "step": 5819
581
  },
582
  {
583
- "epoch": 11.1531190926276,
584
- "grad_norm": 1.138975739479065,
585
  "learning_rate": 5e-05,
586
- "loss": 1.0842,
587
  "step": 5900
588
  },
589
  {
590
- "epoch": 11.342155009451796,
591
- "grad_norm": 1.727989912033081,
592
  "learning_rate": 5e-05,
593
- "loss": 1.0806,
594
  "step": 6000
595
  },
596
  {
597
- "epoch": 11.531190926275993,
598
- "grad_norm": 1.3104358911514282,
599
  "learning_rate": 5e-05,
600
- "loss": 1.0953,
601
  "step": 6100
602
  },
603
  {
604
- "epoch": 11.720226843100189,
605
- "grad_norm": 1.3629850149154663,
606
  "learning_rate": 5e-05,
607
- "loss": 1.0945,
608
  "step": 6200
609
  },
610
  {
611
- "epoch": 11.909262759924385,
612
- "grad_norm": 1.3276056051254272,
613
  "learning_rate": 5e-05,
614
- "loss": 1.1069,
615
  "step": 6300
616
  },
617
  {
618
  "epoch": 12.0,
619
- "eval_accuracy": 0.562,
620
- "eval_loss": 2.3118155002593994,
621
- "eval_runtime": 5.3837,
622
- "eval_samples_per_second": 92.873,
623
- "eval_steps_per_second": 11.702,
624
  "step": 6348
625
  },
626
  {
627
  "epoch": 12.0,
628
- "eval_exact_match": 17.4,
629
- "eval_f1": 24.43924963924964,
630
  "step": 6348
631
  }
632
  ],
633
  "logging_steps": 100,
634
  "max_steps": 26450,
635
- "num_input_tokens_seen": 0,
636
  "num_train_epochs": 50,
637
  "save_steps": 500,
638
- "stateful_callbacks": {
639
- "TrainerControl": {
640
- "args": {
641
- "should_epoch_stop": false,
642
- "should_evaluate": false,
643
- "should_log": false,
644
- "should_save": true,
645
- "should_training_stop": false
646
- },
647
- "attributes": {}
648
- }
649
- },
650
- "total_flos": 3.9105998530543616e+17,
651
- "train_batch_size": 1,
652
  "trial_name": null,
653
  "trial_params": null
654
  }
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.19,
 
13
  "learning_rate": 5e-05,
14
+ "loss": 1.6355,
15
  "step": 100
16
  },
17
  {
18
+ "epoch": 0.38,
 
19
  "learning_rate": 5e-05,
20
+ "loss": 1.4037,
21
  "step": 200
22
  },
23
  {
24
+ "epoch": 0.57,
 
25
  "learning_rate": 5e-05,
26
+ "loss": 1.4183,
27
  "step": 300
28
  },
29
  {
30
+ "epoch": 0.76,
 
31
  "learning_rate": 5e-05,
32
+ "loss": 1.3881,
33
  "step": 400
34
  },
35
  {
36
+ "epoch": 0.95,
 
37
  "learning_rate": 5e-05,
38
+ "loss": 1.3891,
39
  "step": 500
40
  },
41
  {
42
  "epoch": 1.0,
43
+ "eval_accuracy": 0.6137948717948718,
44
+ "eval_loss": 1.3015451431274414,
45
+ "eval_runtime": 4.1821,
46
+ "eval_samples_per_second": 119.557,
47
+ "eval_steps_per_second": 15.064,
48
  "step": 529
49
  },
50
  {
51
  "epoch": 1.0,
52
+ "eval_exact_match": 23.2,
53
+ "eval_f1": 35.17014759120022,
54
  "step": 529
55
  },
56
  {
57
+ "epoch": 1.13,
 
58
  "learning_rate": 5e-05,
59
+ "loss": 1.3808,
60
  "step": 600
61
  },
62
  {
63
+ "epoch": 1.32,
 
64
  "learning_rate": 5e-05,
65
+ "loss": 1.348,
66
  "step": 700
67
  },
68
  {
69
+ "epoch": 1.51,
 
70
  "learning_rate": 5e-05,
71
+ "loss": 1.3627,
72
  "step": 800
73
  },
74
  {
75
+ "epoch": 1.7,
 
76
  "learning_rate": 5e-05,
77
+ "loss": 1.3473,
78
  "step": 900
79
  },
80
  {
81
+ "epoch": 1.89,
 
82
  "learning_rate": 5e-05,
83
+ "loss": 1.3633,
84
  "step": 1000
85
  },
86
  {
87
  "epoch": 2.0,
88
+ "eval_accuracy": 0.6165641025641025,
89
+ "eval_loss": 1.2854759693145752,
90
+ "eval_runtime": 4.1064,
91
+ "eval_samples_per_second": 121.761,
92
+ "eval_steps_per_second": 15.342,
93
  "step": 1058
94
  },
95
  {
96
  "epoch": 2.0,
97
+ "eval_exact_match": 25.8,
98
+ "eval_f1": 37.509047619047635,
99
  "step": 1058
100
  },
101
  {
102
+ "epoch": 2.08,
 
103
  "learning_rate": 5e-05,
104
+ "loss": 1.3358,
105
  "step": 1100
106
  },
107
  {
108
+ "epoch": 2.27,
 
109
  "learning_rate": 5e-05,
110
+ "loss": 1.2917,
111
  "step": 1200
112
  },
113
  {
114
+ "epoch": 2.46,
 
115
  "learning_rate": 5e-05,
116
+ "loss": 1.2948,
117
  "step": 1300
118
  },
119
  {
120
+ "epoch": 2.65,
 
121
  "learning_rate": 5e-05,
122
+ "loss": 1.3144,
123
  "step": 1400
124
  },
125
  {
126
+ "epoch": 2.84,
 
127
  "learning_rate": 5e-05,
128
+ "loss": 1.2929,
129
  "step": 1500
130
  },
131
  {
132
  "epoch": 3.0,
133
+ "eval_accuracy": 0.6176923076923077,
134
+ "eval_loss": 1.2953603267669678,
135
+ "eval_runtime": 4.0992,
136
+ "eval_samples_per_second": 121.974,
137
+ "eval_steps_per_second": 15.369,
138
  "step": 1587
139
  },
140
  {
141
  "epoch": 3.0,
142
+ "eval_exact_match": 25.0,
143
+ "eval_f1": 37.08253968253969,
144
  "step": 1587
145
  },
146
  {
147
+ "epoch": 3.02,
 
148
  "learning_rate": 5e-05,
149
+ "loss": 1.2906,
150
  "step": 1600
151
  },
152
  {
153
+ "epoch": 3.21,
 
154
  "learning_rate": 5e-05,
155
+ "loss": 1.2216,
156
  "step": 1700
157
  },
158
  {
159
+ "epoch": 3.4,
 
160
  "learning_rate": 5e-05,
161
+ "loss": 1.2459,
162
  "step": 1800
163
  },
164
  {
165
+ "epoch": 3.59,
 
166
  "learning_rate": 5e-05,
167
+ "loss": 1.2474,
168
  "step": 1900
169
  },
170
  {
171
+ "epoch": 3.78,
 
172
  "learning_rate": 5e-05,
173
+ "loss": 1.2518,
174
  "step": 2000
175
  },
176
  {
177
+ "epoch": 3.97,
 
178
  "learning_rate": 5e-05,
179
+ "loss": 1.2361,
180
  "step": 2100
181
  },
182
  {
183
  "epoch": 4.0,
184
+ "eval_accuracy": 0.6045128205128205,
185
+ "eval_loss": 1.3488503694534302,
186
+ "eval_runtime": 4.2015,
187
+ "eval_samples_per_second": 119.004,
188
+ "eval_steps_per_second": 14.995,
189
  "step": 2116
190
  },
191
  {
192
  "epoch": 4.0,
193
+ "eval_exact_match": 25.0,
194
+ "eval_f1": 36.82126984126986,
195
  "step": 2116
196
  },
197
  {
198
+ "epoch": 4.16,
 
199
  "learning_rate": 5e-05,
200
+ "loss": 1.1856,
201
  "step": 2200
202
  },
203
  {
204
+ "epoch": 4.35,
 
205
  "learning_rate": 5e-05,
206
+ "loss": 1.1556,
207
  "step": 2300
208
  },
209
  {
210
+ "epoch": 4.54,
 
211
  "learning_rate": 5e-05,
212
+ "loss": 1.1828,
213
  "step": 2400
214
  },
215
  {
216
+ "epoch": 4.73,
 
217
  "learning_rate": 5e-05,
218
+ "loss": 1.1687,
219
  "step": 2500
220
  },
221
  {
222
+ "epoch": 4.91,
 
223
  "learning_rate": 5e-05,
224
+ "loss": 1.1856,
225
  "step": 2600
226
  },
227
  {
228
  "epoch": 5.0,
229
+ "eval_accuracy": 0.6125128205128205,
230
+ "eval_loss": 1.3968186378479004,
231
+ "eval_runtime": 4.3081,
232
+ "eval_samples_per_second": 116.06,
233
+ "eval_steps_per_second": 14.624,
234
  "step": 2645
235
  },
236
  {
237
  "epoch": 5.0,
238
+ "eval_exact_match": 24.6,
239
+ "eval_f1": 37.079523809523806,
240
  "step": 2645
241
  },
242
  {
243
+ "epoch": 5.1,
 
244
  "learning_rate": 5e-05,
245
+ "loss": 1.1376,
246
  "step": 2700
247
  },
248
  {
249
+ "epoch": 5.29,
 
250
  "learning_rate": 5e-05,
251
+ "loss": 1.1079,
252
  "step": 2800
253
  },
254
  {
255
+ "epoch": 5.48,
 
256
  "learning_rate": 5e-05,
257
+ "loss": 1.0954,
258
  "step": 2900
259
  },
260
  {
261
+ "epoch": 5.67,
 
262
  "learning_rate": 5e-05,
263
+ "loss": 1.1317,
264
  "step": 3000
265
  },
266
  {
267
+ "epoch": 5.86,
 
268
  "learning_rate": 5e-05,
269
+ "loss": 1.1098,
270
  "step": 3100
271
  },
272
  {
273
  "epoch": 6.0,
274
+ "eval_accuracy": 0.6114871794871795,
275
+ "eval_loss": 1.4720605611801147,
276
+ "eval_runtime": 4.5,
277
+ "eval_samples_per_second": 111.11,
278
+ "eval_steps_per_second": 14.0,
279
  "step": 3174
280
  },
281
  {
282
  "epoch": 6.0,
283
+ "eval_exact_match": 24.4,
284
+ "eval_f1": 35.78015873015873,
285
  "step": 3174
286
  },
287
  {
288
+ "epoch": 6.05,
 
289
  "learning_rate": 5e-05,
290
+ "loss": 1.0945,
291
  "step": 3200
292
  },
293
  {
294
+ "epoch": 6.24,
 
295
  "learning_rate": 5e-05,
296
+ "loss": 1.0309,
297
  "step": 3300
298
  },
299
  {
300
+ "epoch": 6.43,
 
301
  "learning_rate": 5e-05,
302
+ "loss": 1.0389,
303
  "step": 3400
304
  },
305
  {
306
+ "epoch": 6.62,
 
307
  "learning_rate": 5e-05,
308
+ "loss": 1.0547,
309
  "step": 3500
310
  },
311
  {
312
+ "epoch": 6.81,
 
313
  "learning_rate": 5e-05,
314
+ "loss": 1.0618,
315
  "step": 3600
316
  },
317
  {
318
+ "epoch": 6.99,
 
319
  "learning_rate": 5e-05,
320
+ "loss": 1.0753,
321
  "step": 3700
322
  },
323
  {
324
  "epoch": 7.0,
325
+ "eval_accuracy": 0.6076410256410256,
326
+ "eval_loss": 1.5797967910766602,
327
+ "eval_runtime": 4.8128,
328
+ "eval_samples_per_second": 103.891,
329
+ "eval_steps_per_second": 13.09,
330
  "step": 3703
331
  },
332
  {
333
  "epoch": 7.0,
334
+ "eval_exact_match": 21.8,
335
+ "eval_f1": 33.15111111111111,
336
  "step": 3703
337
  },
338
  {
339
+ "epoch": 7.18,
 
340
  "learning_rate": 5e-05,
341
+ "loss": 0.9727,
342
  "step": 3800
343
  },
344
  {
345
+ "epoch": 7.37,
 
346
  "learning_rate": 5e-05,
347
+ "loss": 0.9815,
348
  "step": 3900
349
  },
350
  {
351
+ "epoch": 7.56,
 
352
  "learning_rate": 5e-05,
353
+ "loss": 0.9777,
354
  "step": 4000
355
  },
356
  {
357
+ "epoch": 7.75,
 
358
  "learning_rate": 5e-05,
359
+ "loss": 0.9984,
360
  "step": 4100
361
  },
362
  {
363
+ "epoch": 7.94,
 
364
  "learning_rate": 5e-05,
365
+ "loss": 1.0048,
366
  "step": 4200
367
  },
368
  {
369
  "epoch": 8.0,
370
+ "eval_accuracy": 0.6084102564102564,
371
+ "eval_loss": 1.6041829586029053,
372
+ "eval_runtime": 4.4087,
373
+ "eval_samples_per_second": 113.411,
374
+ "eval_steps_per_second": 14.29,
375
  "step": 4232
376
  },
377
  {
378
  "epoch": 8.0,
379
+ "eval_exact_match": 22.4,
380
+ "eval_f1": 35.09904761904763,
381
  "step": 4232
382
  },
383
  {
384
+ "epoch": 8.13,
 
385
  "learning_rate": 5e-05,
386
+ "loss": 0.9568,
387
  "step": 4300
388
  },
389
  {
390
+ "epoch": 8.32,
 
391
  "learning_rate": 5e-05,
392
+ "loss": 0.9088,
393
  "step": 4400
394
  },
395
  {
396
+ "epoch": 8.51,
 
397
  "learning_rate": 5e-05,
398
+ "loss": 0.9341,
399
  "step": 4500
400
  },
401
  {
402
+ "epoch": 8.7,
 
403
  "learning_rate": 5e-05,
404
+ "loss": 0.9285,
405
  "step": 4600
406
  },
407
  {
408
+ "epoch": 8.88,
 
409
  "learning_rate": 5e-05,
410
+ "loss": 0.9456,
411
  "step": 4700
412
  },
413
  {
414
  "epoch": 9.0,
415
+ "eval_accuracy": 0.5976923076923077,
416
+ "eval_loss": 1.684310793876648,
417
+ "eval_runtime": 4.9249,
418
+ "eval_samples_per_second": 101.524,
419
+ "eval_steps_per_second": 12.792,
420
  "step": 4761
421
  },
422
  {
423
  "epoch": 9.0,
424
+ "eval_exact_match": 24.6,
425
+ "eval_f1": 36.32493506493508,
426
  "step": 4761
427
  },
428
  {
429
+ "epoch": 9.07,
 
430
  "learning_rate": 5e-05,
431
+ "loss": 0.912,
432
  "step": 4800
433
  },
434
  {
435
+ "epoch": 9.26,
 
436
  "learning_rate": 5e-05,
437
+ "loss": 0.8443,
438
  "step": 4900
439
  },
440
  {
441
+ "epoch": 9.45,
 
442
  "learning_rate": 5e-05,
443
+ "loss": 0.8789,
444
  "step": 5000
445
  },
446
  {
447
+ "epoch": 9.64,
 
448
  "learning_rate": 5e-05,
449
+ "loss": 0.8865,
450
  "step": 5100
451
  },
452
  {
453
+ "epoch": 9.83,
 
454
  "learning_rate": 5e-05,
455
+ "loss": 0.8766,
456
  "step": 5200
457
  },
458
  {
459
  "epoch": 10.0,
460
+ "eval_accuracy": 0.6051282051282051,
461
+ "eval_loss": 1.7828805446624756,
462
+ "eval_runtime": 4.1983,
463
+ "eval_samples_per_second": 119.096,
464
+ "eval_steps_per_second": 15.006,
465
  "step": 5290
466
  },
467
  {
468
  "epoch": 10.0,
469
+ "eval_exact_match": 23.0,
470
+ "eval_f1": 34.336190476190495,
471
  "step": 5290
472
  },
473
  {
474
+ "epoch": 10.02,
 
475
  "learning_rate": 5e-05,
476
+ "loss": 0.8835,
477
  "step": 5300
478
  },
479
  {
480
+ "epoch": 10.21,
 
481
  "learning_rate": 5e-05,
482
+ "loss": 0.8095,
483
  "step": 5400
484
  },
485
  {
486
+ "epoch": 10.4,
 
487
  "learning_rate": 5e-05,
488
+ "loss": 0.8008,
489
  "step": 5500
490
  },
491
  {
492
+ "epoch": 10.59,
 
493
  "learning_rate": 5e-05,
494
+ "loss": 0.8208,
495
  "step": 5600
496
  },
497
  {
498
+ "epoch": 10.78,
 
499
  "learning_rate": 5e-05,
500
+ "loss": 0.8282,
501
  "step": 5700
502
  },
503
  {
504
+ "epoch": 10.96,
 
505
  "learning_rate": 5e-05,
506
+ "loss": 0.8273,
507
  "step": 5800
508
  },
509
  {
510
  "epoch": 11.0,
511
+ "eval_accuracy": 0.6043076923076923,
512
+ "eval_loss": 1.8059720993041992,
513
+ "eval_runtime": 4.2264,
514
+ "eval_samples_per_second": 118.304,
515
+ "eval_steps_per_second": 14.906,
516
  "step": 5819
517
  },
518
  {
519
  "epoch": 11.0,
520
+ "eval_exact_match": 21.6,
521
+ "eval_f1": 32.628095238095234,
522
  "step": 5819
523
  },
524
  {
525
+ "epoch": 11.15,
 
526
  "learning_rate": 5e-05,
527
+ "loss": 0.7637,
528
  "step": 5900
529
  },
530
  {
531
+ "epoch": 11.34,
 
532
  "learning_rate": 5e-05,
533
+ "loss": 0.7581,
534
  "step": 6000
535
  },
536
  {
537
+ "epoch": 11.53,
 
538
  "learning_rate": 5e-05,
539
+ "loss": 0.7571,
540
  "step": 6100
541
  },
542
  {
543
+ "epoch": 11.72,
 
544
  "learning_rate": 5e-05,
545
+ "loss": 0.7637,
546
  "step": 6200
547
  },
548
  {
549
+ "epoch": 11.91,
 
550
  "learning_rate": 5e-05,
551
+ "loss": 0.7755,
552
  "step": 6300
553
  },
554
  {
555
  "epoch": 12.0,
556
+ "eval_accuracy": 0.6018974358974359,
557
+ "eval_loss": 1.8729300498962402,
558
+ "eval_runtime": 4.2001,
559
+ "eval_samples_per_second": 119.045,
560
+ "eval_steps_per_second": 15.0,
561
  "step": 6348
562
  },
563
  {
564
  "epoch": 12.0,
565
+ "eval_exact_match": 16.8,
566
+ "eval_f1": 27.317351339157547,
567
  "step": 6348
568
  }
569
  ],
570
  "logging_steps": 100,
571
  "max_steps": 26450,
 
572
  "num_train_epochs": 50,
573
  "save_steps": 500,
574
+ "total_flos": 1.0726438168187372e+18,
 
 
 
 
 
 
 
 
 
 
 
 
 
575
  "trial_name": null,
576
  "trial_params": null
577
  }
checkpoint-6348/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f01d18f717251d8791982fc93ed5352f2d247add98e1102e2a52cc0a32128bb7
3
- size 5304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:391c04887a1ac777804ab56ff52563bcf9c4927713b353f37947aac9be091742
3
+ size 4728