GaetanMichelet commited on
Commit
635b3b4
1 Parent(s): 3732891

Model save

Browse files
README.md ADDED
@@ -0,0 +1,78 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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-1_120-samples_config-1_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-1_120-samples_config-1_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: 1.5564
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: 8
46
+ - total_train_batch_size: 8
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
+ | 2.3371 | 1.0 | 11 | 2.2386 |
57
+ | 1.7745 | 2.0 | 22 | 1.7577 |
58
+ | 1.2294 | 3.0 | 33 | 1.1298 |
59
+ | 0.9781 | 4.0 | 44 | 0.9825 |
60
+ | 0.8705 | 5.0 | 55 | 0.9252 |
61
+ | 0.8269 | 6.0 | 66 | 0.8844 |
62
+ | 0.6627 | 7.0 | 77 | 0.8656 |
63
+ | 0.6053 | 8.0 | 88 | 0.8689 |
64
+ | 0.5258 | 9.0 | 99 | 0.9307 |
65
+ | 0.3668 | 10.0 | 110 | 1.0742 |
66
+ | 0.248 | 11.0 | 121 | 1.2238 |
67
+ | 0.2039 | 12.0 | 132 | 1.3416 |
68
+ | 0.1357 | 13.0 | 143 | 1.4583 |
69
+ | 0.1016 | 14.0 | 154 | 1.5564 |
70
+
71
+
72
+ ### Framework versions
73
+
74
+ - PEFT 0.12.0
75
+ - Transformers 4.44.0
76
+ - Pytorch 2.1.2+cu121
77
+ - Datasets 2.20.0
78
+ - Tokenizers 0.19.1
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:59b0924a581198bdc86d8e559230269a45f48b874e3ad15710f98a873cd481bd
3
  size 167832240
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b39064505846f92ffab53e7f7393b44873fe10d26cc1e4c95d60c034c47bb566
3
  size 167832240
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 14.0,
3
+ "total_flos": 1.8910971254800384e+16,
4
+ "train_loss": 0.8254564341593098,
5
+ "train_runtime": 1444.0259,
6
+ "train_samples": 88,
7
+ "train_samples_per_second": 3.047,
8
+ "train_steps_per_second": 0.381
9
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 14.0,
3
+ "total_flos": 1.8910971254800384e+16,
4
+ "train_loss": 0.8254564341593098,
5
+ "train_runtime": 1444.0259,
6
+ "train_samples": 88,
7
+ "train_samples_per_second": 3.047,
8
+ "train_steps_per_second": 0.381
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,709 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.8656427264213562,
3
+ "best_model_checkpoint": "data/Llama-31-8B_task-1_120-samples_config-1_full/checkpoint-77",
4
+ "epoch": 14.0,
5
+ "eval_steps": 500,
6
+ "global_step": 154,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.09090909090909091,
13
+ "grad_norm": 2.0200283527374268,
14
+ "learning_rate": 1.818181818181818e-06,
15
+ "loss": 2.4844,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.18181818181818182,
20
+ "grad_norm": 1.8869131803512573,
21
+ "learning_rate": 3.636363636363636e-06,
22
+ "loss": 2.5082,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.36363636363636365,
27
+ "grad_norm": 1.758622169494629,
28
+ "learning_rate": 7.272727272727272e-06,
29
+ "loss": 2.5059,
30
+ "step": 4
31
+ },
32
+ {
33
+ "epoch": 0.5454545454545454,
34
+ "grad_norm": 1.6956913471221924,
35
+ "learning_rate": 1.0909090909090909e-05,
36
+ "loss": 2.4567,
37
+ "step": 6
38
+ },
39
+ {
40
+ "epoch": 0.7272727272727273,
41
+ "grad_norm": 1.5868192911148071,
42
+ "learning_rate": 1.4545454545454545e-05,
43
+ "loss": 2.3684,
44
+ "step": 8
45
+ },
46
+ {
47
+ "epoch": 0.9090909090909091,
48
+ "grad_norm": 1.787414312362671,
49
+ "learning_rate": 1.8181818181818182e-05,
50
+ "loss": 2.3371,
51
+ "step": 10
52
+ },
53
+ {
54
+ "epoch": 1.0,
55
+ "eval_loss": 2.238572835922241,
56
+ "eval_runtime": 9.6136,
57
+ "eval_samples_per_second": 2.496,
58
+ "eval_steps_per_second": 2.496,
59
+ "step": 11
60
+ },
61
+ {
62
+ "epoch": 1.0909090909090908,
63
+ "grad_norm": 1.3648227453231812,
64
+ "learning_rate": 2.1818181818181818e-05,
65
+ "loss": 2.2684,
66
+ "step": 12
67
+ },
68
+ {
69
+ "epoch": 1.2727272727272727,
70
+ "grad_norm": 1.2888587713241577,
71
+ "learning_rate": 2.5454545454545454e-05,
72
+ "loss": 2.2457,
73
+ "step": 14
74
+ },
75
+ {
76
+ "epoch": 1.4545454545454546,
77
+ "grad_norm": 1.0789052248001099,
78
+ "learning_rate": 2.909090909090909e-05,
79
+ "loss": 2.0599,
80
+ "step": 16
81
+ },
82
+ {
83
+ "epoch": 1.6363636363636362,
84
+ "grad_norm": 1.7428549528121948,
85
+ "learning_rate": 3.272727272727273e-05,
86
+ "loss": 2.0556,
87
+ "step": 18
88
+ },
89
+ {
90
+ "epoch": 1.8181818181818183,
91
+ "grad_norm": 0.905511736869812,
92
+ "learning_rate": 3.6363636363636364e-05,
93
+ "loss": 1.919,
94
+ "step": 20
95
+ },
96
+ {
97
+ "epoch": 2.0,
98
+ "grad_norm": 0.7128010988235474,
99
+ "learning_rate": 4e-05,
100
+ "loss": 1.7745,
101
+ "step": 22
102
+ },
103
+ {
104
+ "epoch": 2.0,
105
+ "eval_loss": 1.7576960325241089,
106
+ "eval_runtime": 9.615,
107
+ "eval_samples_per_second": 2.496,
108
+ "eval_steps_per_second": 2.496,
109
+ "step": 22
110
+ },
111
+ {
112
+ "epoch": 2.1818181818181817,
113
+ "grad_norm": 0.8917616605758667,
114
+ "learning_rate": 4.3636363636363636e-05,
115
+ "loss": 1.6951,
116
+ "step": 24
117
+ },
118
+ {
119
+ "epoch": 2.3636363636363638,
120
+ "grad_norm": 0.9202045798301697,
121
+ "learning_rate": 4.7272727272727275e-05,
122
+ "loss": 1.6476,
123
+ "step": 26
124
+ },
125
+ {
126
+ "epoch": 2.5454545454545454,
127
+ "grad_norm": 1.1835784912109375,
128
+ "learning_rate": 5.090909090909091e-05,
129
+ "loss": 1.4571,
130
+ "step": 28
131
+ },
132
+ {
133
+ "epoch": 2.7272727272727275,
134
+ "grad_norm": 1.0429062843322754,
135
+ "learning_rate": 5.4545454545454546e-05,
136
+ "loss": 1.3454,
137
+ "step": 30
138
+ },
139
+ {
140
+ "epoch": 2.909090909090909,
141
+ "grad_norm": 0.8713496327400208,
142
+ "learning_rate": 5.818181818181818e-05,
143
+ "loss": 1.2294,
144
+ "step": 32
145
+ },
146
+ {
147
+ "epoch": 3.0,
148
+ "eval_loss": 1.1298459768295288,
149
+ "eval_runtime": 9.6241,
150
+ "eval_samples_per_second": 2.494,
151
+ "eval_steps_per_second": 2.494,
152
+ "step": 33
153
+ },
154
+ {
155
+ "epoch": 3.090909090909091,
156
+ "grad_norm": 0.6041619777679443,
157
+ "learning_rate": 6.181818181818182e-05,
158
+ "loss": 1.1845,
159
+ "step": 34
160
+ },
161
+ {
162
+ "epoch": 3.2727272727272725,
163
+ "grad_norm": 0.5593976974487305,
164
+ "learning_rate": 6.545454545454546e-05,
165
+ "loss": 1.0476,
166
+ "step": 36
167
+ },
168
+ {
169
+ "epoch": 3.4545454545454546,
170
+ "grad_norm": 0.5526394248008728,
171
+ "learning_rate": 6.90909090909091e-05,
172
+ "loss": 0.989,
173
+ "step": 38
174
+ },
175
+ {
176
+ "epoch": 3.6363636363636362,
177
+ "grad_norm": 0.49133172631263733,
178
+ "learning_rate": 7.272727272727273e-05,
179
+ "loss": 1.0105,
180
+ "step": 40
181
+ },
182
+ {
183
+ "epoch": 3.8181818181818183,
184
+ "grad_norm": 0.5152122378349304,
185
+ "learning_rate": 7.636363636363637e-05,
186
+ "loss": 1.0186,
187
+ "step": 42
188
+ },
189
+ {
190
+ "epoch": 4.0,
191
+ "grad_norm": 0.4263210594654083,
192
+ "learning_rate": 8e-05,
193
+ "loss": 0.9781,
194
+ "step": 44
195
+ },
196
+ {
197
+ "epoch": 4.0,
198
+ "eval_loss": 0.9825031757354736,
199
+ "eval_runtime": 9.6255,
200
+ "eval_samples_per_second": 2.493,
201
+ "eval_steps_per_second": 2.493,
202
+ "step": 44
203
+ },
204
+ {
205
+ "epoch": 4.181818181818182,
206
+ "grad_norm": 0.4594288170337677,
207
+ "learning_rate": 8.363636363636364e-05,
208
+ "loss": 0.9387,
209
+ "step": 46
210
+ },
211
+ {
212
+ "epoch": 4.363636363636363,
213
+ "grad_norm": 0.5193449854850769,
214
+ "learning_rate": 8.727272727272727e-05,
215
+ "loss": 0.9481,
216
+ "step": 48
217
+ },
218
+ {
219
+ "epoch": 4.545454545454545,
220
+ "grad_norm": 0.5140848159790039,
221
+ "learning_rate": 9.090909090909092e-05,
222
+ "loss": 0.932,
223
+ "step": 50
224
+ },
225
+ {
226
+ "epoch": 4.7272727272727275,
227
+ "grad_norm": 0.4637601673603058,
228
+ "learning_rate": 9.454545454545455e-05,
229
+ "loss": 0.9293,
230
+ "step": 52
231
+ },
232
+ {
233
+ "epoch": 4.909090909090909,
234
+ "grad_norm": 0.5280166864395142,
235
+ "learning_rate": 9.818181818181818e-05,
236
+ "loss": 0.8705,
237
+ "step": 54
238
+ },
239
+ {
240
+ "epoch": 5.0,
241
+ "eval_loss": 0.9252349734306335,
242
+ "eval_runtime": 9.6225,
243
+ "eval_samples_per_second": 2.494,
244
+ "eval_steps_per_second": 2.494,
245
+ "step": 55
246
+ },
247
+ {
248
+ "epoch": 5.090909090909091,
249
+ "grad_norm": 0.5434139370918274,
250
+ "learning_rate": 9.999899300364532e-05,
251
+ "loss": 0.8527,
252
+ "step": 56
253
+ },
254
+ {
255
+ "epoch": 5.2727272727272725,
256
+ "grad_norm": 0.5284571647644043,
257
+ "learning_rate": 9.99909372761763e-05,
258
+ "loss": 0.8431,
259
+ "step": 58
260
+ },
261
+ {
262
+ "epoch": 5.454545454545454,
263
+ "grad_norm": 0.5927901268005371,
264
+ "learning_rate": 9.997482711915927e-05,
265
+ "loss": 0.7975,
266
+ "step": 60
267
+ },
268
+ {
269
+ "epoch": 5.636363636363637,
270
+ "grad_norm": 0.622176468372345,
271
+ "learning_rate": 9.99506651282272e-05,
272
+ "loss": 0.8844,
273
+ "step": 62
274
+ },
275
+ {
276
+ "epoch": 5.818181818181818,
277
+ "grad_norm": 0.7427988648414612,
278
+ "learning_rate": 9.991845519630678e-05,
279
+ "loss": 0.7823,
280
+ "step": 64
281
+ },
282
+ {
283
+ "epoch": 6.0,
284
+ "grad_norm": 0.7913317084312439,
285
+ "learning_rate": 9.987820251299122e-05,
286
+ "loss": 0.8269,
287
+ "step": 66
288
+ },
289
+ {
290
+ "epoch": 6.0,
291
+ "eval_loss": 0.884376049041748,
292
+ "eval_runtime": 9.6171,
293
+ "eval_samples_per_second": 2.496,
294
+ "eval_steps_per_second": 2.496,
295
+ "step": 66
296
+ },
297
+ {
298
+ "epoch": 6.181818181818182,
299
+ "grad_norm": 0.7040154337882996,
300
+ "learning_rate": 9.982991356370404e-05,
301
+ "loss": 0.7398,
302
+ "step": 68
303
+ },
304
+ {
305
+ "epoch": 6.363636363636363,
306
+ "grad_norm": 0.7483051419258118,
307
+ "learning_rate": 9.977359612865423e-05,
308
+ "loss": 0.7724,
309
+ "step": 70
310
+ },
311
+ {
312
+ "epoch": 6.545454545454545,
313
+ "grad_norm": 0.5491539239883423,
314
+ "learning_rate": 9.970925928158274e-05,
315
+ "loss": 0.7608,
316
+ "step": 72
317
+ },
318
+ {
319
+ "epoch": 6.7272727272727275,
320
+ "grad_norm": 0.6535021066665649,
321
+ "learning_rate": 9.963691338830044e-05,
322
+ "loss": 0.7548,
323
+ "step": 74
324
+ },
325
+ {
326
+ "epoch": 6.909090909090909,
327
+ "grad_norm": 0.686294436454773,
328
+ "learning_rate": 9.955657010501806e-05,
329
+ "loss": 0.6627,
330
+ "step": 76
331
+ },
332
+ {
333
+ "epoch": 7.0,
334
+ "eval_loss": 0.8656427264213562,
335
+ "eval_runtime": 9.62,
336
+ "eval_samples_per_second": 2.495,
337
+ "eval_steps_per_second": 2.495,
338
+ "step": 77
339
+ },
340
+ {
341
+ "epoch": 7.090909090909091,
342
+ "grad_norm": 0.6294690370559692,
343
+ "learning_rate": 9.946824237646824e-05,
344
+ "loss": 0.6067,
345
+ "step": 78
346
+ },
347
+ {
348
+ "epoch": 7.2727272727272725,
349
+ "grad_norm": 0.732207179069519,
350
+ "learning_rate": 9.937194443381972e-05,
351
+ "loss": 0.6866,
352
+ "step": 80
353
+ },
354
+ {
355
+ "epoch": 7.454545454545454,
356
+ "grad_norm": 0.6583709716796875,
357
+ "learning_rate": 9.926769179238466e-05,
358
+ "loss": 0.5795,
359
+ "step": 82
360
+ },
361
+ {
362
+ "epoch": 7.636363636363637,
363
+ "grad_norm": 0.6547601222991943,
364
+ "learning_rate": 9.915550124911866e-05,
365
+ "loss": 0.5979,
366
+ "step": 84
367
+ },
368
+ {
369
+ "epoch": 7.818181818181818,
370
+ "grad_norm": 0.6767302751541138,
371
+ "learning_rate": 9.903539087991462e-05,
372
+ "loss": 0.6686,
373
+ "step": 86
374
+ },
375
+ {
376
+ "epoch": 8.0,
377
+ "grad_norm": 0.8115944862365723,
378
+ "learning_rate": 9.890738003669029e-05,
379
+ "loss": 0.6053,
380
+ "step": 88
381
+ },
382
+ {
383
+ "epoch": 8.0,
384
+ "eval_loss": 0.868882417678833,
385
+ "eval_runtime": 9.6154,
386
+ "eval_samples_per_second": 2.496,
387
+ "eval_steps_per_second": 2.496,
388
+ "step": 88
389
+ },
390
+ {
391
+ "epoch": 8.181818181818182,
392
+ "grad_norm": 0.7251437902450562,
393
+ "learning_rate": 9.877148934427037e-05,
394
+ "loss": 0.5444,
395
+ "step": 90
396
+ },
397
+ {
398
+ "epoch": 8.363636363636363,
399
+ "grad_norm": 0.9013553261756897,
400
+ "learning_rate": 9.862774069706346e-05,
401
+ "loss": 0.5037,
402
+ "step": 92
403
+ },
404
+ {
405
+ "epoch": 8.545454545454545,
406
+ "grad_norm": 0.9543907046318054,
407
+ "learning_rate": 9.847615725553456e-05,
408
+ "loss": 0.4884,
409
+ "step": 94
410
+ },
411
+ {
412
+ "epoch": 8.727272727272727,
413
+ "grad_norm": 0.8602244853973389,
414
+ "learning_rate": 9.831676344247342e-05,
415
+ "loss": 0.4928,
416
+ "step": 96
417
+ },
418
+ {
419
+ "epoch": 8.909090909090908,
420
+ "grad_norm": 0.9116283059120178,
421
+ "learning_rate": 9.814958493905963e-05,
422
+ "loss": 0.5258,
423
+ "step": 98
424
+ },
425
+ {
426
+ "epoch": 9.0,
427
+ "eval_loss": 0.930729866027832,
428
+ "eval_runtime": 9.6169,
429
+ "eval_samples_per_second": 2.496,
430
+ "eval_steps_per_second": 2.496,
431
+ "step": 99
432
+ },
433
+ {
434
+ "epoch": 9.090909090909092,
435
+ "grad_norm": 0.8587744832038879,
436
+ "learning_rate": 9.797464868072488e-05,
437
+ "loss": 0.4467,
438
+ "step": 100
439
+ },
440
+ {
441
+ "epoch": 9.272727272727273,
442
+ "grad_norm": 1.011325478553772,
443
+ "learning_rate": 9.779198285281325e-05,
444
+ "loss": 0.3718,
445
+ "step": 102
446
+ },
447
+ {
448
+ "epoch": 9.454545454545455,
449
+ "grad_norm": 1.561611294746399,
450
+ "learning_rate": 9.760161688604008e-05,
451
+ "loss": 0.411,
452
+ "step": 104
453
+ },
454
+ {
455
+ "epoch": 9.636363636363637,
456
+ "grad_norm": 1.0310885906219482,
457
+ "learning_rate": 9.740358145174998e-05,
458
+ "loss": 0.3617,
459
+ "step": 106
460
+ },
461
+ {
462
+ "epoch": 9.818181818181818,
463
+ "grad_norm": 0.9703785181045532,
464
+ "learning_rate": 9.719790845697533e-05,
465
+ "loss": 0.3618,
466
+ "step": 108
467
+ },
468
+ {
469
+ "epoch": 10.0,
470
+ "grad_norm": 1.0072264671325684,
471
+ "learning_rate": 9.698463103929542e-05,
472
+ "loss": 0.3668,
473
+ "step": 110
474
+ },
475
+ {
476
+ "epoch": 10.0,
477
+ "eval_loss": 1.0742104053497314,
478
+ "eval_runtime": 9.6189,
479
+ "eval_samples_per_second": 2.495,
480
+ "eval_steps_per_second": 2.495,
481
+ "step": 110
482
+ },
483
+ {
484
+ "epoch": 10.181818181818182,
485
+ "grad_norm": 1.1375013589859009,
486
+ "learning_rate": 9.676378356149734e-05,
487
+ "loss": 0.2664,
488
+ "step": 112
489
+ },
490
+ {
491
+ "epoch": 10.363636363636363,
492
+ "grad_norm": 1.377677083015442,
493
+ "learning_rate": 9.653540160603956e-05,
494
+ "loss": 0.2831,
495
+ "step": 114
496
+ },
497
+ {
498
+ "epoch": 10.545454545454545,
499
+ "grad_norm": 1.225994348526001,
500
+ "learning_rate": 9.629952196931901e-05,
501
+ "loss": 0.2456,
502
+ "step": 116
503
+ },
504
+ {
505
+ "epoch": 10.727272727272727,
506
+ "grad_norm": 1.329970359802246,
507
+ "learning_rate": 9.60561826557425e-05,
508
+ "loss": 0.2814,
509
+ "step": 118
510
+ },
511
+ {
512
+ "epoch": 10.909090909090908,
513
+ "grad_norm": 1.3457225561141968,
514
+ "learning_rate": 9.580542287160348e-05,
515
+ "loss": 0.248,
516
+ "step": 120
517
+ },
518
+ {
519
+ "epoch": 11.0,
520
+ "eval_loss": 1.2238476276397705,
521
+ "eval_runtime": 9.624,
522
+ "eval_samples_per_second": 2.494,
523
+ "eval_steps_per_second": 2.494,
524
+ "step": 121
525
+ },
526
+ {
527
+ "epoch": 11.090909090909092,
528
+ "grad_norm": 1.0499111413955688,
529
+ "learning_rate": 9.554728301876526e-05,
530
+ "loss": 0.1828,
531
+ "step": 122
532
+ },
533
+ {
534
+ "epoch": 11.272727272727273,
535
+ "grad_norm": 1.2574657201766968,
536
+ "learning_rate": 9.528180468815155e-05,
537
+ "loss": 0.1455,
538
+ "step": 124
539
+ },
540
+ {
541
+ "epoch": 11.454545454545455,
542
+ "grad_norm": 1.8661640882492065,
543
+ "learning_rate": 9.50090306530454e-05,
544
+ "loss": 0.1843,
545
+ "step": 126
546
+ },
547
+ {
548
+ "epoch": 11.636363636363637,
549
+ "grad_norm": 1.4496153593063354,
550
+ "learning_rate": 9.472900486219769e-05,
551
+ "loss": 0.181,
552
+ "step": 128
553
+ },
554
+ {
555
+ "epoch": 11.818181818181818,
556
+ "grad_norm": 1.520456075668335,
557
+ "learning_rate": 9.444177243274618e-05,
558
+ "loss": 0.1556,
559
+ "step": 130
560
+ },
561
+ {
562
+ "epoch": 12.0,
563
+ "grad_norm": 1.3242138624191284,
564
+ "learning_rate": 9.414737964294636e-05,
565
+ "loss": 0.2039,
566
+ "step": 132
567
+ },
568
+ {
569
+ "epoch": 12.0,
570
+ "eval_loss": 1.3416190147399902,
571
+ "eval_runtime": 9.6192,
572
+ "eval_samples_per_second": 2.495,
573
+ "eval_steps_per_second": 2.495,
574
+ "step": 132
575
+ },
576
+ {
577
+ "epoch": 12.181818181818182,
578
+ "grad_norm": 1.0514355897903442,
579
+ "learning_rate": 9.384587392471515e-05,
580
+ "loss": 0.1147,
581
+ "step": 134
582
+ },
583
+ {
584
+ "epoch": 12.363636363636363,
585
+ "grad_norm": 1.0401465892791748,
586
+ "learning_rate": 9.353730385598887e-05,
587
+ "loss": 0.1134,
588
+ "step": 136
589
+ },
590
+ {
591
+ "epoch": 12.545454545454545,
592
+ "grad_norm": 1.2559245824813843,
593
+ "learning_rate": 9.322171915289635e-05,
594
+ "loss": 0.1241,
595
+ "step": 138
596
+ },
597
+ {
598
+ "epoch": 12.727272727272727,
599
+ "grad_norm": 1.3274048566818237,
600
+ "learning_rate": 9.289917066174886e-05,
601
+ "loss": 0.1263,
602
+ "step": 140
603
+ },
604
+ {
605
+ "epoch": 12.909090909090908,
606
+ "grad_norm": 1.2401025295257568,
607
+ "learning_rate": 9.256971035084785e-05,
608
+ "loss": 0.1357,
609
+ "step": 142
610
+ },
611
+ {
612
+ "epoch": 13.0,
613
+ "eval_loss": 1.4582844972610474,
614
+ "eval_runtime": 9.6219,
615
+ "eval_samples_per_second": 2.494,
616
+ "eval_steps_per_second": 2.494,
617
+ "step": 143
618
+ },
619
+ {
620
+ "epoch": 13.090909090909092,
621
+ "grad_norm": 1.0610545873641968,
622
+ "learning_rate": 9.223339130211192e-05,
623
+ "loss": 0.0999,
624
+ "step": 144
625
+ },
626
+ {
627
+ "epoch": 13.272727272727273,
628
+ "grad_norm": 1.0816364288330078,
629
+ "learning_rate": 9.189026770252436e-05,
630
+ "loss": 0.0877,
631
+ "step": 146
632
+ },
633
+ {
634
+ "epoch": 13.454545454545455,
635
+ "grad_norm": 0.9884896874427795,
636
+ "learning_rate": 9.154039483540273e-05,
637
+ "loss": 0.0952,
638
+ "step": 148
639
+ },
640
+ {
641
+ "epoch": 13.636363636363637,
642
+ "grad_norm": 1.4226878881454468,
643
+ "learning_rate": 9.118382907149165e-05,
644
+ "loss": 0.0921,
645
+ "step": 150
646
+ },
647
+ {
648
+ "epoch": 13.818181818181818,
649
+ "grad_norm": 1.2806740999221802,
650
+ "learning_rate": 9.082062785988049e-05,
651
+ "loss": 0.0858,
652
+ "step": 152
653
+ },
654
+ {
655
+ "epoch": 14.0,
656
+ "grad_norm": 1.4139348268508911,
657
+ "learning_rate": 9.045084971874738e-05,
658
+ "loss": 0.1016,
659
+ "step": 154
660
+ },
661
+ {
662
+ "epoch": 14.0,
663
+ "eval_loss": 1.5564292669296265,
664
+ "eval_runtime": 9.6208,
665
+ "eval_samples_per_second": 2.495,
666
+ "eval_steps_per_second": 2.495,
667
+ "step": 154
668
+ },
669
+ {
670
+ "epoch": 14.0,
671
+ "step": 154,
672
+ "total_flos": 1.8910971254800384e+16,
673
+ "train_loss": 0.8254564341593098,
674
+ "train_runtime": 1444.0259,
675
+ "train_samples_per_second": 3.047,
676
+ "train_steps_per_second": 0.381
677
+ }
678
+ ],
679
+ "logging_steps": 2,
680
+ "max_steps": 550,
681
+ "num_input_tokens_seen": 0,
682
+ "num_train_epochs": 50,
683
+ "save_steps": 25,
684
+ "stateful_callbacks": {
685
+ "EarlyStoppingCallback": {
686
+ "args": {
687
+ "early_stopping_patience": 7,
688
+ "early_stopping_threshold": 0.0
689
+ },
690
+ "attributes": {
691
+ "early_stopping_patience_counter": 0
692
+ }
693
+ },
694
+ "TrainerControl": {
695
+ "args": {
696
+ "should_epoch_stop": false,
697
+ "should_evaluate": false,
698
+ "should_log": false,
699
+ "should_save": true,
700
+ "should_training_stop": true
701
+ },
702
+ "attributes": {}
703
+ }
704
+ },
705
+ "total_flos": 1.8910971254800384e+16,
706
+ "train_batch_size": 1,
707
+ "trial_name": null,
708
+ "trial_params": null
709
+ }