GaetanMichelet commited on
Commit
bccd625
1 Parent(s): 5c91c21

Model save

Browse files
README.md ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
3
+ library_name: peft
4
+ license: llama3.1
5
+ tags:
6
+ - trl
7
+ - sft
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: Llama-31-8B_task-3_180-samples_config-2_full
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # Llama-31-8B_task-3_180-samples_config-2_full
18
+
19
+ This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on the None dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 1.3190
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 0.0001
41
+ - train_batch_size: 1
42
+ - eval_batch_size: 1
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - gradient_accumulation_steps: 16
46
+ - total_train_batch_size: 16
47
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
+ - lr_scheduler_type: cosine
49
+ - lr_scheduler_warmup_ratio: 0.1
50
+ - num_epochs: 50
51
+
52
+ ### Training results
53
+
54
+ | Training Loss | Epoch | Step | Validation Loss |
55
+ |:-------------:|:-------:|:----:|:---------------:|
56
+ | 1.695 | 0.9412 | 8 | 1.6759 |
57
+ | 1.5868 | 2.0 | 17 | 1.5454 |
58
+ | 1.473 | 2.9412 | 25 | 1.4198 |
59
+ | 1.2035 | 4.0 | 34 | 1.2535 |
60
+ | 1.1293 | 4.9412 | 42 | 1.1991 |
61
+ | 1.1361 | 6.0 | 51 | 1.1733 |
62
+ | 1.1333 | 6.9412 | 59 | 1.1571 |
63
+ | 1.0612 | 8.0 | 68 | 1.1462 |
64
+ | 0.9895 | 8.9412 | 76 | 1.1392 |
65
+ | 0.9858 | 10.0 | 85 | 1.1381 |
66
+ | 0.939 | 10.9412 | 93 | 1.1420 |
67
+ | 0.8747 | 12.0 | 102 | 1.1664 |
68
+ | 0.8694 | 12.9412 | 110 | 1.1780 |
69
+ | 0.8188 | 14.0 | 119 | 1.2246 |
70
+ | 0.697 | 14.9412 | 127 | 1.2348 |
71
+ | 0.6048 | 16.0 | 136 | 1.3102 |
72
+ | 0.5898 | 16.9412 | 144 | 1.3190 |
73
+
74
+
75
+ ### Framework versions
76
+
77
+ - PEFT 0.12.0
78
+ - Transformers 4.44.0
79
+ - Pytorch 2.1.2+cu121
80
+ - Datasets 2.20.0
81
+ - Tokenizers 0.19.1
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d59ef4465df82226ce3e10bdba896ba0ccc72725d4727de351c45c67980048ec
3
  size 167832240
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a59f37bc603f2b8ce83d323c29b3b2d168411ade494b18a1b4afe0275a47652
3
  size 167832240
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 16.941176470588236,
3
+ "total_flos": 1.8132641623362765e+17,
4
+ "train_loss": 1.0652225611524448,
5
+ "train_runtime": 6655.309,
6
+ "train_samples": 136,
7
+ "train_samples_per_second": 1.022,
8
+ "train_steps_per_second": 0.06
9
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 16.941176470588236,
3
+ "total_flos": 1.8132641623362765e+17,
4
+ "train_loss": 1.0652225611524448,
5
+ "train_runtime": 6655.309,
6
+ "train_samples": 136,
7
+ "train_samples_per_second": 1.022,
8
+ "train_steps_per_second": 0.06
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,698 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.1381473541259766,
3
+ "best_model_checkpoint": "data/Llama-31-8B_task-3_180-samples_config-2_full/checkpoint-85",
4
+ "epoch": 16.941176470588236,
5
+ "eval_steps": 500,
6
+ "global_step": 144,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.11764705882352941,
13
+ "grad_norm": 0.5176953077316284,
14
+ "learning_rate": 2.5e-06,
15
+ "loss": 1.7131,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.23529411764705882,
20
+ "grad_norm": 0.5487761497497559,
21
+ "learning_rate": 5e-06,
22
+ "loss": 1.7301,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.47058823529411764,
27
+ "grad_norm": 0.4699436128139496,
28
+ "learning_rate": 1e-05,
29
+ "loss": 1.7256,
30
+ "step": 4
31
+ },
32
+ {
33
+ "epoch": 0.7058823529411765,
34
+ "grad_norm": 0.42270535230636597,
35
+ "learning_rate": 1.5e-05,
36
+ "loss": 1.6691,
37
+ "step": 6
38
+ },
39
+ {
40
+ "epoch": 0.9411764705882353,
41
+ "grad_norm": 0.41488713026046753,
42
+ "learning_rate": 2e-05,
43
+ "loss": 1.695,
44
+ "step": 8
45
+ },
46
+ {
47
+ "epoch": 0.9411764705882353,
48
+ "eval_loss": 1.6759352684020996,
49
+ "eval_runtime": 31.8571,
50
+ "eval_samples_per_second": 1.13,
51
+ "eval_steps_per_second": 1.13,
52
+ "step": 8
53
+ },
54
+ {
55
+ "epoch": 1.1764705882352942,
56
+ "grad_norm": 0.5180110335350037,
57
+ "learning_rate": 2.5e-05,
58
+ "loss": 1.6882,
59
+ "step": 10
60
+ },
61
+ {
62
+ "epoch": 1.4117647058823528,
63
+ "grad_norm": 0.35783591866493225,
64
+ "learning_rate": 3e-05,
65
+ "loss": 1.6225,
66
+ "step": 12
67
+ },
68
+ {
69
+ "epoch": 1.6470588235294117,
70
+ "grad_norm": 0.4034726619720459,
71
+ "learning_rate": 3.5e-05,
72
+ "loss": 1.5951,
73
+ "step": 14
74
+ },
75
+ {
76
+ "epoch": 1.8823529411764706,
77
+ "grad_norm": 0.4607871472835541,
78
+ "learning_rate": 4e-05,
79
+ "loss": 1.5868,
80
+ "step": 16
81
+ },
82
+ {
83
+ "epoch": 2.0,
84
+ "eval_loss": 1.5454449653625488,
85
+ "eval_runtime": 31.8528,
86
+ "eval_samples_per_second": 1.13,
87
+ "eval_steps_per_second": 1.13,
88
+ "step": 17
89
+ },
90
+ {
91
+ "epoch": 2.1176470588235294,
92
+ "grad_norm": 0.2619361877441406,
93
+ "learning_rate": 4.5e-05,
94
+ "loss": 1.5575,
95
+ "step": 18
96
+ },
97
+ {
98
+ "epoch": 2.3529411764705883,
99
+ "grad_norm": 0.23799021542072296,
100
+ "learning_rate": 5e-05,
101
+ "loss": 1.4926,
102
+ "step": 20
103
+ },
104
+ {
105
+ "epoch": 2.588235294117647,
106
+ "grad_norm": 0.2394094467163086,
107
+ "learning_rate": 5.500000000000001e-05,
108
+ "loss": 1.4593,
109
+ "step": 22
110
+ },
111
+ {
112
+ "epoch": 2.8235294117647056,
113
+ "grad_norm": 0.255198210477829,
114
+ "learning_rate": 6e-05,
115
+ "loss": 1.473,
116
+ "step": 24
117
+ },
118
+ {
119
+ "epoch": 2.9411764705882355,
120
+ "eval_loss": 1.4197732210159302,
121
+ "eval_runtime": 31.8597,
122
+ "eval_samples_per_second": 1.13,
123
+ "eval_steps_per_second": 1.13,
124
+ "step": 25
125
+ },
126
+ {
127
+ "epoch": 3.0588235294117645,
128
+ "grad_norm": 0.23705199360847473,
129
+ "learning_rate": 6.500000000000001e-05,
130
+ "loss": 1.3998,
131
+ "step": 26
132
+ },
133
+ {
134
+ "epoch": 3.2941176470588234,
135
+ "grad_norm": 0.2624831199645996,
136
+ "learning_rate": 7e-05,
137
+ "loss": 1.3573,
138
+ "step": 28
139
+ },
140
+ {
141
+ "epoch": 3.5294117647058822,
142
+ "grad_norm": 0.27011838555336,
143
+ "learning_rate": 7.500000000000001e-05,
144
+ "loss": 1.2873,
145
+ "step": 30
146
+ },
147
+ {
148
+ "epoch": 3.764705882352941,
149
+ "grad_norm": 0.2953531742095947,
150
+ "learning_rate": 8e-05,
151
+ "loss": 1.3048,
152
+ "step": 32
153
+ },
154
+ {
155
+ "epoch": 4.0,
156
+ "grad_norm": 0.2443869560956955,
157
+ "learning_rate": 8.5e-05,
158
+ "loss": 1.2035,
159
+ "step": 34
160
+ },
161
+ {
162
+ "epoch": 4.0,
163
+ "eval_loss": 1.2535014152526855,
164
+ "eval_runtime": 31.8533,
165
+ "eval_samples_per_second": 1.13,
166
+ "eval_steps_per_second": 1.13,
167
+ "step": 34
168
+ },
169
+ {
170
+ "epoch": 4.235294117647059,
171
+ "grad_norm": 0.2208809107542038,
172
+ "learning_rate": 9e-05,
173
+ "loss": 1.1907,
174
+ "step": 36
175
+ },
176
+ {
177
+ "epoch": 4.470588235294118,
178
+ "grad_norm": 0.1921616941690445,
179
+ "learning_rate": 9.5e-05,
180
+ "loss": 1.1664,
181
+ "step": 38
182
+ },
183
+ {
184
+ "epoch": 4.705882352941177,
185
+ "grad_norm": 0.15039095282554626,
186
+ "learning_rate": 0.0001,
187
+ "loss": 1.1716,
188
+ "step": 40
189
+ },
190
+ {
191
+ "epoch": 4.9411764705882355,
192
+ "grad_norm": 0.15168997645378113,
193
+ "learning_rate": 9.999238475781957e-05,
194
+ "loss": 1.1293,
195
+ "step": 42
196
+ },
197
+ {
198
+ "epoch": 4.9411764705882355,
199
+ "eval_loss": 1.199088454246521,
200
+ "eval_runtime": 31.8617,
201
+ "eval_samples_per_second": 1.13,
202
+ "eval_steps_per_second": 1.13,
203
+ "step": 42
204
+ },
205
+ {
206
+ "epoch": 5.176470588235294,
207
+ "grad_norm": 0.16245824098587036,
208
+ "learning_rate": 9.99695413509548e-05,
209
+ "loss": 1.0927,
210
+ "step": 44
211
+ },
212
+ {
213
+ "epoch": 5.411764705882353,
214
+ "grad_norm": 0.17652744054794312,
215
+ "learning_rate": 9.99314767377287e-05,
216
+ "loss": 1.1417,
217
+ "step": 46
218
+ },
219
+ {
220
+ "epoch": 5.647058823529412,
221
+ "grad_norm": 0.18893443048000336,
222
+ "learning_rate": 9.987820251299122e-05,
223
+ "loss": 1.1069,
224
+ "step": 48
225
+ },
226
+ {
227
+ "epoch": 5.882352941176471,
228
+ "grad_norm": 0.16078542172908783,
229
+ "learning_rate": 9.980973490458728e-05,
230
+ "loss": 1.1361,
231
+ "step": 50
232
+ },
233
+ {
234
+ "epoch": 6.0,
235
+ "eval_loss": 1.1733344793319702,
236
+ "eval_runtime": 31.8555,
237
+ "eval_samples_per_second": 1.13,
238
+ "eval_steps_per_second": 1.13,
239
+ "step": 51
240
+ },
241
+ {
242
+ "epoch": 6.117647058823529,
243
+ "grad_norm": 0.1555337905883789,
244
+ "learning_rate": 9.972609476841367e-05,
245
+ "loss": 1.0797,
246
+ "step": 52
247
+ },
248
+ {
249
+ "epoch": 6.352941176470588,
250
+ "grad_norm": 0.14737282693386078,
251
+ "learning_rate": 9.962730758206611e-05,
252
+ "loss": 1.0578,
253
+ "step": 54
254
+ },
255
+ {
256
+ "epoch": 6.588235294117647,
257
+ "grad_norm": 0.16058345139026642,
258
+ "learning_rate": 9.951340343707852e-05,
259
+ "loss": 1.0911,
260
+ "step": 56
261
+ },
262
+ {
263
+ "epoch": 6.823529411764706,
264
+ "grad_norm": 0.14137417078018188,
265
+ "learning_rate": 9.938441702975689e-05,
266
+ "loss": 1.1333,
267
+ "step": 58
268
+ },
269
+ {
270
+ "epoch": 6.9411764705882355,
271
+ "eval_loss": 1.1571494340896606,
272
+ "eval_runtime": 31.8579,
273
+ "eval_samples_per_second": 1.13,
274
+ "eval_steps_per_second": 1.13,
275
+ "step": 59
276
+ },
277
+ {
278
+ "epoch": 7.0588235294117645,
279
+ "grad_norm": 0.15388211607933044,
280
+ "learning_rate": 9.924038765061042e-05,
281
+ "loss": 1.0784,
282
+ "step": 60
283
+ },
284
+ {
285
+ "epoch": 7.294117647058823,
286
+ "grad_norm": 0.1491321623325348,
287
+ "learning_rate": 9.908135917238321e-05,
288
+ "loss": 1.07,
289
+ "step": 62
290
+ },
291
+ {
292
+ "epoch": 7.529411764705882,
293
+ "grad_norm": 0.17386263608932495,
294
+ "learning_rate": 9.890738003669029e-05,
295
+ "loss": 1.0382,
296
+ "step": 64
297
+ },
298
+ {
299
+ "epoch": 7.764705882352941,
300
+ "grad_norm": 0.17576885223388672,
301
+ "learning_rate": 9.871850323926177e-05,
302
+ "loss": 1.0377,
303
+ "step": 66
304
+ },
305
+ {
306
+ "epoch": 8.0,
307
+ "grad_norm": 0.1866431087255478,
308
+ "learning_rate": 9.851478631379982e-05,
309
+ "loss": 1.0612,
310
+ "step": 68
311
+ },
312
+ {
313
+ "epoch": 8.0,
314
+ "eval_loss": 1.1461663246154785,
315
+ "eval_runtime": 31.8577,
316
+ "eval_samples_per_second": 1.13,
317
+ "eval_steps_per_second": 1.13,
318
+ "step": 68
319
+ },
320
+ {
321
+ "epoch": 8.235294117647058,
322
+ "grad_norm": 0.1761140078306198,
323
+ "learning_rate": 9.829629131445342e-05,
324
+ "loss": 1.0261,
325
+ "step": 70
326
+ },
327
+ {
328
+ "epoch": 8.470588235294118,
329
+ "grad_norm": 0.19430580735206604,
330
+ "learning_rate": 9.806308479691595e-05,
331
+ "loss": 1.0604,
332
+ "step": 72
333
+ },
334
+ {
335
+ "epoch": 8.705882352941176,
336
+ "grad_norm": 0.19356723129749298,
337
+ "learning_rate": 9.781523779815179e-05,
338
+ "loss": 1.0369,
339
+ "step": 74
340
+ },
341
+ {
342
+ "epoch": 8.941176470588236,
343
+ "grad_norm": 0.21120838820934296,
344
+ "learning_rate": 9.755282581475769e-05,
345
+ "loss": 0.9895,
346
+ "step": 76
347
+ },
348
+ {
349
+ "epoch": 8.941176470588236,
350
+ "eval_loss": 1.1391992568969727,
351
+ "eval_runtime": 31.8668,
352
+ "eval_samples_per_second": 1.13,
353
+ "eval_steps_per_second": 1.13,
354
+ "step": 76
355
+ },
356
+ {
357
+ "epoch": 9.176470588235293,
358
+ "grad_norm": 0.16879647970199585,
359
+ "learning_rate": 9.727592877996585e-05,
360
+ "loss": 1.0034,
361
+ "step": 78
362
+ },
363
+ {
364
+ "epoch": 9.411764705882353,
365
+ "grad_norm": 0.1657697558403015,
366
+ "learning_rate": 9.698463103929542e-05,
367
+ "loss": 1.0109,
368
+ "step": 80
369
+ },
370
+ {
371
+ "epoch": 9.647058823529411,
372
+ "grad_norm": 0.19832511246204376,
373
+ "learning_rate": 9.667902132486009e-05,
374
+ "loss": 1.003,
375
+ "step": 82
376
+ },
377
+ {
378
+ "epoch": 9.882352941176471,
379
+ "grad_norm": 0.219316765666008,
380
+ "learning_rate": 9.635919272833938e-05,
381
+ "loss": 0.9858,
382
+ "step": 84
383
+ },
384
+ {
385
+ "epoch": 10.0,
386
+ "eval_loss": 1.1381473541259766,
387
+ "eval_runtime": 31.8595,
388
+ "eval_samples_per_second": 1.13,
389
+ "eval_steps_per_second": 1.13,
390
+ "step": 85
391
+ },
392
+ {
393
+ "epoch": 10.117647058823529,
394
+ "grad_norm": 0.2147984355688095,
395
+ "learning_rate": 9.602524267262203e-05,
396
+ "loss": 0.9601,
397
+ "step": 86
398
+ },
399
+ {
400
+ "epoch": 10.352941176470589,
401
+ "grad_norm": 0.21176777780056,
402
+ "learning_rate": 9.567727288213005e-05,
403
+ "loss": 0.9891,
404
+ "step": 88
405
+ },
406
+ {
407
+ "epoch": 10.588235294117647,
408
+ "grad_norm": 0.22931328415870667,
409
+ "learning_rate": 9.53153893518325e-05,
410
+ "loss": 0.9392,
411
+ "step": 90
412
+ },
413
+ {
414
+ "epoch": 10.823529411764707,
415
+ "grad_norm": 0.24265824258327484,
416
+ "learning_rate": 9.493970231495835e-05,
417
+ "loss": 0.939,
418
+ "step": 92
419
+ },
420
+ {
421
+ "epoch": 10.941176470588236,
422
+ "eval_loss": 1.1420247554779053,
423
+ "eval_runtime": 31.8533,
424
+ "eval_samples_per_second": 1.13,
425
+ "eval_steps_per_second": 1.13,
426
+ "step": 93
427
+ },
428
+ {
429
+ "epoch": 11.058823529411764,
430
+ "grad_norm": 0.24943289160728455,
431
+ "learning_rate": 9.45503262094184e-05,
432
+ "loss": 0.9836,
433
+ "step": 94
434
+ },
435
+ {
436
+ "epoch": 11.294117647058824,
437
+ "grad_norm": 0.61884605884552,
438
+ "learning_rate": 9.414737964294636e-05,
439
+ "loss": 0.9779,
440
+ "step": 96
441
+ },
442
+ {
443
+ "epoch": 11.529411764705882,
444
+ "grad_norm": 0.3478313982486725,
445
+ "learning_rate": 9.373098535696979e-05,
446
+ "loss": 0.8764,
447
+ "step": 98
448
+ },
449
+ {
450
+ "epoch": 11.764705882352942,
451
+ "grad_norm": 0.31895795464515686,
452
+ "learning_rate": 9.330127018922194e-05,
453
+ "loss": 0.9216,
454
+ "step": 100
455
+ },
456
+ {
457
+ "epoch": 12.0,
458
+ "grad_norm": 0.32360953092575073,
459
+ "learning_rate": 9.285836503510562e-05,
460
+ "loss": 0.8747,
461
+ "step": 102
462
+ },
463
+ {
464
+ "epoch": 12.0,
465
+ "eval_loss": 1.1664069890975952,
466
+ "eval_runtime": 31.8594,
467
+ "eval_samples_per_second": 1.13,
468
+ "eval_steps_per_second": 1.13,
469
+ "step": 102
470
+ },
471
+ {
472
+ "epoch": 12.235294117647058,
473
+ "grad_norm": 0.38271987438201904,
474
+ "learning_rate": 9.24024048078213e-05,
475
+ "loss": 0.8447,
476
+ "step": 104
477
+ },
478
+ {
479
+ "epoch": 12.470588235294118,
480
+ "grad_norm": 0.4208544194698334,
481
+ "learning_rate": 9.193352839727121e-05,
482
+ "loss": 0.8554,
483
+ "step": 106
484
+ },
485
+ {
486
+ "epoch": 12.705882352941176,
487
+ "grad_norm": 0.3989667296409607,
488
+ "learning_rate": 9.145187862775209e-05,
489
+ "loss": 0.8843,
490
+ "step": 108
491
+ },
492
+ {
493
+ "epoch": 12.941176470588236,
494
+ "grad_norm": 0.384284645318985,
495
+ "learning_rate": 9.09576022144496e-05,
496
+ "loss": 0.8694,
497
+ "step": 110
498
+ },
499
+ {
500
+ "epoch": 12.941176470588236,
501
+ "eval_loss": 1.1779769659042358,
502
+ "eval_runtime": 31.8607,
503
+ "eval_samples_per_second": 1.13,
504
+ "eval_steps_per_second": 1.13,
505
+ "step": 110
506
+ },
507
+ {
508
+ "epoch": 13.176470588235293,
509
+ "grad_norm": 0.375377893447876,
510
+ "learning_rate": 9.045084971874738e-05,
511
+ "loss": 0.8364,
512
+ "step": 112
513
+ },
514
+ {
515
+ "epoch": 13.411764705882353,
516
+ "grad_norm": 0.4887729585170746,
517
+ "learning_rate": 8.993177550236464e-05,
518
+ "loss": 0.7637,
519
+ "step": 114
520
+ },
521
+ {
522
+ "epoch": 13.647058823529411,
523
+ "grad_norm": 0.48660561442375183,
524
+ "learning_rate": 8.940053768033609e-05,
525
+ "loss": 0.7966,
526
+ "step": 116
527
+ },
528
+ {
529
+ "epoch": 13.882352941176471,
530
+ "grad_norm": 0.5761192440986633,
531
+ "learning_rate": 8.885729807284856e-05,
532
+ "loss": 0.8188,
533
+ "step": 118
534
+ },
535
+ {
536
+ "epoch": 14.0,
537
+ "eval_loss": 1.2246310710906982,
538
+ "eval_runtime": 31.8527,
539
+ "eval_samples_per_second": 1.13,
540
+ "eval_steps_per_second": 1.13,
541
+ "step": 119
542
+ },
543
+ {
544
+ "epoch": 14.117647058823529,
545
+ "grad_norm": 0.523510754108429,
546
+ "learning_rate": 8.83022221559489e-05,
547
+ "loss": 0.8017,
548
+ "step": 120
549
+ },
550
+ {
551
+ "epoch": 14.352941176470589,
552
+ "grad_norm": 0.6305602192878723,
553
+ "learning_rate": 8.773547901113862e-05,
554
+ "loss": 0.6666,
555
+ "step": 122
556
+ },
557
+ {
558
+ "epoch": 14.588235294117647,
559
+ "grad_norm": 0.5910943746566772,
560
+ "learning_rate": 8.715724127386972e-05,
561
+ "loss": 0.7669,
562
+ "step": 124
563
+ },
564
+ {
565
+ "epoch": 14.823529411764707,
566
+ "grad_norm": 0.6196286082267761,
567
+ "learning_rate": 8.656768508095853e-05,
568
+ "loss": 0.697,
569
+ "step": 126
570
+ },
571
+ {
572
+ "epoch": 14.941176470588236,
573
+ "eval_loss": 1.234824538230896,
574
+ "eval_runtime": 31.858,
575
+ "eval_samples_per_second": 1.13,
576
+ "eval_steps_per_second": 1.13,
577
+ "step": 127
578
+ },
579
+ {
580
+ "epoch": 15.058823529411764,
581
+ "grad_norm": 0.6877910494804382,
582
+ "learning_rate": 8.596699001693255e-05,
583
+ "loss": 0.7504,
584
+ "step": 128
585
+ },
586
+ {
587
+ "epoch": 15.294117647058824,
588
+ "grad_norm": 1.5988885164260864,
589
+ "learning_rate": 8.535533905932738e-05,
590
+ "loss": 0.6642,
591
+ "step": 130
592
+ },
593
+ {
594
+ "epoch": 15.529411764705882,
595
+ "grad_norm": 0.8035158514976501,
596
+ "learning_rate": 8.473291852294987e-05,
597
+ "loss": 0.6829,
598
+ "step": 132
599
+ },
600
+ {
601
+ "epoch": 15.764705882352942,
602
+ "grad_norm": 0.7143783569335938,
603
+ "learning_rate": 8.409991800312493e-05,
604
+ "loss": 0.6849,
605
+ "step": 134
606
+ },
607
+ {
608
+ "epoch": 16.0,
609
+ "grad_norm": 0.6947578191757202,
610
+ "learning_rate": 8.345653031794292e-05,
611
+ "loss": 0.6048,
612
+ "step": 136
613
+ },
614
+ {
615
+ "epoch": 16.0,
616
+ "eval_loss": 1.310172200202942,
617
+ "eval_runtime": 31.8619,
618
+ "eval_samples_per_second": 1.13,
619
+ "eval_steps_per_second": 1.13,
620
+ "step": 136
621
+ },
622
+ {
623
+ "epoch": 16.235294117647058,
624
+ "grad_norm": 2.3495612144470215,
625
+ "learning_rate": 8.280295144952536e-05,
626
+ "loss": 0.552,
627
+ "step": 138
628
+ },
629
+ {
630
+ "epoch": 16.470588235294116,
631
+ "grad_norm": 0.9533553719520569,
632
+ "learning_rate": 8.213938048432697e-05,
633
+ "loss": 0.6525,
634
+ "step": 140
635
+ },
636
+ {
637
+ "epoch": 16.705882352941178,
638
+ "grad_norm": 0.9292975664138794,
639
+ "learning_rate": 8.146601955249188e-05,
640
+ "loss": 0.5725,
641
+ "step": 142
642
+ },
643
+ {
644
+ "epoch": 16.941176470588236,
645
+ "grad_norm": 0.9577359557151794,
646
+ "learning_rate": 8.07830737662829e-05,
647
+ "loss": 0.5898,
648
+ "step": 144
649
+ },
650
+ {
651
+ "epoch": 16.941176470588236,
652
+ "eval_loss": 1.3189616203308105,
653
+ "eval_runtime": 31.8633,
654
+ "eval_samples_per_second": 1.13,
655
+ "eval_steps_per_second": 1.13,
656
+ "step": 144
657
+ },
658
+ {
659
+ "epoch": 16.941176470588236,
660
+ "step": 144,
661
+ "total_flos": 1.8132641623362765e+17,
662
+ "train_loss": 1.0652225611524448,
663
+ "train_runtime": 6655.309,
664
+ "train_samples_per_second": 1.022,
665
+ "train_steps_per_second": 0.06
666
+ }
667
+ ],
668
+ "logging_steps": 2,
669
+ "max_steps": 400,
670
+ "num_input_tokens_seen": 0,
671
+ "num_train_epochs": 50,
672
+ "save_steps": 25,
673
+ "stateful_callbacks": {
674
+ "EarlyStoppingCallback": {
675
+ "args": {
676
+ "early_stopping_patience": 7,
677
+ "early_stopping_threshold": 0.0
678
+ },
679
+ "attributes": {
680
+ "early_stopping_patience_counter": 0
681
+ }
682
+ },
683
+ "TrainerControl": {
684
+ "args": {
685
+ "should_epoch_stop": false,
686
+ "should_evaluate": false,
687
+ "should_log": false,
688
+ "should_save": true,
689
+ "should_training_stop": true
690
+ },
691
+ "attributes": {}
692
+ }
693
+ },
694
+ "total_flos": 1.8132641623362765e+17,
695
+ "train_batch_size": 1,
696
+ "trial_name": null,
697
+ "trial_params": null
698
+ }