tyzhu commited on
Commit
7b938bc
1 Parent(s): 73c6171

Training in progress, epoch 34, checkpoint

Browse files
checkpoint-8500/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-8500/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1033854b724415c3b8697de0ed4d5461ebca182b1a64570546156c8f4e73e929
3
+ size 143269386
checkpoint-8500/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-8500/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:41a28b8f4a29dcf861781506e342933ebaea7b920201898fcc77367acea8cfea
3
- size 224537202
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a326f16660f3ce5f2d6b52184cc5d59a3c83bec26effef285159782d38ae0678
3
+ size 286585234
checkpoint-8500/rng_state_0.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:026027772fa37788e8dfb295079b05b901fa451848b60d7d3da8208c27f3cedd
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8a8b32bc19a8b41a568236e0eae6796e31a7c9c8ef25ee985e952fd9f50fc01
3
  size 15024
checkpoint-8500/rng_state_1.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c08ab1a9cf9d7156ce777a687ec61ed55e14ad166148ffc340944a2bf891b8ce
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18971e11f3937b501c01013eacf4b48af5df6f4f5ac967e1b8c697f0cdb029b0
3
  size 15024
checkpoint-8500/rng_state_2.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:88b587e0f370204e89b20a0d48c2ab38534b0ec182389c533dd7aca1ce07fe3d
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6a768cf77b0a0e8079561c9b3718d3df5e5d1c032db41a10b73935a3297896e
3
  size 15024
checkpoint-8500/rng_state_3.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:50572f0d8906e71ef969ef080ae7308f6111d4c4c75d753cb77d6e1f06b72a7b
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:751f6054a0f5700b4e1692ae142121ded9832a7229d7615a7c73c6a8af2832da
3
  size 15024
checkpoint-8500/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-8500/tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
checkpoint-8500/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
checkpoint-8500/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-8500/trainer_state.json CHANGED
@@ -10,1129 +10,1030 @@
10
  "log_history": [
11
  {
12
  "epoch": 0.4,
13
- "grad_norm": 0.48472875356674194,
14
  "learning_rate": 0.0003,
15
- "loss": 2.3219,
16
  "step": 100
17
  },
18
  {
19
  "epoch": 0.8,
20
- "grad_norm": 0.4583953619003296,
21
  "learning_rate": 0.0003,
22
- "loss": 2.2487,
23
  "step": 200
24
  },
25
  {
26
  "epoch": 1.0,
27
- "eval_accuracy": 0.5168253968253969,
28
- "eval_loss": 2.311781167984009,
29
- "eval_runtime": 5.4425,
30
- "eval_samples_per_second": 91.87,
31
- "eval_steps_per_second": 11.576,
32
  "step": 250
33
  },
34
  {
35
  "epoch": 1.0,
36
- "eval_exact_match": 18.6,
37
- "eval_f1": 26.999121042356347,
38
  "step": 250
39
  },
40
  {
41
  "epoch": 1.2,
42
- "grad_norm": 0.6610094904899597,
43
  "learning_rate": 0.0003,
44
- "loss": 2.083,
45
  "step": 300
46
  },
47
  {
48
  "epoch": 1.6,
49
- "grad_norm": 0.7361629605293274,
50
  "learning_rate": 0.0003,
51
- "loss": 1.9274,
52
  "step": 400
53
  },
54
  {
55
  "epoch": 2.0,
56
- "grad_norm": 0.6793610453605652,
57
  "learning_rate": 0.0003,
58
- "loss": 1.9113,
59
  "step": 500
60
  },
61
  {
62
  "epoch": 2.0,
63
- "eval_accuracy": 0.5153650793650794,
64
- "eval_loss": 2.3749232292175293,
65
- "eval_runtime": 5.2167,
66
- "eval_samples_per_second": 95.846,
67
- "eval_steps_per_second": 12.077,
68
  "step": 500
69
  },
70
  {
71
  "epoch": 2.0,
72
- "eval_exact_match": 18.2,
73
- "eval_f1": 26.209704184704197,
74
  "step": 500
75
  },
76
  {
77
  "epoch": 2.4,
78
- "grad_norm": 0.8843489289283752,
79
  "learning_rate": 0.0003,
80
- "loss": 1.4602,
81
  "step": 600
82
  },
83
  {
84
  "epoch": 2.8,
85
- "grad_norm": 0.8380613923072815,
86
  "learning_rate": 0.0003,
87
- "loss": 1.4924,
88
  "step": 700
89
  },
90
  {
91
  "epoch": 3.0,
92
- "eval_accuracy": 0.5098730158730159,
93
- "eval_loss": 2.5289740562438965,
94
- "eval_runtime": 5.3631,
95
- "eval_samples_per_second": 93.23,
96
- "eval_steps_per_second": 11.747,
97
  "step": 750
98
  },
99
  {
100
  "epoch": 3.0,
101
- "eval_exact_match": 16.6,
102
- "eval_f1": 24.526153291153296,
103
  "step": 750
104
  },
105
  {
106
  "epoch": 3.2,
107
- "grad_norm": 1.068247675895691,
108
  "learning_rate": 0.0003,
109
- "loss": 1.2869,
110
  "step": 800
111
  },
112
  {
113
  "epoch": 3.6,
114
- "grad_norm": 1.041741967201233,
115
  "learning_rate": 0.0003,
116
- "loss": 1.0749,
117
  "step": 900
118
  },
119
  {
120
  "epoch": 4.0,
121
- "grad_norm": 1.267968773841858,
122
  "learning_rate": 0.0003,
123
- "loss": 1.1157,
124
  "step": 1000
125
  },
126
  {
127
  "epoch": 4.0,
128
- "eval_accuracy": 0.5038412698412699,
129
- "eval_loss": 2.794942855834961,
130
- "eval_runtime": 5.0088,
131
- "eval_samples_per_second": 99.825,
132
- "eval_steps_per_second": 12.578,
133
  "step": 1000
134
  },
135
  {
136
  "epoch": 4.0,
137
- "eval_exact_match": 16.4,
138
- "eval_f1": 23.978561993561993,
139
  "step": 1000
140
  },
141
  {
142
  "epoch": 4.4,
143
- "grad_norm": 1.1971005201339722,
144
  "learning_rate": 0.0003,
145
- "loss": 0.7367,
146
  "step": 1100
147
  },
148
  {
149
  "epoch": 4.8,
150
- "grad_norm": 1.2383127212524414,
151
  "learning_rate": 0.0003,
152
- "loss": 0.7814,
153
  "step": 1200
154
  },
155
  {
156
  "epoch": 5.0,
157
- "eval_accuracy": 0.49853968253968256,
158
- "eval_loss": 3.0137181282043457,
159
- "eval_runtime": 5.1368,
160
- "eval_samples_per_second": 97.337,
161
- "eval_steps_per_second": 12.264,
162
  "step": 1250
163
  },
164
  {
165
  "epoch": 5.0,
166
- "eval_exact_match": 14.8,
167
- "eval_f1": 21.880294705294705,
168
  "step": 1250
169
  },
170
  {
171
  "epoch": 5.2,
172
- "grad_norm": 0.9558005928993225,
173
  "learning_rate": 0.0003,
174
- "loss": 0.6786,
175
  "step": 1300
176
  },
177
  {
178
  "epoch": 5.6,
179
- "grad_norm": 1.1341361999511719,
180
  "learning_rate": 0.0003,
181
- "loss": 0.5741,
182
  "step": 1400
183
  },
184
  {
185
  "epoch": 6.0,
186
- "grad_norm": 1.1375033855438232,
187
  "learning_rate": 0.0003,
188
- "loss": 0.605,
189
  "step": 1500
190
  },
191
  {
192
  "epoch": 6.0,
193
- "eval_accuracy": 0.49657142857142855,
194
- "eval_loss": 3.220507860183716,
195
- "eval_runtime": 5.2708,
196
- "eval_samples_per_second": 94.863,
197
- "eval_steps_per_second": 11.953,
198
  "step": 1500
199
  },
200
  {
201
  "epoch": 6.0,
202
- "eval_exact_match": 15.4,
203
- "eval_f1": 22.281807987334307,
204
  "step": 1500
205
  },
206
  {
207
  "epoch": 6.4,
208
- "grad_norm": 1.118528962135315,
209
  "learning_rate": 0.0003,
210
- "loss": 0.4684,
211
  "step": 1600
212
  },
213
  {
214
  "epoch": 6.8,
215
- "grad_norm": 0.9311291575431824,
216
  "learning_rate": 0.0003,
217
- "loss": 0.4968,
218
  "step": 1700
219
  },
220
  {
221
  "epoch": 7.0,
222
- "eval_accuracy": 0.4946984126984127,
223
- "eval_loss": 3.3760199546813965,
224
- "eval_runtime": 5.0734,
225
- "eval_samples_per_second": 98.554,
226
- "eval_steps_per_second": 12.418,
227
  "step": 1750
228
  },
229
  {
230
  "epoch": 7.0,
231
- "eval_exact_match": 14.8,
232
- "eval_f1": 21.79480811586075,
233
  "step": 1750
234
  },
235
  {
236
  "epoch": 7.2,
237
- "grad_norm": 1.0299904346466064,
238
  "learning_rate": 0.0003,
239
- "loss": 0.4631,
240
  "step": 1800
241
  },
242
  {
243
  "epoch": 7.6,
244
- "grad_norm": 0.7518107295036316,
245
  "learning_rate": 0.0003,
246
- "loss": 0.4411,
247
  "step": 1900
248
  },
249
  {
250
  "epoch": 8.0,
251
- "grad_norm": 1.1140515804290771,
252
  "learning_rate": 0.0003,
253
- "loss": 0.4597,
254
  "step": 2000
255
  },
256
  {
257
  "epoch": 8.0,
258
- "eval_accuracy": 0.49234920634920637,
259
- "eval_loss": 3.4634218215942383,
260
- "eval_runtime": 5.8652,
261
- "eval_samples_per_second": 85.249,
262
- "eval_steps_per_second": 10.741,
263
  "step": 2000
264
  },
265
  {
266
  "epoch": 8.0,
267
- "eval_exact_match": 15.0,
268
- "eval_f1": 21.617115196631318,
269
  "step": 2000
270
  },
271
  {
272
  "epoch": 8.4,
273
- "grad_norm": 1.2363754510879517,
274
  "learning_rate": 0.0003,
275
- "loss": 0.4026,
276
  "step": 2100
277
  },
278
  {
279
  "epoch": 8.8,
280
- "grad_norm": 0.8174892067909241,
281
  "learning_rate": 0.0003,
282
- "loss": 0.4242,
283
  "step": 2200
284
  },
285
  {
286
  "epoch": 9.0,
287
- "eval_accuracy": 0.4955873015873016,
288
- "eval_loss": 3.4379212856292725,
289
- "eval_runtime": 5.214,
290
- "eval_samples_per_second": 95.896,
291
- "eval_steps_per_second": 12.083,
292
  "step": 2250
293
  },
294
  {
295
  "epoch": 9.0,
296
- "eval_exact_match": 15.0,
297
- "eval_f1": 22.36426184926185,
298
  "step": 2250
299
  },
300
  {
301
  "epoch": 9.2,
302
- "grad_norm": 0.551261842250824,
303
  "learning_rate": 0.0003,
304
- "loss": 0.4135,
305
  "step": 2300
306
  },
307
  {
308
  "epoch": 9.6,
309
- "grad_norm": 0.6361530423164368,
310
  "learning_rate": 0.0003,
311
- "loss": 0.3999,
312
  "step": 2400
313
  },
314
  {
315
  "epoch": 10.0,
316
- "grad_norm": 0.9359388947486877,
317
  "learning_rate": 0.0003,
318
- "loss": 0.4223,
319
  "step": 2500
320
  },
321
  {
322
  "epoch": 10.0,
323
- "eval_accuracy": 0.494952380952381,
324
- "eval_loss": 3.600064277648926,
325
- "eval_runtime": 5.8832,
326
- "eval_samples_per_second": 84.988,
327
- "eval_steps_per_second": 10.709,
328
  "step": 2500
329
  },
330
  {
331
  "epoch": 10.0,
332
- "eval_exact_match": 14.8,
333
- "eval_f1": 21.784345099345103,
334
  "step": 2500
335
  },
336
  {
337
  "epoch": 10.4,
338
- "grad_norm": 0.7834428548812866,
339
  "learning_rate": 0.0003,
340
- "loss": 0.3855,
341
  "step": 2600
342
  },
343
  {
344
  "epoch": 10.8,
345
- "grad_norm": 1.0201836824417114,
346
  "learning_rate": 0.0003,
347
- "loss": 0.4,
348
  "step": 2700
349
  },
350
  {
351
  "epoch": 11.0,
352
- "eval_accuracy": 0.4955555555555556,
353
- "eval_loss": 3.571756362915039,
354
- "eval_runtime": 5.6646,
355
- "eval_samples_per_second": 88.267,
356
- "eval_steps_per_second": 11.122,
357
  "step": 2750
358
  },
359
  {
360
  "epoch": 11.0,
361
- "eval_exact_match": 15.0,
362
- "eval_f1": 22.02530358530358,
363
  "step": 2750
364
  },
365
  {
366
  "epoch": 11.2,
367
- "grad_norm": 0.738467812538147,
368
  "learning_rate": 0.0003,
369
- "loss": 0.3977,
370
  "step": 2800
371
  },
372
  {
373
  "epoch": 11.6,
374
- "grad_norm": 0.535826563835144,
375
  "learning_rate": 0.0003,
376
- "loss": 0.3904,
377
  "step": 2900
378
  },
379
  {
380
  "epoch": 12.0,
381
- "grad_norm": 0.8586340546607971,
382
  "learning_rate": 0.0003,
383
- "loss": 0.4007,
384
  "step": 3000
385
  },
386
  {
387
  "epoch": 12.0,
388
- "eval_accuracy": 0.49323809523809525,
389
- "eval_loss": 3.568419933319092,
390
- "eval_runtime": 5.5173,
391
- "eval_samples_per_second": 90.624,
392
- "eval_steps_per_second": 11.419,
393
  "step": 3000
394
  },
395
  {
396
  "epoch": 12.0,
397
- "eval_exact_match": 15.2,
398
- "eval_f1": 21.94431457431457,
399
  "step": 3000
400
  },
401
  {
402
  "epoch": 12.4,
403
- "grad_norm": 0.6024760007858276,
404
  "learning_rate": 0.0003,
405
- "loss": 0.3721,
406
  "step": 3100
407
  },
408
  {
409
  "epoch": 12.8,
410
- "grad_norm": 0.7764168381690979,
411
  "learning_rate": 0.0003,
412
- "loss": 0.3929,
413
  "step": 3200
414
  },
415
  {
416
  "epoch": 13.0,
417
- "eval_accuracy": 0.49314285714285716,
418
- "eval_loss": 3.602919816970825,
419
- "eval_runtime": 5.1504,
420
- "eval_samples_per_second": 97.08,
421
- "eval_steps_per_second": 12.232,
422
  "step": 3250
423
  },
424
  {
425
  "epoch": 13.0,
426
- "eval_exact_match": 13.8,
427
- "eval_f1": 20.855486735486743,
428
  "step": 3250
429
  },
430
  {
431
  "epoch": 13.2,
432
- "grad_norm": 0.668953537940979,
433
  "learning_rate": 0.0003,
434
- "loss": 0.3886,
435
  "step": 3300
436
  },
437
  {
438
  "epoch": 13.6,
439
- "grad_norm": 0.7968106865882874,
440
  "learning_rate": 0.0003,
441
- "loss": 0.3845,
442
  "step": 3400
443
  },
444
  {
445
  "epoch": 14.0,
446
- "grad_norm": 0.7598584294319153,
447
  "learning_rate": 0.0003,
448
- "loss": 0.4003,
449
  "step": 3500
450
  },
451
  {
452
  "epoch": 14.0,
453
- "eval_accuracy": 0.49206349206349204,
454
- "eval_loss": 3.584076404571533,
455
- "eval_runtime": 5.5138,
456
- "eval_samples_per_second": 90.682,
457
- "eval_steps_per_second": 11.426,
458
  "step": 3500
459
  },
460
  {
461
  "epoch": 14.0,
462
- "eval_exact_match": 15.6,
463
- "eval_f1": 22.729776889776886,
464
  "step": 3500
465
  },
466
  {
467
  "epoch": 14.4,
468
- "grad_norm": 0.7232977151870728,
469
  "learning_rate": 0.0003,
470
- "loss": 0.3713,
471
  "step": 3600
472
  },
473
  {
474
  "epoch": 14.8,
475
- "grad_norm": 0.6823480725288391,
476
  "learning_rate": 0.0003,
477
- "loss": 0.3834,
478
  "step": 3700
479
  },
480
  {
481
  "epoch": 15.0,
482
- "eval_accuracy": 0.4925079365079365,
483
- "eval_loss": 3.6552798748016357,
484
- "eval_runtime": 5.0332,
485
- "eval_samples_per_second": 99.341,
486
- "eval_steps_per_second": 12.517,
487
  "step": 3750
488
  },
489
  {
490
  "epoch": 15.0,
491
- "eval_exact_match": 13.8,
492
- "eval_f1": 20.994360084360082,
493
  "step": 3750
494
  },
495
  {
496
  "epoch": 15.2,
497
- "grad_norm": 0.5730023384094238,
498
  "learning_rate": 0.0003,
499
- "loss": 0.3795,
500
  "step": 3800
501
  },
502
  {
503
  "epoch": 15.6,
504
- "grad_norm": 0.943156898021698,
505
  "learning_rate": 0.0003,
506
- "loss": 0.3786,
507
  "step": 3900
508
  },
509
  {
510
  "epoch": 16.0,
511
- "grad_norm": 0.8560875654220581,
512
  "learning_rate": 0.0003,
513
- "loss": 0.3955,
514
  "step": 4000
515
  },
516
  {
517
  "epoch": 16.0,
518
- "eval_accuracy": 0.49196825396825394,
519
- "eval_loss": 3.6384711265563965,
520
- "eval_runtime": 5.73,
521
- "eval_samples_per_second": 87.26,
522
- "eval_steps_per_second": 10.995,
523
  "step": 4000
524
  },
525
  {
526
  "epoch": 16.0,
527
- "eval_exact_match": 14.4,
528
- "eval_f1": 21.078827926459503,
529
  "step": 4000
530
  },
531
  {
532
  "epoch": 16.4,
533
- "grad_norm": 0.5993083715438843,
534
  "learning_rate": 0.0003,
535
- "loss": 0.3651,
536
  "step": 4100
537
  },
538
  {
539
  "epoch": 16.8,
540
- "grad_norm": 0.8744499683380127,
541
  "learning_rate": 0.0003,
542
- "loss": 0.3843,
543
  "step": 4200
544
  },
545
  {
546
  "epoch": 17.0,
547
- "eval_accuracy": 0.49082539682539683,
548
- "eval_loss": 3.658356189727783,
549
- "eval_runtime": 5.2853,
550
- "eval_samples_per_second": 94.602,
551
- "eval_steps_per_second": 11.92,
552
  "step": 4250
553
  },
554
  {
555
  "epoch": 17.0,
556
- "eval_exact_match": 15.0,
557
- "eval_f1": 20.76140470640471,
558
  "step": 4250
559
  },
560
  {
561
  "epoch": 17.2,
562
- "grad_norm": 0.7331182360649109,
563
  "learning_rate": 0.0003,
564
- "loss": 0.3791,
565
  "step": 4300
566
  },
567
  {
568
  "epoch": 17.6,
569
- "grad_norm": 0.9493301510810852,
570
  "learning_rate": 0.0003,
571
- "loss": 0.3737,
572
  "step": 4400
573
  },
574
  {
575
  "epoch": 18.0,
576
- "grad_norm": 0.7808287143707275,
577
  "learning_rate": 0.0003,
578
- "loss": 0.3916,
579
  "step": 4500
580
  },
581
  {
582
  "epoch": 18.0,
583
- "eval_accuracy": 0.49247619047619046,
584
- "eval_loss": 3.659234046936035,
585
- "eval_runtime": 5.7356,
586
- "eval_samples_per_second": 87.176,
587
- "eval_steps_per_second": 10.984,
588
  "step": 4500
589
  },
590
  {
591
  "epoch": 18.0,
592
- "eval_exact_match": 14.8,
593
- "eval_f1": 21.322143412143415,
594
  "step": 4500
595
  },
596
  {
597
  "epoch": 18.4,
598
- "grad_norm": 0.7472721934318542,
599
  "learning_rate": 0.0003,
600
- "loss": 0.3692,
601
  "step": 4600
602
  },
603
  {
604
  "epoch": 18.8,
605
- "grad_norm": 0.5398291349411011,
606
  "learning_rate": 0.0003,
607
- "loss": 0.3825,
608
  "step": 4700
609
  },
610
  {
611
  "epoch": 19.0,
612
- "eval_accuracy": 0.4913015873015873,
613
- "eval_loss": 3.660428762435913,
614
- "eval_runtime": 5.6417,
615
- "eval_samples_per_second": 88.626,
616
- "eval_steps_per_second": 11.167,
617
  "step": 4750
618
  },
619
  {
620
  "epoch": 19.0,
621
- "eval_exact_match": 15.8,
622
- "eval_f1": 23.026702741702746,
623
  "step": 4750
624
  },
625
  {
626
  "epoch": 19.2,
627
- "grad_norm": 0.5717331171035767,
628
  "learning_rate": 0.0003,
629
- "loss": 0.3746,
630
  "step": 4800
631
  },
632
  {
633
  "epoch": 19.6,
634
- "grad_norm": 0.8405028581619263,
635
  "learning_rate": 0.0003,
636
- "loss": 0.3756,
637
  "step": 4900
638
  },
639
  {
640
  "epoch": 20.0,
641
- "grad_norm": 0.7853202223777771,
642
  "learning_rate": 0.0003,
643
- "loss": 0.387,
644
  "step": 5000
645
  },
646
  {
647
  "epoch": 20.0,
648
- "eval_accuracy": 0.49044444444444446,
649
- "eval_loss": 3.7399282455444336,
650
- "eval_runtime": 5.2464,
651
- "eval_samples_per_second": 95.303,
652
- "eval_steps_per_second": 12.008,
653
  "step": 5000
654
  },
655
  {
656
  "epoch": 20.0,
657
- "eval_exact_match": 13.8,
658
- "eval_f1": 21.514708624708625,
659
  "step": 5000
660
  },
661
  {
662
  "epoch": 20.4,
663
- "grad_norm": 0.8046431541442871,
664
  "learning_rate": 0.0003,
665
- "loss": 0.3643,
666
  "step": 5100
667
  },
668
  {
669
  "epoch": 20.8,
670
- "grad_norm": 0.6186045408248901,
671
  "learning_rate": 0.0003,
672
- "loss": 0.3832,
673
  "step": 5200
674
  },
675
  {
676
  "epoch": 21.0,
677
- "eval_accuracy": 0.4913968253968254,
678
- "eval_loss": 3.6563804149627686,
679
- "eval_runtime": 5.7336,
680
- "eval_samples_per_second": 87.206,
681
- "eval_steps_per_second": 10.988,
682
  "step": 5250
683
  },
684
  {
685
  "epoch": 21.0,
686
- "eval_exact_match": 13.8,
687
- "eval_f1": 20.836753246753247,
688
  "step": 5250
689
  },
690
  {
691
  "epoch": 21.2,
692
- "grad_norm": 0.4458799660205841,
693
  "learning_rate": 0.0003,
694
- "loss": 0.3733,
695
  "step": 5300
696
  },
697
  {
698
  "epoch": 21.6,
699
- "grad_norm": 0.6755269765853882,
700
  "learning_rate": 0.0003,
701
- "loss": 0.3675,
702
  "step": 5400
703
  },
704
  {
705
  "epoch": 22.0,
706
- "grad_norm": 0.8081554174423218,
707
  "learning_rate": 0.0003,
708
- "loss": 0.384,
709
  "step": 5500
710
  },
711
  {
712
  "epoch": 22.0,
713
- "eval_accuracy": 0.4899047619047619,
714
- "eval_loss": 3.686178207397461,
715
- "eval_runtime": 5.0623,
716
- "eval_samples_per_second": 98.77,
717
- "eval_steps_per_second": 12.445,
718
  "step": 5500
719
  },
720
  {
721
  "epoch": 22.0,
722
- "eval_exact_match": 15.6,
723
- "eval_f1": 22.18729215229216,
724
  "step": 5500
725
  },
726
  {
727
  "epoch": 22.4,
728
- "grad_norm": 0.5928174257278442,
729
  "learning_rate": 0.0003,
730
- "loss": 0.3561,
731
  "step": 5600
732
  },
733
  {
734
  "epoch": 22.8,
735
- "grad_norm": 0.5147334933280945,
736
  "learning_rate": 0.0003,
737
- "loss": 0.3753,
738
  "step": 5700
739
  },
740
  {
741
  "epoch": 23.0,
742
- "eval_accuracy": 0.49063492063492065,
743
- "eval_loss": 3.6690633296966553,
744
- "eval_runtime": 5.6639,
745
- "eval_samples_per_second": 88.278,
746
- "eval_steps_per_second": 11.123,
747
  "step": 5750
748
  },
749
  {
750
  "epoch": 23.0,
751
- "eval_exact_match": 14.2,
752
- "eval_f1": 21.33359729642339,
753
  "step": 5750
754
  },
755
  {
756
  "epoch": 23.2,
757
- "grad_norm": 0.8577277660369873,
758
  "learning_rate": 0.0003,
759
- "loss": 0.3655,
760
  "step": 5800
761
  },
762
  {
763
  "epoch": 23.6,
764
- "grad_norm": 0.6065458655357361,
765
  "learning_rate": 0.0003,
766
- "loss": 0.3609,
767
  "step": 5900
768
  },
769
  {
770
  "epoch": 24.0,
771
- "grad_norm": 1.1300039291381836,
772
  "learning_rate": 0.0003,
773
- "loss": 0.3816,
774
  "step": 6000
775
  },
776
  {
777
  "epoch": 24.0,
778
- "eval_accuracy": 0.4908571428571429,
779
- "eval_loss": 3.7181262969970703,
780
- "eval_runtime": 5.7672,
781
- "eval_samples_per_second": 86.698,
782
- "eval_steps_per_second": 10.924,
783
  "step": 6000
784
  },
785
  {
786
  "epoch": 24.0,
787
- "eval_exact_match": 13.6,
788
- "eval_f1": 20.043032849503437,
789
  "step": 6000
790
  },
791
  {
792
  "epoch": 24.4,
793
- "grad_norm": 0.722024142742157,
794
  "learning_rate": 0.0003,
795
- "loss": 0.3555,
796
  "step": 6100
797
  },
798
  {
799
  "epoch": 24.8,
800
- "grad_norm": 0.773781955242157,
801
  "learning_rate": 0.0003,
802
- "loss": 0.3711,
803
  "step": 6200
804
  },
805
  {
806
  "epoch": 25.0,
807
- "eval_accuracy": 0.4895555555555556,
808
- "eval_loss": 3.715923309326172,
809
- "eval_runtime": 5.35,
810
- "eval_samples_per_second": 93.457,
811
- "eval_steps_per_second": 11.776,
812
  "step": 6250
813
  },
814
  {
815
  "epoch": 25.0,
816
- "eval_exact_match": 14.2,
817
- "eval_f1": 21.817198315959928,
818
  "step": 6250
819
  },
820
  {
821
  "epoch": 25.2,
822
- "grad_norm": 0.5301027894020081,
823
  "learning_rate": 0.0003,
824
- "loss": 0.365,
825
  "step": 6300
826
  },
827
  {
828
  "epoch": 25.6,
829
- "grad_norm": 0.6550167202949524,
830
  "learning_rate": 0.0003,
831
- "loss": 0.3632,
832
  "step": 6400
833
  },
834
  {
835
  "epoch": 26.0,
836
- "grad_norm": 0.9597300291061401,
837
  "learning_rate": 0.0003,
838
- "loss": 0.3758,
839
  "step": 6500
840
  },
841
  {
842
  "epoch": 26.0,
843
- "eval_accuracy": 0.49073015873015874,
844
- "eval_loss": 3.7265737056732178,
845
- "eval_runtime": 5.7349,
846
- "eval_samples_per_second": 87.185,
847
- "eval_steps_per_second": 10.985,
848
  "step": 6500
849
  },
850
  {
851
  "epoch": 26.0,
852
- "eval_exact_match": 13.0,
853
- "eval_f1": 19.8548159030512,
854
  "step": 6500
855
  },
856
  {
857
  "epoch": 26.4,
858
- "grad_norm": 0.5059230327606201,
859
  "learning_rate": 0.0003,
860
- "loss": 0.3536,
861
  "step": 6600
862
  },
863
  {
864
  "epoch": 26.8,
865
- "grad_norm": 0.5931850671768188,
866
  "learning_rate": 0.0003,
867
- "loss": 0.372,
868
  "step": 6700
869
  },
870
  {
871
  "epoch": 27.0,
872
- "eval_accuracy": 0.4915873015873016,
873
- "eval_loss": 3.7632603645324707,
874
- "eval_runtime": 4.9892,
875
- "eval_samples_per_second": 100.216,
876
- "eval_steps_per_second": 12.627,
877
  "step": 6750
878
  },
879
  {
880
  "epoch": 27.0,
881
- "eval_exact_match": 14.6,
882
- "eval_f1": 21.288471691706988,
883
  "step": 6750
884
  },
885
  {
886
  "epoch": 27.2,
887
- "grad_norm": 0.8211339712142944,
888
  "learning_rate": 0.0003,
889
- "loss": 0.3667,
890
  "step": 6800
891
  },
892
  {
893
  "epoch": 27.6,
894
- "grad_norm": 0.7241945862770081,
895
  "learning_rate": 0.0003,
896
- "loss": 0.3655,
897
  "step": 6900
898
  },
899
  {
900
  "epoch": 28.0,
901
- "grad_norm": 0.8376013040542603,
902
  "learning_rate": 0.0003,
903
- "loss": 0.3813,
904
  "step": 7000
905
  },
906
  {
907
  "epoch": 28.0,
908
- "eval_accuracy": 0.48926984126984124,
909
- "eval_loss": 3.751114845275879,
910
- "eval_runtime": 5.7876,
911
- "eval_samples_per_second": 86.391,
912
- "eval_steps_per_second": 10.885,
913
  "step": 7000
914
  },
915
  {
916
  "epoch": 28.0,
917
- "eval_exact_match": 14.4,
918
- "eval_f1": 21.447649735885033,
919
  "step": 7000
920
  },
921
  {
922
  "epoch": 28.4,
923
- "grad_norm": 1.0523216724395752,
924
  "learning_rate": 0.0003,
925
- "loss": 0.3588,
926
  "step": 7100
927
  },
928
  {
929
  "epoch": 28.8,
930
- "grad_norm": 1.0963407754898071,
931
  "learning_rate": 0.0003,
932
- "loss": 0.3699,
933
  "step": 7200
934
  },
935
  {
936
  "epoch": 29.0,
937
- "eval_accuracy": 0.4903174603174603,
938
- "eval_loss": 3.7290866374969482,
939
- "eval_runtime": 5.3378,
940
- "eval_samples_per_second": 93.671,
941
- "eval_steps_per_second": 11.803,
942
  "step": 7250
943
  },
944
  {
945
  "epoch": 29.0,
946
- "eval_exact_match": 14.4,
947
- "eval_f1": 21.026401539636833,
948
  "step": 7250
949
  },
950
  {
951
  "epoch": 29.2,
952
- "grad_norm": 0.6481629610061646,
953
  "learning_rate": 0.0003,
954
- "loss": 0.3655,
955
  "step": 7300
956
  },
957
  {
958
  "epoch": 29.6,
959
- "grad_norm": 0.6700736284255981,
960
  "learning_rate": 0.0003,
961
- "loss": 0.3657,
962
  "step": 7400
963
  },
964
  {
965
  "epoch": 30.0,
966
- "grad_norm": 0.6046570539474487,
967
  "learning_rate": 0.0003,
968
- "loss": 0.3772,
969
  "step": 7500
970
  },
971
  {
972
  "epoch": 30.0,
973
- "eval_accuracy": 0.48857142857142855,
974
- "eval_loss": 3.7827305793762207,
975
- "eval_runtime": 4.9957,
976
- "eval_samples_per_second": 100.086,
977
- "eval_steps_per_second": 12.611,
978
  "step": 7500
979
  },
980
  {
981
  "epoch": 30.0,
982
- "eval_exact_match": 14.2,
983
- "eval_f1": 20.93110366757426,
984
  "step": 7500
985
  },
986
  {
987
  "epoch": 30.4,
988
- "grad_norm": 0.5793018937110901,
989
  "learning_rate": 0.0003,
990
- "loss": 0.3515,
991
  "step": 7600
992
  },
993
  {
994
  "epoch": 30.8,
995
- "grad_norm": 0.733734667301178,
996
  "learning_rate": 0.0003,
997
- "loss": 0.3673,
998
  "step": 7700
999
  },
1000
  {
1001
  "epoch": 31.0,
1002
- "eval_accuracy": 0.48917460317460315,
1003
- "eval_loss": 3.8032257556915283,
1004
- "eval_runtime": 5.7717,
1005
- "eval_samples_per_second": 86.629,
1006
- "eval_steps_per_second": 10.915,
1007
  "step": 7750
1008
  },
1009
  {
1010
  "epoch": 31.0,
1011
- "eval_exact_match": 13.2,
1012
- "eval_f1": 19.896597454832754,
1013
  "step": 7750
1014
  },
1015
  {
1016
  "epoch": 31.2,
1017
- "grad_norm": 0.8759809732437134,
1018
  "learning_rate": 0.0003,
1019
- "loss": 0.3642,
1020
  "step": 7800
1021
  },
1022
  {
1023
  "epoch": 31.6,
1024
- "grad_norm": 0.7398084998130798,
1025
  "learning_rate": 0.0003,
1026
- "loss": 0.3647,
1027
  "step": 7900
1028
  },
1029
  {
1030
  "epoch": 32.0,
1031
- "grad_norm": 0.9378136992454529,
1032
  "learning_rate": 0.0003,
1033
- "loss": 0.3708,
1034
  "step": 8000
1035
  },
1036
  {
1037
  "epoch": 32.0,
1038
- "eval_accuracy": 0.4894603174603175,
1039
- "eval_loss": 3.830268144607544,
1040
- "eval_runtime": 5.6155,
1041
- "eval_samples_per_second": 89.039,
1042
- "eval_steps_per_second": 11.219,
1043
  "step": 8000
1044
  },
1045
  {
1046
  "epoch": 32.0,
1047
- "eval_exact_match": 14.0,
1048
- "eval_f1": 20.862043120278422,
1049
  "step": 8000
1050
  },
1051
  {
1052
  "epoch": 32.4,
1053
- "grad_norm": 0.4455881118774414,
1054
  "learning_rate": 0.0003,
1055
- "loss": 0.346,
1056
  "step": 8100
1057
  },
1058
  {
1059
  "epoch": 32.8,
1060
- "grad_norm": 0.6031106114387512,
1061
  "learning_rate": 0.0003,
1062
- "loss": 0.3632,
1063
  "step": 8200
1064
  },
1065
  {
1066
  "epoch": 33.0,
1067
- "eval_accuracy": 0.4886984126984127,
1068
- "eval_loss": 3.821800947189331,
1069
- "eval_runtime": 5.21,
1070
- "eval_samples_per_second": 95.968,
1071
- "eval_steps_per_second": 12.092,
1072
  "step": 8250
1073
  },
1074
  {
1075
  "epoch": 33.0,
1076
- "eval_exact_match": 14.0,
1077
- "eval_f1": 20.771498109733407,
1078
  "step": 8250
1079
  },
1080
  {
1081
  "epoch": 33.2,
1082
- "grad_norm": 0.6439555287361145,
1083
  "learning_rate": 0.0003,
1084
- "loss": 0.3553,
1085
  "step": 8300
1086
  },
1087
  {
1088
  "epoch": 33.6,
1089
- "grad_norm": 0.6980018019676208,
1090
  "learning_rate": 0.0003,
1091
- "loss": 0.3533,
1092
  "step": 8400
1093
  },
1094
  {
1095
  "epoch": 34.0,
1096
- "grad_norm": 0.6426189541816711,
1097
  "learning_rate": 0.0003,
1098
- "loss": 0.3692,
1099
  "step": 8500
1100
  },
1101
  {
1102
  "epoch": 34.0,
1103
- "eval_accuracy": 0.49,
1104
- "eval_loss": 3.7488105297088623,
1105
- "eval_runtime": 4.9468,
1106
- "eval_samples_per_second": 101.075,
1107
- "eval_steps_per_second": 12.735,
1108
  "step": 8500
1109
  },
1110
  {
1111
  "epoch": 34.0,
1112
- "eval_exact_match": 14.8,
1113
- "eval_f1": 21.277384282384293,
1114
  "step": 8500
1115
  }
1116
  ],
1117
  "logging_steps": 100,
1118
  "max_steps": 12500,
1119
- "num_input_tokens_seen": 0,
1120
  "num_train_epochs": 50,
1121
  "save_steps": 500,
1122
- "stateful_callbacks": {
1123
- "TrainerControl": {
1124
- "args": {
1125
- "should_epoch_stop": false,
1126
- "should_evaluate": false,
1127
- "should_log": false,
1128
- "should_save": true,
1129
- "should_training_stop": false
1130
- },
1131
- "attributes": {}
1132
- }
1133
- },
1134
- "total_flos": 2.199015614536745e+17,
1135
- "train_batch_size": 1,
1136
  "trial_name": null,
1137
  "trial_params": null
1138
  }
 
10
  "log_history": [
11
  {
12
  "epoch": 0.4,
 
13
  "learning_rate": 0.0003,
14
+ "loss": 1.9296,
15
  "step": 100
16
  },
17
  {
18
  "epoch": 0.8,
 
19
  "learning_rate": 0.0003,
20
+ "loss": 1.7554,
21
  "step": 200
22
  },
23
  {
24
  "epoch": 1.0,
25
+ "eval_accuracy": 0.6093164556962025,
26
+ "eval_loss": 1.7939746379852295,
27
+ "eval_runtime": 5.0177,
28
+ "eval_samples_per_second": 99.648,
29
+ "eval_steps_per_second": 12.556,
30
  "step": 250
31
  },
32
  {
33
  "epoch": 1.0,
34
+ "eval_exact_match": 26.0,
35
+ "eval_f1": 36.71136363636365,
36
  "step": 250
37
  },
38
  {
39
  "epoch": 1.2,
 
40
  "learning_rate": 0.0003,
41
+ "loss": 1.6394,
42
  "step": 300
43
  },
44
  {
45
  "epoch": 1.6,
 
46
  "learning_rate": 0.0003,
47
+ "loss": 1.5315,
48
  "step": 400
49
  },
50
  {
51
  "epoch": 2.0,
 
52
  "learning_rate": 0.0003,
53
+ "loss": 1.5248,
54
  "step": 500
55
  },
56
  {
57
  "epoch": 2.0,
58
+ "eval_accuracy": 0.6085063291139241,
59
+ "eval_loss": 1.8273799419403076,
60
+ "eval_runtime": 4.4023,
61
+ "eval_samples_per_second": 113.576,
62
+ "eval_steps_per_second": 14.311,
63
  "step": 500
64
  },
65
  {
66
  "epoch": 2.0,
67
+ "eval_exact_match": 24.4,
68
+ "eval_f1": 33.57723554223556,
69
  "step": 500
70
  },
71
  {
72
  "epoch": 2.4,
 
73
  "learning_rate": 0.0003,
74
+ "loss": 1.1692,
75
  "step": 600
76
  },
77
  {
78
  "epoch": 2.8,
 
79
  "learning_rate": 0.0003,
80
+ "loss": 1.2054,
81
  "step": 700
82
  },
83
  {
84
  "epoch": 3.0,
85
+ "eval_accuracy": 0.6027088607594937,
86
+ "eval_loss": 1.9717934131622314,
87
+ "eval_runtime": 5.1273,
88
+ "eval_samples_per_second": 97.518,
89
+ "eval_steps_per_second": 12.287,
90
  "step": 750
91
  },
92
  {
93
  "epoch": 3.0,
94
+ "eval_exact_match": 20.8,
95
+ "eval_f1": 30.584243697478996,
96
  "step": 750
97
  },
98
  {
99
  "epoch": 3.2,
 
100
  "learning_rate": 0.0003,
101
+ "loss": 1.0252,
102
  "step": 800
103
  },
104
  {
105
  "epoch": 3.6,
 
106
  "learning_rate": 0.0003,
107
+ "loss": 0.8608,
108
  "step": 900
109
  },
110
  {
111
  "epoch": 4.0,
 
112
  "learning_rate": 0.0003,
113
+ "loss": 0.8989,
114
  "step": 1000
115
  },
116
  {
117
  "epoch": 4.0,
118
+ "eval_accuracy": 0.5986835443037974,
119
+ "eval_loss": 2.151914596557617,
120
+ "eval_runtime": 5.3325,
121
+ "eval_samples_per_second": 93.765,
122
+ "eval_steps_per_second": 11.814,
123
  "step": 1000
124
  },
125
  {
126
  "epoch": 4.0,
127
+ "eval_exact_match": 21.0,
128
+ "eval_f1": 30.363229548229558,
129
  "step": 1000
130
  },
131
  {
132
  "epoch": 4.4,
 
133
  "learning_rate": 0.0003,
134
+ "loss": 0.5842,
135
  "step": 1100
136
  },
137
  {
138
  "epoch": 4.8,
 
139
  "learning_rate": 0.0003,
140
+ "loss": 0.6306,
141
  "step": 1200
142
  },
143
  {
144
  "epoch": 5.0,
145
+ "eval_accuracy": 0.5960506329113924,
146
+ "eval_loss": 2.329284191131592,
147
+ "eval_runtime": 4.5585,
148
+ "eval_samples_per_second": 109.686,
149
+ "eval_steps_per_second": 13.82,
150
  "step": 1250
151
  },
152
  {
153
  "epoch": 5.0,
154
+ "eval_exact_match": 20.4,
155
+ "eval_f1": 30.02542624042625,
156
  "step": 1250
157
  },
158
  {
159
  "epoch": 5.2,
 
160
  "learning_rate": 0.0003,
161
+ "loss": 0.5368,
162
  "step": 1300
163
  },
164
  {
165
  "epoch": 5.6,
 
166
  "learning_rate": 0.0003,
167
+ "loss": 0.4495,
168
  "step": 1400
169
  },
170
  {
171
  "epoch": 6.0,
 
172
  "learning_rate": 0.0003,
173
+ "loss": 0.4712,
174
  "step": 1500
175
  },
176
  {
177
  "epoch": 6.0,
178
+ "eval_accuracy": 0.5935696202531645,
179
+ "eval_loss": 2.5598793029785156,
180
+ "eval_runtime": 4.4094,
181
+ "eval_samples_per_second": 113.395,
182
+ "eval_steps_per_second": 14.288,
183
  "step": 1500
184
  },
185
  {
186
  "epoch": 6.0,
187
+ "eval_exact_match": 20.6,
188
+ "eval_f1": 30.108149073149082,
189
  "step": 1500
190
  },
191
  {
192
  "epoch": 6.4,
 
193
  "learning_rate": 0.0003,
194
+ "loss": 0.3625,
195
  "step": 1600
196
  },
197
  {
198
  "epoch": 6.8,
 
199
  "learning_rate": 0.0003,
200
+ "loss": 0.3797,
201
  "step": 1700
202
  },
203
  {
204
  "epoch": 7.0,
205
+ "eval_accuracy": 0.5935696202531645,
206
+ "eval_loss": 2.732860803604126,
207
+ "eval_runtime": 5.2701,
208
+ "eval_samples_per_second": 94.876,
209
+ "eval_steps_per_second": 11.954,
210
  "step": 1750
211
  },
212
  {
213
  "epoch": 7.0,
214
+ "eval_exact_match": 19.6,
215
+ "eval_f1": 29.002709512709526,
216
  "step": 1750
217
  },
218
  {
219
  "epoch": 7.2,
 
220
  "learning_rate": 0.0003,
221
+ "loss": 0.3582,
222
  "step": 1800
223
  },
224
  {
225
  "epoch": 7.6,
 
226
  "learning_rate": 0.0003,
227
+ "loss": 0.3399,
228
  "step": 1900
229
  },
230
  {
231
  "epoch": 8.0,
 
232
  "learning_rate": 0.0003,
233
+ "loss": 0.3527,
234
  "step": 2000
235
  },
236
  {
237
  "epoch": 8.0,
238
+ "eval_accuracy": 0.5912911392405064,
239
+ "eval_loss": 2.8185083866119385,
240
+ "eval_runtime": 4.4105,
241
+ "eval_samples_per_second": 113.366,
242
+ "eval_steps_per_second": 14.284,
243
  "step": 2000
244
  },
245
  {
246
  "epoch": 8.0,
247
+ "eval_exact_match": 20.8,
248
+ "eval_f1": 29.191525696525694,
249
  "step": 2000
250
  },
251
  {
252
  "epoch": 8.4,
 
253
  "learning_rate": 0.0003,
254
+ "loss": 0.3202,
255
  "step": 2100
256
  },
257
  {
258
  "epoch": 8.8,
 
259
  "learning_rate": 0.0003,
260
+ "loss": 0.3314,
261
  "step": 2200
262
  },
263
  {
264
  "epoch": 9.0,
265
+ "eval_accuracy": 0.592,
266
+ "eval_loss": 2.824962854385376,
267
+ "eval_runtime": 4.5596,
268
+ "eval_samples_per_second": 109.66,
269
+ "eval_steps_per_second": 13.817,
270
  "step": 2250
271
  },
272
  {
273
  "epoch": 9.0,
274
+ "eval_exact_match": 21.0,
275
+ "eval_f1": 29.434571539571547,
276
  "step": 2250
277
  },
278
  {
279
  "epoch": 9.2,
 
280
  "learning_rate": 0.0003,
281
+ "loss": 0.3174,
282
  "step": 2300
283
  },
284
  {
285
  "epoch": 9.6,
 
286
  "learning_rate": 0.0003,
287
+ "loss": 0.3157,
288
  "step": 2400
289
  },
290
  {
291
  "epoch": 10.0,
 
292
  "learning_rate": 0.0003,
293
+ "loss": 0.3265,
294
  "step": 2500
295
  },
296
  {
297
  "epoch": 10.0,
298
+ "eval_accuracy": 0.5911392405063292,
299
+ "eval_loss": 2.9242382049560547,
300
+ "eval_runtime": 4.7857,
301
+ "eval_samples_per_second": 104.477,
302
+ "eval_steps_per_second": 13.164,
303
  "step": 2500
304
  },
305
  {
306
  "epoch": 10.0,
307
+ "eval_exact_match": 20.8,
308
+ "eval_f1": 29.524657009657016,
309
  "step": 2500
310
  },
311
  {
312
  "epoch": 10.4,
 
313
  "learning_rate": 0.0003,
314
+ "loss": 0.2989,
315
  "step": 2600
316
  },
317
  {
318
  "epoch": 10.8,
 
319
  "learning_rate": 0.0003,
320
+ "loss": 0.3148,
321
  "step": 2700
322
  },
323
  {
324
  "epoch": 11.0,
325
+ "eval_accuracy": 0.591240506329114,
326
+ "eval_loss": 3.0012593269348145,
327
+ "eval_runtime": 4.5582,
328
+ "eval_samples_per_second": 109.691,
329
+ "eval_steps_per_second": 13.821,
330
  "step": 2750
331
  },
332
  {
333
  "epoch": 11.0,
334
+ "eval_exact_match": 20.0,
335
+ "eval_f1": 28.128880727116037,
336
  "step": 2750
337
  },
338
  {
339
  "epoch": 11.2,
 
340
  "learning_rate": 0.0003,
341
+ "loss": 0.3028,
342
  "step": 2800
343
  },
344
  {
345
  "epoch": 11.6,
 
346
  "learning_rate": 0.0003,
347
+ "loss": 0.3047,
348
  "step": 2900
349
  },
350
  {
351
  "epoch": 12.0,
 
352
  "learning_rate": 0.0003,
353
+ "loss": 0.3184,
354
  "step": 3000
355
  },
356
  {
357
  "epoch": 12.0,
358
+ "eval_accuracy": 0.590632911392405,
359
+ "eval_loss": 2.931525707244873,
360
+ "eval_runtime": 4.4989,
361
+ "eval_samples_per_second": 111.138,
362
+ "eval_steps_per_second": 14.003,
363
  "step": 3000
364
  },
365
  {
366
  "epoch": 12.0,
367
+ "eval_exact_match": 21.2,
368
+ "eval_f1": 30.12248362748364,
369
  "step": 3000
370
  },
371
  {
372
  "epoch": 12.4,
 
373
  "learning_rate": 0.0003,
374
+ "loss": 0.3,
375
  "step": 3100
376
  },
377
  {
378
  "epoch": 12.8,
 
379
  "learning_rate": 0.0003,
380
+ "loss": 0.3101,
381
  "step": 3200
382
  },
383
  {
384
  "epoch": 13.0,
385
+ "eval_accuracy": 0.5896962025316456,
386
+ "eval_loss": 2.9116382598876953,
387
+ "eval_runtime": 4.6155,
388
+ "eval_samples_per_second": 108.331,
389
+ "eval_steps_per_second": 13.65,
390
  "step": 3250
391
  },
392
  {
393
  "epoch": 13.0,
394
+ "eval_exact_match": 19.2,
395
+ "eval_f1": 27.90991785991786,
396
  "step": 3250
397
  },
398
  {
399
  "epoch": 13.2,
 
400
  "learning_rate": 0.0003,
401
+ "loss": 0.3063,
402
  "step": 3300
403
  },
404
  {
405
  "epoch": 13.6,
 
406
  "learning_rate": 0.0003,
407
+ "loss": 0.3041,
408
  "step": 3400
409
  },
410
  {
411
  "epoch": 14.0,
 
412
  "learning_rate": 0.0003,
413
+ "loss": 0.3164,
414
  "step": 3500
415
  },
416
  {
417
  "epoch": 14.0,
418
+ "eval_accuracy": 0.5902025316455696,
419
+ "eval_loss": 2.920793056488037,
420
+ "eval_runtime": 4.4066,
421
+ "eval_samples_per_second": 113.467,
422
+ "eval_steps_per_second": 14.297,
423
  "step": 3500
424
  },
425
  {
426
  "epoch": 14.0,
427
+ "eval_exact_match": 18.0,
428
+ "eval_f1": 27.415170385170395,
429
  "step": 3500
430
  },
431
  {
432
  "epoch": 14.4,
 
433
  "learning_rate": 0.0003,
434
+ "loss": 0.2957,
435
  "step": 3600
436
  },
437
  {
438
  "epoch": 14.8,
 
439
  "learning_rate": 0.0003,
440
+ "loss": 0.3074,
441
  "step": 3700
442
  },
443
  {
444
  "epoch": 15.0,
445
+ "eval_accuracy": 0.5908607594936709,
446
+ "eval_loss": 2.9385440349578857,
447
+ "eval_runtime": 4.8175,
448
+ "eval_samples_per_second": 103.788,
449
+ "eval_steps_per_second": 13.077,
450
  "step": 3750
451
  },
452
  {
453
  "epoch": 15.0,
454
+ "eval_exact_match": 21.2,
455
+ "eval_f1": 29.687630702630706,
456
  "step": 3750
457
  },
458
  {
459
  "epoch": 15.2,
 
460
  "learning_rate": 0.0003,
461
+ "loss": 0.3013,
462
  "step": 3800
463
  },
464
  {
465
  "epoch": 15.6,
 
466
  "learning_rate": 0.0003,
467
+ "loss": 0.3002,
468
  "step": 3900
469
  },
470
  {
471
  "epoch": 16.0,
 
472
  "learning_rate": 0.0003,
473
+ "loss": 0.3107,
474
  "step": 4000
475
  },
476
  {
477
  "epoch": 16.0,
478
+ "eval_accuracy": 0.5891645569620253,
479
+ "eval_loss": 2.9519243240356445,
480
+ "eval_runtime": 4.5619,
481
+ "eval_samples_per_second": 109.603,
482
+ "eval_steps_per_second": 13.81,
483
  "step": 4000
484
  },
485
  {
486
  "epoch": 16.0,
487
+ "eval_exact_match": 19.8,
488
+ "eval_f1": 28.670117105117104,
489
  "step": 4000
490
  },
491
  {
492
  "epoch": 16.4,
 
493
  "learning_rate": 0.0003,
494
+ "loss": 0.2892,
495
  "step": 4100
496
  },
497
  {
498
  "epoch": 16.8,
 
499
  "learning_rate": 0.0003,
500
+ "loss": 0.3054,
501
  "step": 4200
502
  },
503
  {
504
  "epoch": 17.0,
505
+ "eval_accuracy": 0.5898227848101266,
506
+ "eval_loss": 3.010847568511963,
507
+ "eval_runtime": 4.718,
508
+ "eval_samples_per_second": 105.977,
509
+ "eval_steps_per_second": 13.353,
510
  "step": 4250
511
  },
512
  {
513
  "epoch": 17.0,
514
+ "eval_exact_match": 19.6,
515
+ "eval_f1": 27.744197632432932,
516
  "step": 4250
517
  },
518
  {
519
  "epoch": 17.2,
 
520
  "learning_rate": 0.0003,
521
+ "loss": 0.2959,
522
  "step": 4300
523
  },
524
  {
525
  "epoch": 17.6,
 
526
  "learning_rate": 0.0003,
527
+ "loss": 0.297,
528
  "step": 4400
529
  },
530
  {
531
  "epoch": 18.0,
 
532
  "learning_rate": 0.0003,
533
+ "loss": 0.309,
534
  "step": 4500
535
  },
536
  {
537
  "epoch": 18.0,
538
+ "eval_accuracy": 0.5903544303797469,
539
+ "eval_loss": 3.003683567047119,
540
+ "eval_runtime": 4.4085,
541
+ "eval_samples_per_second": 113.418,
542
+ "eval_steps_per_second": 14.291,
543
  "step": 4500
544
  },
545
  {
546
  "epoch": 18.0,
547
+ "eval_exact_match": 20.6,
548
+ "eval_f1": 28.64154694978225,
549
  "step": 4500
550
  },
551
  {
552
  "epoch": 18.4,
 
553
  "learning_rate": 0.0003,
554
+ "loss": 0.2883,
555
  "step": 4600
556
  },
557
  {
558
  "epoch": 18.8,
 
559
  "learning_rate": 0.0003,
560
+ "loss": 0.3005,
561
  "step": 4700
562
  },
563
  {
564
  "epoch": 19.0,
565
+ "eval_accuracy": 0.5898481012658228,
566
+ "eval_loss": 3.0279438495635986,
567
+ "eval_runtime": 4.7099,
568
+ "eval_samples_per_second": 106.16,
569
+ "eval_steps_per_second": 13.376,
570
  "step": 4750
571
  },
572
  {
573
  "epoch": 19.0,
574
+ "eval_exact_match": 21.0,
575
+ "eval_f1": 28.707790706026003,
576
  "step": 4750
577
  },
578
  {
579
  "epoch": 19.2,
 
580
  "learning_rate": 0.0003,
581
+ "loss": 0.2959,
582
  "step": 4800
583
  },
584
  {
585
  "epoch": 19.6,
 
586
  "learning_rate": 0.0003,
587
+ "loss": 0.2911,
588
  "step": 4900
589
  },
590
  {
591
  "epoch": 20.0,
 
592
  "learning_rate": 0.0003,
593
+ "loss": 0.3127,
594
  "step": 5000
595
  },
596
  {
597
  "epoch": 20.0,
598
+ "eval_accuracy": 0.5883291139240506,
599
+ "eval_loss": 2.9650285243988037,
600
+ "eval_runtime": 4.711,
601
+ "eval_samples_per_second": 106.135,
602
+ "eval_steps_per_second": 13.373,
603
  "step": 5000
604
  },
605
  {
606
  "epoch": 20.0,
607
+ "eval_exact_match": 11.4,
608
+ "eval_f1": 20.32415197623728,
609
  "step": 5000
610
  },
611
  {
612
  "epoch": 20.4,
 
613
  "learning_rate": 0.0003,
614
+ "loss": 0.2868,
615
  "step": 5100
616
  },
617
  {
618
  "epoch": 20.8,
 
619
  "learning_rate": 0.0003,
620
+ "loss": 0.2989,
621
  "step": 5200
622
  },
623
  {
624
  "epoch": 21.0,
625
+ "eval_accuracy": 0.5895696202531645,
626
+ "eval_loss": 3.013432502746582,
627
+ "eval_runtime": 4.8101,
628
+ "eval_samples_per_second": 103.947,
629
+ "eval_steps_per_second": 13.097,
630
  "step": 5250
631
  },
632
  {
633
  "epoch": 21.0,
634
+ "eval_exact_match": 10.8,
635
+ "eval_f1": 17.55567536105352,
636
  "step": 5250
637
  },
638
  {
639
  "epoch": 21.2,
 
640
  "learning_rate": 0.0003,
641
+ "loss": 0.2948,
642
  "step": 5300
643
  },
644
  {
645
  "epoch": 21.6,
 
646
  "learning_rate": 0.0003,
647
+ "loss": 0.2921,
648
  "step": 5400
649
  },
650
  {
651
  "epoch": 22.0,
 
652
  "learning_rate": 0.0003,
653
+ "loss": 0.3072,
654
  "step": 5500
655
  },
656
  {
657
  "epoch": 22.0,
658
+ "eval_accuracy": 0.5885822784810126,
659
+ "eval_loss": 3.0298521518707275,
660
+ "eval_runtime": 4.7113,
661
+ "eval_samples_per_second": 106.128,
662
+ "eval_steps_per_second": 13.372,
663
  "step": 5500
664
  },
665
  {
666
  "epoch": 22.0,
667
+ "eval_exact_match": 0.0,
668
+ "eval_f1": 0.7666666666666667,
669
  "step": 5500
670
  },
671
  {
672
  "epoch": 22.4,
 
673
  "learning_rate": 0.0003,
674
+ "loss": 0.2914,
675
  "step": 5600
676
  },
677
  {
678
  "epoch": 22.8,
 
679
  "learning_rate": 0.0003,
680
+ "loss": 0.2982,
681
  "step": 5700
682
  },
683
  {
684
  "epoch": 23.0,
685
+ "eval_accuracy": 0.5870126582278481,
686
+ "eval_loss": 3.0277092456817627,
687
+ "eval_runtime": 4.4065,
688
+ "eval_samples_per_second": 113.468,
689
+ "eval_steps_per_second": 14.297,
690
  "step": 5750
691
  },
692
  {
693
  "epoch": 23.0,
694
+ "eval_exact_match": 7.6,
695
+ "eval_f1": 9.586507936507935,
696
  "step": 5750
697
  },
698
  {
699
  "epoch": 23.2,
 
700
  "learning_rate": 0.0003,
701
+ "loss": 0.3014,
702
  "step": 5800
703
  },
704
  {
705
  "epoch": 23.6,
 
706
  "learning_rate": 0.0003,
707
+ "loss": 0.2907,
708
  "step": 5900
709
  },
710
  {
711
  "epoch": 24.0,
 
712
  "learning_rate": 0.0003,
713
+ "loss": 0.3029,
714
  "step": 6000
715
  },
716
  {
717
  "epoch": 24.0,
718
+ "eval_accuracy": 0.5877215189873418,
719
+ "eval_loss": 3.02402925491333,
720
+ "eval_runtime": 4.7089,
721
+ "eval_samples_per_second": 106.182,
722
+ "eval_steps_per_second": 13.379,
723
  "step": 6000
724
  },
725
  {
726
  "epoch": 24.0,
727
+ "eval_exact_match": 4.2,
728
+ "eval_f1": 7.583745822888591,
729
  "step": 6000
730
  },
731
  {
732
  "epoch": 24.4,
 
733
  "learning_rate": 0.0003,
734
+ "loss": 0.2853,
735
  "step": 6100
736
  },
737
  {
738
  "epoch": 24.8,
 
739
  "learning_rate": 0.0003,
740
+ "loss": 0.2989,
741
  "step": 6200
742
  },
743
  {
744
  "epoch": 25.0,
745
+ "eval_accuracy": 0.5874430379746836,
746
+ "eval_loss": 3.0006086826324463,
747
+ "eval_runtime": 4.5617,
748
+ "eval_samples_per_second": 109.609,
749
+ "eval_steps_per_second": 13.811,
750
  "step": 6250
751
  },
752
  {
753
  "epoch": 25.0,
754
+ "eval_exact_match": 0.0,
755
+ "eval_f1": 0.5121500721500722,
756
  "step": 6250
757
  },
758
  {
759
  "epoch": 25.2,
 
760
  "learning_rate": 0.0003,
761
+ "loss": 0.2904,
762
  "step": 6300
763
  },
764
  {
765
  "epoch": 25.6,
 
766
  "learning_rate": 0.0003,
767
+ "loss": 0.29,
768
  "step": 6400
769
  },
770
  {
771
  "epoch": 26.0,
 
772
  "learning_rate": 0.0003,
773
+ "loss": 0.3009,
774
  "step": 6500
775
  },
776
  {
777
  "epoch": 26.0,
778
+ "eval_accuracy": 0.5873417721518988,
779
+ "eval_loss": 3.058345317840576,
780
+ "eval_runtime": 4.708,
781
+ "eval_samples_per_second": 106.202,
782
+ "eval_steps_per_second": 13.381,
783
  "step": 6500
784
  },
785
  {
786
  "epoch": 26.0,
787
+ "eval_exact_match": 2.2,
788
+ "eval_f1": 4.8883016983016985,
789
  "step": 6500
790
  },
791
  {
792
  "epoch": 26.4,
 
793
  "learning_rate": 0.0003,
794
+ "loss": 0.2793,
795
  "step": 6600
796
  },
797
  {
798
  "epoch": 26.8,
 
799
  "learning_rate": 0.0003,
800
+ "loss": 0.2905,
801
  "step": 6700
802
  },
803
  {
804
  "epoch": 27.0,
805
+ "eval_accuracy": 0.5867848101265822,
806
+ "eval_loss": 3.1007750034332275,
807
+ "eval_runtime": 4.5013,
808
+ "eval_samples_per_second": 111.079,
809
+ "eval_steps_per_second": 13.996,
810
  "step": 6750
811
  },
812
  {
813
  "epoch": 27.0,
814
+ "eval_exact_match": 1.2,
815
+ "eval_f1": 2.0656084656084652,
816
  "step": 6750
817
  },
818
  {
819
  "epoch": 27.2,
 
820
  "learning_rate": 0.0003,
821
+ "loss": 0.2964,
822
  "step": 6800
823
  },
824
  {
825
  "epoch": 27.6,
 
826
  "learning_rate": 0.0003,
827
+ "loss": 0.3021,
828
  "step": 6900
829
  },
830
  {
831
  "epoch": 28.0,
 
832
  "learning_rate": 0.0003,
833
+ "loss": 0.3237,
834
  "step": 7000
835
  },
836
  {
837
  "epoch": 28.0,
838
+ "eval_accuracy": 0.5851139240506329,
839
+ "eval_loss": 3.192138671875,
840
+ "eval_runtime": 4.5595,
841
+ "eval_samples_per_second": 109.661,
842
+ "eval_steps_per_second": 13.817,
843
  "step": 7000
844
  },
845
  {
846
  "epoch": 28.0,
847
+ "eval_exact_match": 0.8,
848
+ "eval_f1": 1.5059791467236656,
849
  "step": 7000
850
  },
851
  {
852
  "epoch": 28.4,
 
853
  "learning_rate": 0.0003,
854
+ "loss": 0.3584,
855
  "step": 7100
856
  },
857
  {
858
  "epoch": 28.8,
 
859
  "learning_rate": 0.0003,
860
+ "loss": 0.4738,
861
  "step": 7200
862
  },
863
  {
864
  "epoch": 29.0,
865
+ "eval_accuracy": 0.5789873417721519,
866
+ "eval_loss": 3.32121205329895,
867
+ "eval_runtime": 4.4073,
868
+ "eval_samples_per_second": 113.447,
869
+ "eval_steps_per_second": 14.294,
870
  "step": 7250
871
  },
872
  {
873
  "epoch": 29.0,
874
+ "eval_exact_match": 3.8,
875
+ "eval_f1": 6.689863653841047,
876
  "step": 7250
877
  },
878
  {
879
  "epoch": 29.2,
 
880
  "learning_rate": 0.0003,
881
+ "loss": 0.8318,
882
  "step": 7300
883
  },
884
  {
885
  "epoch": 29.6,
 
886
  "learning_rate": 0.0003,
887
+ "loss": 1.646,
888
  "step": 7400
889
  },
890
  {
891
  "epoch": 30.0,
 
892
  "learning_rate": 0.0003,
893
+ "loss": 2.5106,
894
  "step": 7500
895
  },
896
  {
897
  "epoch": 30.0,
898
+ "eval_accuracy": 0.5371645569620254,
899
+ "eval_loss": 3.9834704399108887,
900
+ "eval_runtime": 5.1142,
901
+ "eval_samples_per_second": 97.766,
902
+ "eval_steps_per_second": 12.319,
903
  "step": 7500
904
  },
905
  {
906
  "epoch": 30.0,
907
+ "eval_exact_match": 3.8,
908
+ "eval_f1": 5.247779866332498,
909
  "step": 7500
910
  },
911
  {
912
  "epoch": 30.4,
 
913
  "learning_rate": 0.0003,
914
+ "loss": 2.5833,
915
  "step": 7600
916
  },
917
  {
918
  "epoch": 30.8,
 
919
  "learning_rate": 0.0003,
920
+ "loss": 2.4442,
921
  "step": 7700
922
  },
923
  {
924
  "epoch": 31.0,
925
+ "eval_accuracy": 0.5435443037974683,
926
+ "eval_loss": 3.8099365234375,
927
+ "eval_runtime": 5.1354,
928
+ "eval_samples_per_second": 97.364,
929
+ "eval_steps_per_second": 12.268,
930
  "step": 7750
931
  },
932
  {
933
  "epoch": 31.0,
934
+ "eval_exact_match": 8.0,
935
+ "eval_f1": 11.652945717183004,
936
  "step": 7750
937
  },
938
  {
939
  "epoch": 31.2,
 
940
  "learning_rate": 0.0003,
941
+ "loss": 2.1299,
942
  "step": 7800
943
  },
944
  {
945
  "epoch": 31.6,
 
946
  "learning_rate": 0.0003,
947
+ "loss": 2.3675,
948
  "step": 7900
949
  },
950
  {
951
  "epoch": 32.0,
 
952
  "learning_rate": 0.0003,
953
+ "loss": 2.5329,
954
  "step": 8000
955
  },
956
  {
957
  "epoch": 32.0,
958
+ "eval_accuracy": 0.5304810126582279,
959
+ "eval_loss": 4.046270847320557,
960
+ "eval_runtime": 5.1297,
961
+ "eval_samples_per_second": 97.471,
962
+ "eval_steps_per_second": 12.281,
963
  "step": 8000
964
  },
965
  {
966
  "epoch": 32.0,
967
+ "eval_exact_match": 8.0,
968
+ "eval_f1": 11.118349417066474,
969
  "step": 8000
970
  },
971
  {
972
  "epoch": 32.4,
 
973
  "learning_rate": 0.0003,
974
+ "loss": 2.9966,
975
  "step": 8100
976
  },
977
  {
978
  "epoch": 32.8,
 
979
  "learning_rate": 0.0003,
980
+ "loss": 3.9607,
981
  "step": 8200
982
  },
983
  {
984
  "epoch": 33.0,
985
+ "eval_accuracy": 0.45944303797468355,
986
+ "eval_loss": 5.853953838348389,
987
+ "eval_runtime": 4.7634,
988
+ "eval_samples_per_second": 104.966,
989
+ "eval_steps_per_second": 13.226,
990
  "step": 8250
991
  },
992
  {
993
  "epoch": 33.0,
994
+ "eval_exact_match": 0.8,
995
+ "eval_f1": 2.1406916099773245,
996
  "step": 8250
997
  },
998
  {
999
  "epoch": 33.2,
 
1000
  "learning_rate": 0.0003,
1001
+ "loss": 5.3072,
1002
  "step": 8300
1003
  },
1004
  {
1005
  "epoch": 33.6,
 
1006
  "learning_rate": 0.0003,
1007
+ "loss": 8.0875,
1008
  "step": 8400
1009
  },
1010
  {
1011
  "epoch": 34.0,
 
1012
  "learning_rate": 0.0003,
1013
+ "loss": 8.2643,
1014
  "step": 8500
1015
  },
1016
  {
1017
  "epoch": 34.0,
1018
+ "eval_accuracy": 0.36860759493670886,
1019
+ "eval_loss": 9.332476615905762,
1020
+ "eval_runtime": 4.4983,
1021
+ "eval_samples_per_second": 111.154,
1022
+ "eval_steps_per_second": 14.005,
1023
  "step": 8500
1024
  },
1025
  {
1026
  "epoch": 34.0,
1027
+ "eval_exact_match": 0.0,
1028
+ "eval_f1": 0.12853014882332628,
1029
  "step": 8500
1030
  }
1031
  ],
1032
  "logging_steps": 100,
1033
  "max_steps": 12500,
 
1034
  "num_train_epochs": 50,
1035
  "save_steps": 500,
1036
+ "total_flos": 5.435707759033057e+17,
 
 
 
 
 
 
 
 
 
 
 
 
 
1037
  "trial_name": null,
1038
  "trial_params": null
1039
  }
checkpoint-8500/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:dae73bb50794744b259c1eaaeb67f6ae0b1d0f39b0a32b5f1162fc82e3a5d6a6
3
- size 5304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eef995e0a4a8da1af8ae275504624b1fcb2d18639b8c53b553ebed1c9a9b7fd7
3
+ size 4728