tyzhu commited on
Commit
61e932e
1 Parent(s): 2f2d115

Training in progress, epoch 10, checkpoint

Browse files
checkpoint-5290/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-5290/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed35235efe20323b866e40c10fd00add97bc2af32cae6228d8d152b53edf234a
3
+ size 143269386
checkpoint-5290/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-5290/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:407f90370e378825aa4f6bb645dc0cfe58c80bc5147e6a195dd3cda28364747f
3
- size 224537202
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17c7e3a96e4d7d1ecd15b12532addb9a80ac860888f06a501311cc196492508c
3
+ size 286585234
checkpoint-5290/rng_state_0.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:534472b69b4a10a5c79da5f83155c440b712f0b7cda8084663c9a5177168ae82
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8ea0e1cb8ca6671a4e42b993c09f1722bb0b5a5a728da168164be6fb8364893
3
  size 15024
checkpoint-5290/rng_state_1.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0814534bb070be08f43edda8ddf1336eac7a821519bf8f839f79a7f69a0e6124
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:57e32635c71ed1370bc1c6faa1803aa9dde27323aae603dd9e739965cd26f4d5
3
  size 15024
checkpoint-5290/rng_state_2.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:66b90a38d81689f180c460ef4727a85672c0efd72c8f566395450211203ef1e3
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fed98f922404836b922412ae9b255ad56bcc599d4a03cb9b581fc8cf98941e54
3
  size 15024
checkpoint-5290/rng_state_3.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:16a26582935f43448d93d3bd56fc6c65be1d30fb9ade216495350d16817984cd
3
  size 15024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21944e99d49031879e9c5f9f36d7c2c8e1a279765595d6b36e0b330082fbb10b
3
  size 15024
checkpoint-5290/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-5290/tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
checkpoint-5290/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
checkpoint-5290/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-5290/trainer_state.json CHANGED
@@ -9,539 +9,473 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 0.1890359168241966,
13
- "grad_norm": 0.5558584332466125,
14
  "learning_rate": 3e-05,
15
- "loss": 2.1103,
16
  "step": 100
17
  },
18
  {
19
- "epoch": 0.3780718336483932,
20
- "grad_norm": 0.4733140468597412,
21
  "learning_rate": 3e-05,
22
- "loss": 1.931,
23
  "step": 200
24
  },
25
  {
26
- "epoch": 0.5671077504725898,
27
- "grad_norm": 0.48580050468444824,
28
  "learning_rate": 3e-05,
29
- "loss": 1.8948,
30
  "step": 300
31
  },
32
  {
33
- "epoch": 0.7561436672967864,
34
- "grad_norm": 0.40655234456062317,
35
  "learning_rate": 3e-05,
36
- "loss": 1.8988,
37
  "step": 400
38
  },
39
  {
40
- "epoch": 0.945179584120983,
41
- "grad_norm": 0.6303955316543579,
42
  "learning_rate": 3e-05,
43
- "loss": 1.8745,
44
  "step": 500
45
  },
46
  {
47
  "epoch": 1.0,
48
- "eval_accuracy": 0.5701538461538461,
49
- "eval_loss": 1.6455246210098267,
50
- "eval_runtime": 5.0805,
51
- "eval_samples_per_second": 98.415,
52
- "eval_steps_per_second": 12.4,
53
  "step": 529
54
  },
55
  {
56
  "epoch": 1.0,
57
- "eval_exact_match": 15.0,
58
- "eval_f1": 23.927619047619068,
59
  "step": 529
60
  },
61
  {
62
- "epoch": 1.1342155009451795,
63
- "grad_norm": 0.4278818666934967,
64
  "learning_rate": 3e-05,
65
- "loss": 1.8798,
66
  "step": 600
67
  },
68
  {
69
- "epoch": 1.3232514177693762,
70
- "grad_norm": 0.449830025434494,
71
  "learning_rate": 3e-05,
72
- "loss": 1.8613,
73
  "step": 700
74
  },
75
  {
76
- "epoch": 1.5122873345935728,
77
- "grad_norm": 0.5130725502967834,
78
  "learning_rate": 3e-05,
79
- "loss": 1.8709,
80
  "step": 800
81
  },
82
  {
83
- "epoch": 1.7013232514177694,
84
- "grad_norm": 0.4889715611934662,
85
  "learning_rate": 3e-05,
86
- "loss": 1.8448,
87
  "step": 900
88
  },
89
  {
90
- "epoch": 1.8903591682419658,
91
- "grad_norm": 11.186380386352539,
92
  "learning_rate": 3e-05,
93
- "loss": 1.8764,
94
  "step": 1000
95
  },
96
  {
97
  "epoch": 2.0,
98
- "eval_accuracy": 0.5723589743589743,
99
- "eval_loss": 1.6143561601638794,
100
- "eval_runtime": 5.0466,
101
- "eval_samples_per_second": 99.077,
102
- "eval_steps_per_second": 12.484,
103
  "step": 1058
104
  },
105
  {
106
  "epoch": 2.0,
107
- "eval_exact_match": 15.6,
108
- "eval_f1": 25.825238095238113,
109
  "step": 1058
110
  },
111
  {
112
- "epoch": 2.0793950850661624,
113
- "grad_norm": 0.5414931774139404,
114
  "learning_rate": 3e-05,
115
- "loss": 1.8398,
116
  "step": 1100
117
  },
118
  {
119
- "epoch": 2.268431001890359,
120
- "grad_norm": 7.441005706787109,
121
  "learning_rate": 3e-05,
122
- "loss": 1.8348,
123
  "step": 1200
124
  },
125
  {
126
- "epoch": 2.4574669187145557,
127
- "grad_norm": 24.312145233154297,
128
  "learning_rate": 3e-05,
129
- "loss": 1.8421,
130
  "step": 1300
131
  },
132
  {
133
- "epoch": 2.6465028355387523,
134
- "grad_norm": 0.7407662868499756,
135
  "learning_rate": 3e-05,
136
- "loss": 1.8292,
137
  "step": 1400
138
  },
139
  {
140
- "epoch": 2.835538752362949,
141
- "grad_norm": 0.9653642177581787,
142
  "learning_rate": 3e-05,
143
- "loss": 1.828,
144
  "step": 1500
145
  },
146
  {
147
  "epoch": 3.0,
148
- "eval_accuracy": 0.5734358974358974,
149
- "eval_loss": 1.617864727973938,
150
- "eval_runtime": 5.2153,
151
- "eval_samples_per_second": 95.872,
152
- "eval_steps_per_second": 12.08,
153
  "step": 1587
154
  },
155
  {
156
  "epoch": 3.0,
157
- "eval_exact_match": 16.2,
158
- "eval_f1": 25.811428571428586,
159
  "step": 1587
160
  },
161
  {
162
- "epoch": 3.0245746691871456,
163
- "grad_norm": 450.6055603027344,
164
  "learning_rate": 3e-05,
165
- "loss": 1.8204,
166
  "step": 1600
167
  },
168
  {
169
- "epoch": 3.213610586011342,
170
- "grad_norm": 2.1127028465270996,
171
  "learning_rate": 3e-05,
172
- "loss": 1.8361,
173
  "step": 1700
174
  },
175
  {
176
- "epoch": 3.402646502835539,
177
- "grad_norm": 564.9320068359375,
178
  "learning_rate": 3e-05,
179
- "loss": 1.8338,
180
  "step": 1800
181
  },
182
  {
183
- "epoch": 3.5916824196597354,
184
- "grad_norm": 52.5078010559082,
185
  "learning_rate": 3e-05,
186
- "loss": 1.8075,
187
  "step": 1900
188
  },
189
  {
190
- "epoch": 3.780718336483932,
191
- "grad_norm": 27.87557601928711,
192
  "learning_rate": 3e-05,
193
- "loss": 1.8175,
194
  "step": 2000
195
  },
196
  {
197
- "epoch": 3.9697542533081287,
198
- "grad_norm": 4.810656547546387,
199
  "learning_rate": 3e-05,
200
- "loss": 1.8218,
201
  "step": 2100
202
  },
203
  {
204
  "epoch": 4.0,
205
- "eval_accuracy": 0.5728205128205128,
206
- "eval_loss": 1.6224406957626343,
207
- "eval_runtime": 5.0588,
208
- "eval_samples_per_second": 98.838,
209
- "eval_steps_per_second": 12.454,
210
  "step": 2116
211
  },
212
  {
213
  "epoch": 4.0,
214
- "eval_exact_match": 17.4,
215
- "eval_f1": 27.194285714285726,
216
  "step": 2116
217
  },
218
  {
219
- "epoch": 4.158790170132325,
220
- "grad_norm": 13.004076957702637,
221
  "learning_rate": 3e-05,
222
- "loss": 1.8089,
223
  "step": 2200
224
  },
225
  {
226
- "epoch": 4.3478260869565215,
227
- "grad_norm": 2.23368501663208,
228
  "learning_rate": 3e-05,
229
- "loss": 1.8292,
230
  "step": 2300
231
  },
232
  {
233
- "epoch": 4.536862003780718,
234
- "grad_norm": 0.8635586500167847,
235
  "learning_rate": 3e-05,
236
- "loss": 1.7992,
237
  "step": 2400
238
  },
239
  {
240
- "epoch": 4.725897920604915,
241
- "grad_norm": 6.5326738357543945,
242
  "learning_rate": 3e-05,
243
- "loss": 1.7955,
244
  "step": 2500
245
  },
246
  {
247
- "epoch": 4.914933837429111,
248
- "grad_norm": 463.4647521972656,
249
  "learning_rate": 3e-05,
250
- "loss": 1.7953,
251
  "step": 2600
252
  },
253
  {
254
  "epoch": 5.0,
255
- "eval_accuracy": 0.5740512820512821,
256
- "eval_loss": 1.6217337846755981,
257
- "eval_runtime": 5.1898,
258
- "eval_samples_per_second": 96.342,
259
- "eval_steps_per_second": 12.139,
260
  "step": 2645
261
  },
262
  {
263
  "epoch": 5.0,
264
- "eval_exact_match": 17.6,
265
- "eval_f1": 26.161428571428587,
266
  "step": 2645
267
  },
268
  {
269
- "epoch": 5.103969754253308,
270
- "grad_norm": 203.82815551757812,
271
  "learning_rate": 3e-05,
272
- "loss": 1.7845,
273
  "step": 2700
274
  },
275
  {
276
- "epoch": 5.293005671077505,
277
- "grad_norm": 660.8076782226562,
278
  "learning_rate": 3e-05,
279
- "loss": 1.7909,
280
  "step": 2800
281
  },
282
  {
283
- "epoch": 5.482041587901701,
284
- "grad_norm": 5727.51806640625,
285
  "learning_rate": 3e-05,
286
- "loss": 1.7909,
287
  "step": 2900
288
  },
289
  {
290
- "epoch": 5.671077504725898,
291
- "grad_norm": 4880698.0,
292
  "learning_rate": 3e-05,
293
- "loss": 3.5781,
294
  "step": 3000
295
  },
296
  {
297
- "epoch": 5.8601134215500945,
298
- "grad_norm": 1192.6517333984375,
299
  "learning_rate": 3e-05,
300
- "loss": 3.8703,
301
  "step": 3100
302
  },
303
  {
304
  "epoch": 6.0,
305
- "eval_accuracy": 0.5433846153846154,
306
- "eval_loss": 1.886108636856079,
307
- "eval_runtime": 5.5162,
308
- "eval_samples_per_second": 90.643,
309
- "eval_steps_per_second": 11.421,
310
  "step": 3174
311
  },
312
  {
313
  "epoch": 6.0,
314
- "eval_exact_match": 11.4,
315
- "eval_f1": 20.358571428571437,
316
  "step": 3174
317
  },
318
  {
319
- "epoch": 6.049149338374291,
320
- "grad_norm": 532.3778686523438,
321
  "learning_rate": 3e-05,
322
- "loss": 2.1701,
323
  "step": 3200
324
  },
325
  {
326
- "epoch": 6.238185255198488,
327
- "grad_norm": 565.6697998046875,
328
  "learning_rate": 3e-05,
329
- "loss": 1.8533,
330
  "step": 3300
331
  },
332
  {
333
- "epoch": 6.427221172022684,
334
- "grad_norm": 25206.482421875,
335
  "learning_rate": 3e-05,
336
- "loss": 2.623,
337
  "step": 3400
338
  },
339
  {
340
- "epoch": 6.616257088846881,
341
- "grad_norm": 71.88556671142578,
342
  "learning_rate": 3e-05,
343
- "loss": 3.1573,
344
  "step": 3500
345
  },
346
  {
347
- "epoch": 6.805293005671078,
348
- "grad_norm": 269563.15625,
349
  "learning_rate": 3e-05,
350
- "loss": 2.2696,
351
  "step": 3600
352
  },
353
  {
354
- "epoch": 6.994328922495274,
355
- "grad_norm": 11768.7880859375,
356
  "learning_rate": 3e-05,
357
- "loss": 3.1506,
358
  "step": 3700
359
  },
360
  {
361
  "epoch": 7.0,
362
- "eval_accuracy": 0.2834871794871795,
363
- "eval_loss": 8.311681747436523,
364
- "eval_runtime": 5.3223,
365
- "eval_samples_per_second": 93.944,
366
- "eval_steps_per_second": 11.837,
367
  "step": 3703
368
  },
369
  {
370
  "epoch": 7.0,
371
- "eval_exact_match": 0.0,
372
- "eval_f1": 0.027458851580988985,
373
  "step": 3703
374
  },
375
  {
376
- "epoch": 7.183364839319471,
377
- "grad_norm": 6220.75244140625,
378
  "learning_rate": 3e-05,
379
- "loss": 4.273,
380
  "step": 3800
381
  },
382
  {
383
- "epoch": 7.3724007561436675,
384
- "grad_norm": 3387.953857421875,
385
  "learning_rate": 3e-05,
386
- "loss": 7.7199,
387
  "step": 3900
388
  },
389
  {
390
- "epoch": 7.561436672967864,
391
- "grad_norm": 372.8681640625,
392
  "learning_rate": 3e-05,
393
- "loss": 7.7045,
394
  "step": 4000
395
  },
396
  {
397
- "epoch": 7.750472589792061,
398
- "grad_norm": 159.64283752441406,
399
  "learning_rate": 3e-05,
400
- "loss": 6.8901,
401
  "step": 4100
402
  },
403
  {
404
- "epoch": 7.939508506616257,
405
- "grad_norm": 751876.125,
406
  "learning_rate": 3e-05,
407
- "loss": 6.5238,
408
  "step": 4200
409
  },
410
  {
411
  "epoch": 8.0,
412
- "eval_accuracy": 0.22764102564102565,
413
- "eval_loss": 7.12347936630249,
414
- "eval_runtime": 5.3818,
415
- "eval_samples_per_second": 92.905,
416
- "eval_steps_per_second": 11.706,
417
  "step": 4232
418
  },
419
  {
420
  "epoch": 8.0,
421
- "eval_exact_match": 0.0,
422
- "eval_f1": 0.013333333333333336,
423
  "step": 4232
424
  },
425
  {
426
- "epoch": 8.128544423440454,
427
- "grad_norm": 1035664128.0,
428
  "learning_rate": 3e-05,
429
- "loss": 5.4157,
430
  "step": 4300
431
  },
432
  {
433
- "epoch": 8.31758034026465,
434
- "grad_norm": 11571242.0,
435
  "learning_rate": 3e-05,
436
- "loss": 5.8257,
437
  "step": 4400
438
  },
439
  {
440
- "epoch": 8.506616257088847,
441
- "grad_norm": 88236256.0,
442
  "learning_rate": 3e-05,
443
- "loss": 6.3609,
444
  "step": 4500
445
  },
446
  {
447
- "epoch": 8.695652173913043,
448
- "grad_norm": 105426080.0,
449
  "learning_rate": 3e-05,
450
- "loss": 6.1824,
451
  "step": 4600
452
  },
453
  {
454
- "epoch": 8.88468809073724,
455
- "grad_norm": 51438844.0,
456
  "learning_rate": 3e-05,
457
- "loss": 6.1818,
458
  "step": 4700
459
  },
460
  {
461
  "epoch": 9.0,
462
- "eval_accuracy": 0.1995897435897436,
463
- "eval_loss": 11.187646865844727,
464
- "eval_runtime": 5.154,
465
- "eval_samples_per_second": 97.013,
466
- "eval_steps_per_second": 12.224,
467
  "step": 4761
468
  },
469
  {
470
  "epoch": 9.0,
471
- "eval_exact_match": 0.0,
472
- "eval_f1": 0.0,
473
  "step": 4761
474
  },
475
  {
476
- "epoch": 9.073724007561436,
477
- "grad_norm": 8742571.0,
478
  "learning_rate": 3e-05,
479
- "loss": 7.5029,
480
  "step": 4800
481
  },
482
  {
483
- "epoch": 9.262759924385634,
484
- "grad_norm": 4885384192.0,
485
  "learning_rate": 3e-05,
486
- "loss": 7.805,
487
  "step": 4900
488
  },
489
  {
490
- "epoch": 9.45179584120983,
491
- "grad_norm": 919074304.0,
492
  "learning_rate": 3e-05,
493
- "loss": 8.0712,
494
  "step": 5000
495
  },
496
  {
497
- "epoch": 9.640831758034027,
498
- "grad_norm": 9706883072.0,
499
  "learning_rate": 3e-05,
500
- "loss": 8.4915,
501
  "step": 5100
502
  },
503
  {
504
- "epoch": 9.829867674858223,
505
- "grad_norm": 35723780096.0,
506
  "learning_rate": 3e-05,
507
- "loss": 8.3286,
508
  "step": 5200
509
  },
510
  {
511
  "epoch": 10.0,
512
- "eval_accuracy": 0.1995897435897436,
513
- "eval_loss": 13.374640464782715,
514
- "eval_runtime": 5.2089,
515
- "eval_samples_per_second": 95.99,
516
- "eval_steps_per_second": 12.095,
517
  "step": 5290
518
  },
519
  {
520
  "epoch": 10.0,
521
- "eval_exact_match": 0.0,
522
- "eval_f1": 0.0,
523
  "step": 5290
524
  }
525
  ],
526
  "logging_steps": 100,
527
  "max_steps": 26450,
528
- "num_input_tokens_seen": 0,
529
  "num_train_epochs": 50,
530
  "save_steps": 500,
531
- "stateful_callbacks": {
532
- "TrainerControl": {
533
- "args": {
534
- "should_epoch_stop": false,
535
- "should_evaluate": false,
536
- "should_log": false,
537
- "should_save": true,
538
- "should_training_stop": false
539
- },
540
- "attributes": {}
541
- }
542
- },
543
- "total_flos": 4.3074604156688794e+17,
544
- "train_batch_size": 2,
545
  "trial_name": null,
546
  "trial_params": null
547
  }
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.19,
 
13
  "learning_rate": 3e-05,
14
+ "loss": 1.7194,
15
  "step": 100
16
  },
17
  {
18
+ "epoch": 0.38,
 
19
  "learning_rate": 3e-05,
20
+ "loss": 1.4369,
21
  "step": 200
22
  },
23
  {
24
+ "epoch": 0.57,
 
25
  "learning_rate": 3e-05,
26
+ "loss": 1.4257,
27
  "step": 300
28
  },
29
  {
30
+ "epoch": 0.76,
 
31
  "learning_rate": 3e-05,
32
+ "loss": 1.3947,
33
  "step": 400
34
  },
35
  {
36
+ "epoch": 0.95,
 
37
  "learning_rate": 3e-05,
38
+ "loss": 1.3948,
39
  "step": 500
40
  },
41
  {
42
  "epoch": 1.0,
43
+ "eval_accuracy": 0.6132307692307692,
44
+ "eval_loss": 1.3087366819381714,
45
+ "eval_runtime": 4.3553,
46
+ "eval_samples_per_second": 114.803,
47
+ "eval_steps_per_second": 14.465,
48
  "step": 529
49
  },
50
  {
51
  "epoch": 1.0,
52
+ "eval_exact_match": 24.4,
53
+ "eval_f1": 37.614420394420414,
54
  "step": 529
55
  },
56
  {
57
+ "epoch": 1.13,
 
58
  "learning_rate": 3e-05,
59
+ "loss": 1.3927,
60
  "step": 600
61
  },
62
  {
63
+ "epoch": 1.32,
 
64
  "learning_rate": 3e-05,
65
+ "loss": 1.3644,
66
  "step": 700
67
  },
68
  {
69
+ "epoch": 1.51,
 
70
  "learning_rate": 3e-05,
71
+ "loss": 1.3791,
72
  "step": 800
73
  },
74
  {
75
+ "epoch": 1.7,
 
76
  "learning_rate": 3e-05,
77
+ "loss": 1.3627,
78
  "step": 900
79
  },
80
  {
81
+ "epoch": 1.89,
 
82
  "learning_rate": 3e-05,
83
+ "loss": 1.3789,
84
  "step": 1000
85
  },
86
  {
87
  "epoch": 2.0,
88
+ "eval_accuracy": 0.6146153846153846,
89
+ "eval_loss": 1.289727807044983,
90
+ "eval_runtime": 4.4281,
91
+ "eval_samples_per_second": 112.915,
92
+ "eval_steps_per_second": 14.227,
93
  "step": 1058
94
  },
95
  {
96
  "epoch": 2.0,
97
+ "eval_exact_match": 23.2,
98
+ "eval_f1": 35.56775335775336,
99
  "step": 1058
100
  },
101
  {
102
+ "epoch": 2.08,
 
103
  "learning_rate": 3e-05,
104
+ "loss": 1.3605,
105
  "step": 1100
106
  },
107
  {
108
+ "epoch": 2.27,
 
109
  "learning_rate": 3e-05,
110
+ "loss": 1.3306,
111
  "step": 1200
112
  },
113
  {
114
+ "epoch": 2.46,
 
115
  "learning_rate": 3e-05,
116
+ "loss": 1.3318,
117
  "step": 1300
118
  },
119
  {
120
+ "epoch": 2.65,
 
121
  "learning_rate": 3e-05,
122
+ "loss": 1.3491,
123
  "step": 1400
124
  },
125
  {
126
+ "epoch": 2.84,
 
127
  "learning_rate": 3e-05,
128
+ "loss": 1.3259,
129
  "step": 1500
130
  },
131
  {
132
  "epoch": 3.0,
133
+ "eval_accuracy": 0.6178974358974358,
134
+ "eval_loss": 1.2849316596984863,
135
+ "eval_runtime": 4.875,
136
+ "eval_samples_per_second": 102.565,
137
+ "eval_steps_per_second": 12.923,
138
  "step": 1587
139
  },
140
  {
141
  "epoch": 3.0,
142
+ "eval_exact_match": 25.2,
143
+ "eval_f1": 37.32698412698414,
144
  "step": 1587
145
  },
146
  {
147
+ "epoch": 3.02,
 
148
  "learning_rate": 3e-05,
149
+ "loss": 1.3255,
150
  "step": 1600
151
  },
152
  {
153
+ "epoch": 3.21,
 
154
  "learning_rate": 3e-05,
155
+ "loss": 1.28,
156
  "step": 1700
157
  },
158
  {
159
+ "epoch": 3.4,
 
160
  "learning_rate": 3e-05,
161
+ "loss": 1.2995,
162
  "step": 1800
163
  },
164
  {
165
+ "epoch": 3.59,
 
166
  "learning_rate": 3e-05,
167
+ "loss": 1.2991,
168
  "step": 1900
169
  },
170
  {
171
+ "epoch": 3.78,
 
172
  "learning_rate": 3e-05,
173
+ "loss": 1.3024,
174
  "step": 2000
175
  },
176
  {
177
+ "epoch": 3.97,
 
178
  "learning_rate": 3e-05,
179
+ "loss": 1.2853,
180
  "step": 2100
181
  },
182
  {
183
  "epoch": 4.0,
184
+ "eval_accuracy": 0.615948717948718,
185
+ "eval_loss": 1.316924810409546,
186
+ "eval_runtime": 5.3385,
187
+ "eval_samples_per_second": 93.659,
188
+ "eval_steps_per_second": 11.801,
189
  "step": 2116
190
  },
191
  {
192
  "epoch": 4.0,
193
+ "eval_exact_match": 23.8,
194
+ "eval_f1": 36.284920634920645,
195
  "step": 2116
196
  },
197
  {
198
+ "epoch": 4.16,
 
199
  "learning_rate": 3e-05,
200
+ "loss": 1.2575,
201
  "step": 2200
202
  },
203
  {
204
+ "epoch": 4.35,
 
205
  "learning_rate": 3e-05,
206
+ "loss": 1.2261,
207
  "step": 2300
208
  },
209
  {
210
+ "epoch": 4.54,
 
211
  "learning_rate": 3e-05,
212
+ "loss": 1.2536,
213
  "step": 2400
214
  },
215
  {
216
+ "epoch": 4.73,
 
217
  "learning_rate": 3e-05,
218
+ "loss": 1.2376,
219
  "step": 2500
220
  },
221
  {
222
+ "epoch": 4.91,
 
223
  "learning_rate": 3e-05,
224
+ "loss": 1.2556,
225
  "step": 2600
226
  },
227
  {
228
  "epoch": 5.0,
229
+ "eval_accuracy": 0.6131794871794872,
230
+ "eval_loss": 1.3531945943832397,
231
+ "eval_runtime": 4.4382,
232
+ "eval_samples_per_second": 112.659,
233
+ "eval_steps_per_second": 14.195,
234
  "step": 2645
235
  },
236
  {
237
  "epoch": 5.0,
238
+ "eval_exact_match": 25.8,
239
+ "eval_f1": 37.67571428571429,
240
  "step": 2645
241
  },
242
  {
243
+ "epoch": 5.1,
 
244
  "learning_rate": 3e-05,
245
+ "loss": 1.2195,
246
  "step": 2700
247
  },
248
  {
249
+ "epoch": 5.29,
 
250
  "learning_rate": 3e-05,
251
+ "loss": 1.2003,
252
  "step": 2800
253
  },
254
  {
255
+ "epoch": 5.48,
 
256
  "learning_rate": 3e-05,
257
+ "loss": 1.1864,
258
  "step": 2900
259
  },
260
  {
261
+ "epoch": 5.67,
 
262
  "learning_rate": 3e-05,
263
+ "loss": 1.2231,
264
  "step": 3000
265
  },
266
  {
267
+ "epoch": 5.86,
 
268
  "learning_rate": 3e-05,
269
+ "loss": 1.1972,
270
  "step": 3100
271
  },
272
  {
273
  "epoch": 6.0,
274
+ "eval_accuracy": 0.6125641025641025,
275
+ "eval_loss": 1.4135066270828247,
276
+ "eval_runtime": 5.0312,
277
+ "eval_samples_per_second": 99.38,
278
+ "eval_steps_per_second": 12.522,
279
  "step": 3174
280
  },
281
  {
282
  "epoch": 6.0,
283
+ "eval_exact_match": 24.0,
284
+ "eval_f1": 36.140952380952385,
285
  "step": 3174
286
  },
287
  {
288
+ "epoch": 6.05,
 
289
  "learning_rate": 3e-05,
290
+ "loss": 1.1889,
291
  "step": 3200
292
  },
293
  {
294
+ "epoch": 6.24,
 
295
  "learning_rate": 3e-05,
296
+ "loss": 1.1471,
297
  "step": 3300
298
  },
299
  {
300
+ "epoch": 6.43,
 
301
  "learning_rate": 3e-05,
302
+ "loss": 1.1509,
303
  "step": 3400
304
  },
305
  {
306
+ "epoch": 6.62,
 
307
  "learning_rate": 3e-05,
308
+ "loss": 1.1654,
309
  "step": 3500
310
  },
311
  {
312
+ "epoch": 6.81,
 
313
  "learning_rate": 3e-05,
314
+ "loss": 1.1716,
315
  "step": 3600
316
  },
317
  {
318
+ "epoch": 6.99,
 
319
  "learning_rate": 3e-05,
320
+ "loss": 1.1839,
321
  "step": 3700
322
  },
323
  {
324
  "epoch": 7.0,
325
+ "eval_accuracy": 0.6081025641025641,
326
+ "eval_loss": 1.500714898109436,
327
+ "eval_runtime": 4.4677,
328
+ "eval_samples_per_second": 111.916,
329
+ "eval_steps_per_second": 14.101,
330
  "step": 3703
331
  },
332
  {
333
  "epoch": 7.0,
334
+ "eval_exact_match": 22.8,
335
+ "eval_f1": 33.44777777777779,
336
  "step": 3703
337
  },
338
  {
339
+ "epoch": 7.18,
 
340
  "learning_rate": 3e-05,
341
+ "loss": 1.1059,
342
  "step": 3800
343
  },
344
  {
345
+ "epoch": 7.37,
 
346
  "learning_rate": 3e-05,
347
+ "loss": 1.1146,
348
  "step": 3900
349
  },
350
  {
351
+ "epoch": 7.56,
 
352
  "learning_rate": 3e-05,
353
+ "loss": 1.1121,
354
  "step": 4000
355
  },
356
  {
357
+ "epoch": 7.75,
 
358
  "learning_rate": 3e-05,
359
+ "loss": 1.1286,
360
  "step": 4100
361
  },
362
  {
363
+ "epoch": 7.94,
 
364
  "learning_rate": 3e-05,
365
+ "loss": 1.1334,
366
  "step": 4200
367
  },
368
  {
369
  "epoch": 8.0,
370
+ "eval_accuracy": 0.6074358974358974,
371
+ "eval_loss": 1.524243950843811,
372
+ "eval_runtime": 4.2161,
373
+ "eval_samples_per_second": 118.592,
374
+ "eval_steps_per_second": 14.943,
375
  "step": 4232
376
  },
377
  {
378
  "epoch": 8.0,
379
+ "eval_exact_match": 23.2,
380
+ "eval_f1": 34.507792207792214,
381
  "step": 4232
382
  },
383
  {
384
+ "epoch": 8.13,
 
385
  "learning_rate": 3e-05,
386
+ "loss": 1.1072,
387
  "step": 4300
388
  },
389
  {
390
+ "epoch": 8.32,
 
391
  "learning_rate": 3e-05,
392
+ "loss": 1.0626,
393
  "step": 4400
394
  },
395
  {
396
+ "epoch": 8.51,
 
397
  "learning_rate": 3e-05,
398
+ "loss": 1.0842,
399
  "step": 4500
400
  },
401
  {
402
+ "epoch": 8.7,
 
403
  "learning_rate": 3e-05,
404
+ "loss": 1.0812,
405
  "step": 4600
406
  },
407
  {
408
+ "epoch": 8.88,
 
409
  "learning_rate": 3e-05,
410
+ "loss": 1.0966,
411
  "step": 4700
412
  },
413
  {
414
  "epoch": 9.0,
415
+ "eval_accuracy": 0.5802564102564103,
416
+ "eval_loss": 1.6106736660003662,
417
+ "eval_runtime": 4.2194,
418
+ "eval_samples_per_second": 118.5,
419
+ "eval_steps_per_second": 14.931,
420
  "step": 4761
421
  },
422
  {
423
  "epoch": 9.0,
424
+ "eval_exact_match": 22.8,
425
+ "eval_f1": 34.45666666666666,
426
  "step": 4761
427
  },
428
  {
429
+ "epoch": 9.07,
 
430
  "learning_rate": 3e-05,
431
+ "loss": 1.0755,
432
  "step": 4800
433
  },
434
  {
435
+ "epoch": 9.26,
 
436
  "learning_rate": 3e-05,
437
+ "loss": 1.0189,
438
  "step": 4900
439
  },
440
  {
441
+ "epoch": 9.45,
 
442
  "learning_rate": 3e-05,
443
+ "loss": 1.0565,
444
  "step": 5000
445
  },
446
  {
447
+ "epoch": 9.64,
 
448
  "learning_rate": 3e-05,
449
+ "loss": 1.0635,
450
  "step": 5100
451
  },
452
  {
453
+ "epoch": 9.83,
 
454
  "learning_rate": 3e-05,
455
+ "loss": 1.0485,
456
  "step": 5200
457
  },
458
  {
459
  "epoch": 10.0,
460
+ "eval_accuracy": 0.6049230769230769,
461
+ "eval_loss": 1.6749216318130493,
462
+ "eval_runtime": 4.8521,
463
+ "eval_samples_per_second": 103.048,
464
+ "eval_steps_per_second": 12.984,
465
  "step": 5290
466
  },
467
  {
468
  "epoch": 10.0,
469
+ "eval_exact_match": 22.8,
470
+ "eval_f1": 34.3535064935065,
471
  "step": 5290
472
  }
473
  ],
474
  "logging_steps": 100,
475
  "max_steps": 26450,
 
476
  "num_train_epochs": 50,
477
  "save_steps": 500,
478
+ "total_flos": 8.937870479180431e+17,
 
 
 
 
 
 
 
 
 
 
 
 
 
479
  "trial_name": null,
480
  "trial_params": null
481
  }
checkpoint-5290/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3895667ea94410ecd283430024ce6f8557816c9d86efd03783de342ef988d79c
3
- size 5304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec9dd2a1e083127818a65cb4d99f38a0a803e54a476491027d2e996644eba47c
3
+ size 4728