nthakur commited on
Commit
d294fdb
·
verified ·
1 Parent(s): 98ec7f7

Model save

Browse files
Files changed (4) hide show
  1. README.md +69 -0
  2. all_results.json +9 -0
  3. train_results.json +9 -0
  4. trainer_state.json +1011 -0
README.md ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: llama3
3
+ library_name: peft
4
+ tags:
5
+ - trl
6
+ - sft
7
+ - generated_from_trainer
8
+ base_model: meta-llama/Meta-Llama-3-8B-Instruct
9
+ model-index:
10
+ - name: Meta-Llama-3-8B-Instruct-nomiracl-sft
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
+ # Meta-Llama-3-8B-Instruct-nomiracl-sft
18
+
19
+ This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on an unknown dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 1.6358
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: 1e-05
41
+ - train_batch_size: 4
42
+ - eval_batch_size: 4
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - num_devices: 4
46
+ - gradient_accumulation_steps: 2
47
+ - total_train_batch_size: 32
48
+ - total_eval_batch_size: 16
49
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
50
+ - lr_scheduler_type: cosine
51
+ - lr_scheduler_warmup_ratio: 0.1
52
+ - num_epochs: 1
53
+
54
+ ### Training results
55
+
56
+ | Training Loss | Epoch | Step | Validation Loss |
57
+ |:-------------:|:------:|:----:|:---------------:|
58
+ | 1.6576 | 0.2981 | 200 | 1.6656 |
59
+ | 1.6447 | 0.5961 | 400 | 1.6409 |
60
+ | 1.6245 | 0.8942 | 600 | 1.6358 |
61
+
62
+
63
+ ### Framework versions
64
+
65
+ - PEFT 0.7.1
66
+ - Transformers 4.41.2
67
+ - Pytorch 2.3.0+cu121
68
+ - Datasets 2.20.0
69
+ - Tokenizers 0.19.1
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "total_flos": 919616052789248.0,
4
+ "train_loss": 1.6831368206154926,
5
+ "train_runtime": 18035.2779,
6
+ "train_samples": 21471,
7
+ "train_samples_per_second": 1.191,
8
+ "train_steps_per_second": 0.037
9
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "total_flos": 919616052789248.0,
4
+ "train_loss": 1.6831368206154926,
5
+ "train_runtime": 18035.2779,
6
+ "train_samples": 21471,
7
+ "train_samples_per_second": 1.191,
8
+ "train_steps_per_second": 0.037
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,1011 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.6357628107070923,
3
+ "best_model_checkpoint": "/mnt/users/n3thakur/research_new/miracl-unanswerable/models/Meta-Llama-3-8B-Instruct-nomiracl-sft/checkpoint-600",
4
+ "epoch": 1.0,
5
+ "eval_steps": 200,
6
+ "global_step": 671,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0014903129657228018,
13
+ "grad_norm": 1.2864242608707517,
14
+ "learning_rate": 1.4705882352941178e-07,
15
+ "loss": 1.9819,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.007451564828614009,
20
+ "grad_norm": 1.3601881935430797,
21
+ "learning_rate": 7.352941176470589e-07,
22
+ "loss": 2.0209,
23
+ "step": 5
24
+ },
25
+ {
26
+ "epoch": 0.014903129657228018,
27
+ "grad_norm": 1.308324783880395,
28
+ "learning_rate": 1.4705882352941177e-06,
29
+ "loss": 2.0789,
30
+ "step": 10
31
+ },
32
+ {
33
+ "epoch": 0.022354694485842028,
34
+ "grad_norm": 1.3402040033505984,
35
+ "learning_rate": 2.2058823529411767e-06,
36
+ "loss": 2.1346,
37
+ "step": 15
38
+ },
39
+ {
40
+ "epoch": 0.029806259314456036,
41
+ "grad_norm": 1.252511058906224,
42
+ "learning_rate": 2.9411764705882355e-06,
43
+ "loss": 1.9887,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.037257824143070044,
48
+ "grad_norm": 1.1571538128626155,
49
+ "learning_rate": 3.6764705882352946e-06,
50
+ "loss": 1.9678,
51
+ "step": 25
52
+ },
53
+ {
54
+ "epoch": 0.044709388971684055,
55
+ "grad_norm": 0.9688203213594095,
56
+ "learning_rate": 4.411764705882353e-06,
57
+ "loss": 1.9569,
58
+ "step": 30
59
+ },
60
+ {
61
+ "epoch": 0.05216095380029806,
62
+ "grad_norm": 0.9113616550249095,
63
+ "learning_rate": 5.147058823529411e-06,
64
+ "loss": 1.9236,
65
+ "step": 35
66
+ },
67
+ {
68
+ "epoch": 0.05961251862891207,
69
+ "grad_norm": 0.8080169675485035,
70
+ "learning_rate": 5.882352941176471e-06,
71
+ "loss": 1.9358,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 0.06706408345752608,
76
+ "grad_norm": 0.6008229695631025,
77
+ "learning_rate": 6.61764705882353e-06,
78
+ "loss": 1.8434,
79
+ "step": 45
80
+ },
81
+ {
82
+ "epoch": 0.07451564828614009,
83
+ "grad_norm": 0.5413811186895102,
84
+ "learning_rate": 7.352941176470589e-06,
85
+ "loss": 1.8786,
86
+ "step": 50
87
+ },
88
+ {
89
+ "epoch": 0.08196721311475409,
90
+ "grad_norm": 0.4866905959375867,
91
+ "learning_rate": 8.088235294117648e-06,
92
+ "loss": 1.7887,
93
+ "step": 55
94
+ },
95
+ {
96
+ "epoch": 0.08941877794336811,
97
+ "grad_norm": 0.3991956099552151,
98
+ "learning_rate": 8.823529411764707e-06,
99
+ "loss": 1.8388,
100
+ "step": 60
101
+ },
102
+ {
103
+ "epoch": 0.09687034277198212,
104
+ "grad_norm": 0.4089753043909763,
105
+ "learning_rate": 9.558823529411766e-06,
106
+ "loss": 1.7719,
107
+ "step": 65
108
+ },
109
+ {
110
+ "epoch": 0.10432190760059612,
111
+ "grad_norm": 0.3567858319466126,
112
+ "learning_rate": 9.999728567909404e-06,
113
+ "loss": 1.7757,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.11177347242921014,
118
+ "grad_norm": 0.3733731735267369,
119
+ "learning_rate": 9.996675295326345e-06,
120
+ "loss": 1.7357,
121
+ "step": 75
122
+ },
123
+ {
124
+ "epoch": 0.11922503725782414,
125
+ "grad_norm": 0.4120045550242149,
126
+ "learning_rate": 9.99023153874608e-06,
127
+ "loss": 1.76,
128
+ "step": 80
129
+ },
130
+ {
131
+ "epoch": 0.12667660208643816,
132
+ "grad_norm": 0.47838141094777215,
133
+ "learning_rate": 9.980401670566705e-06,
134
+ "loss": 1.8124,
135
+ "step": 85
136
+ },
137
+ {
138
+ "epoch": 0.13412816691505217,
139
+ "grad_norm": 0.34098601847525956,
140
+ "learning_rate": 9.967192360825558e-06,
141
+ "loss": 1.7454,
142
+ "step": 90
143
+ },
144
+ {
145
+ "epoch": 0.14157973174366617,
146
+ "grad_norm": 0.3091360382279767,
147
+ "learning_rate": 9.950612572673255e-06,
148
+ "loss": 1.7528,
149
+ "step": 95
150
+ },
151
+ {
152
+ "epoch": 0.14903129657228018,
153
+ "grad_norm": 0.2937982462814676,
154
+ "learning_rate": 9.930673556291789e-06,
155
+ "loss": 1.6999,
156
+ "step": 100
157
+ },
158
+ {
159
+ "epoch": 0.15648286140089418,
160
+ "grad_norm": 0.3033053874366656,
161
+ "learning_rate": 9.907388841260723e-06,
162
+ "loss": 1.6328,
163
+ "step": 105
164
+ },
165
+ {
166
+ "epoch": 0.16393442622950818,
167
+ "grad_norm": 0.3147533435412934,
168
+ "learning_rate": 9.880774227376727e-06,
169
+ "loss": 1.7202,
170
+ "step": 110
171
+ },
172
+ {
173
+ "epoch": 0.17138599105812222,
174
+ "grad_norm": 0.30138135147702194,
175
+ "learning_rate": 9.850847773932656e-06,
176
+ "loss": 1.7147,
177
+ "step": 115
178
+ },
179
+ {
180
+ "epoch": 0.17883755588673622,
181
+ "grad_norm": 0.2821814099944737,
182
+ "learning_rate": 9.817629787463456e-06,
183
+ "loss": 1.7141,
184
+ "step": 120
185
+ },
186
+ {
187
+ "epoch": 0.18628912071535023,
188
+ "grad_norm": 0.278421347158755,
189
+ "learning_rate": 9.781142807967205e-06,
190
+ "loss": 1.7812,
191
+ "step": 125
192
+ },
193
+ {
194
+ "epoch": 0.19374068554396423,
195
+ "grad_norm": 0.2745120712204819,
196
+ "learning_rate": 9.741411593610635e-06,
197
+ "loss": 1.6429,
198
+ "step": 130
199
+ },
200
+ {
201
+ "epoch": 0.20119225037257824,
202
+ "grad_norm": 0.2915048113702178,
203
+ "learning_rate": 9.698463103929542e-06,
204
+ "loss": 1.6622,
205
+ "step": 135
206
+ },
207
+ {
208
+ "epoch": 0.20864381520119224,
209
+ "grad_norm": 0.29052352322277714,
210
+ "learning_rate": 9.652326481535434e-06,
211
+ "loss": 1.7323,
212
+ "step": 140
213
+ },
214
+ {
215
+ "epoch": 0.21609538002980627,
216
+ "grad_norm": 0.30795018674862495,
217
+ "learning_rate": 9.603033032340875e-06,
218
+ "loss": 1.7064,
219
+ "step": 145
220
+ },
221
+ {
222
+ "epoch": 0.22354694485842028,
223
+ "grad_norm": 0.27173782182479533,
224
+ "learning_rate": 9.550616204316922e-06,
225
+ "loss": 1.7384,
226
+ "step": 150
227
+ },
228
+ {
229
+ "epoch": 0.23099850968703428,
230
+ "grad_norm": 0.3106086974094058,
231
+ "learning_rate": 9.495111564797073e-06,
232
+ "loss": 1.7267,
233
+ "step": 155
234
+ },
235
+ {
236
+ "epoch": 0.23845007451564829,
237
+ "grad_norm": 0.3021509489363488,
238
+ "learning_rate": 9.43655677634312e-06,
239
+ "loss": 1.6479,
240
+ "step": 160
241
+ },
242
+ {
243
+ "epoch": 0.2459016393442623,
244
+ "grad_norm": 0.29504453527495433,
245
+ "learning_rate": 9.374991571189292e-06,
246
+ "loss": 1.6919,
247
+ "step": 165
248
+ },
249
+ {
250
+ "epoch": 0.2533532041728763,
251
+ "grad_norm": 0.3337725680370752,
252
+ "learning_rate": 9.310457724282034e-06,
253
+ "loss": 1.7475,
254
+ "step": 170
255
+ },
256
+ {
257
+ "epoch": 0.2608047690014903,
258
+ "grad_norm": 0.271616871396071,
259
+ "learning_rate": 9.242999024933694e-06,
260
+ "loss": 1.6249,
261
+ "step": 175
262
+ },
263
+ {
264
+ "epoch": 0.26825633383010433,
265
+ "grad_norm": 0.27469314451092647,
266
+ "learning_rate": 9.172661247109382e-06,
267
+ "loss": 1.6289,
268
+ "step": 180
269
+ },
270
+ {
271
+ "epoch": 0.2757078986587183,
272
+ "grad_norm": 0.32247904554795664,
273
+ "learning_rate": 9.099492118367123e-06,
274
+ "loss": 1.6759,
275
+ "step": 185
276
+ },
277
+ {
278
+ "epoch": 0.28315946348733234,
279
+ "grad_norm": 0.32346400909431516,
280
+ "learning_rate": 9.023541287472435e-06,
281
+ "loss": 1.7517,
282
+ "step": 190
283
+ },
284
+ {
285
+ "epoch": 0.2906110283159464,
286
+ "grad_norm": 0.28010968189921137,
287
+ "learning_rate": 8.944860290709245e-06,
288
+ "loss": 1.6495,
289
+ "step": 195
290
+ },
291
+ {
292
+ "epoch": 0.29806259314456035,
293
+ "grad_norm": 0.3073056244750479,
294
+ "learning_rate": 8.863502516910058e-06,
295
+ "loss": 1.6576,
296
+ "step": 200
297
+ },
298
+ {
299
+ "epoch": 0.29806259314456035,
300
+ "eval_loss": 1.6655718088150024,
301
+ "eval_runtime": 581.6685,
302
+ "eval_samples_per_second": 4.102,
303
+ "eval_steps_per_second": 0.258,
304
+ "step": 200
305
+ },
306
+ {
307
+ "epoch": 0.3055141579731744,
308
+ "grad_norm": 0.2982943497092564,
309
+ "learning_rate": 8.779523171229061e-06,
310
+ "loss": 1.6927,
311
+ "step": 205
312
+ },
313
+ {
314
+ "epoch": 0.31296572280178836,
315
+ "grad_norm": 0.3097195272484475,
316
+ "learning_rate": 8.692979237682785e-06,
317
+ "loss": 1.6619,
318
+ "step": 210
319
+ },
320
+ {
321
+ "epoch": 0.3204172876304024,
322
+ "grad_norm": 0.34367552210427565,
323
+ "learning_rate": 8.603929440483714e-06,
324
+ "loss": 1.684,
325
+ "step": 215
326
+ },
327
+ {
328
+ "epoch": 0.32786885245901637,
329
+ "grad_norm": 0.3023959348087803,
330
+ "learning_rate": 8.51243420419308e-06,
331
+ "loss": 1.5742,
332
+ "step": 220
333
+ },
334
+ {
335
+ "epoch": 0.3353204172876304,
336
+ "grad_norm": 0.2908694884413669,
337
+ "learning_rate": 8.418555612719912e-06,
338
+ "loss": 1.6211,
339
+ "step": 225
340
+ },
341
+ {
342
+ "epoch": 0.34277198211624443,
343
+ "grad_norm": 0.3265065297068578,
344
+ "learning_rate": 8.32235736719411e-06,
345
+ "loss": 1.6614,
346
+ "step": 230
347
+ },
348
+ {
349
+ "epoch": 0.3502235469448584,
350
+ "grad_norm": 0.29528676002450155,
351
+ "learning_rate": 8.223904742742182e-06,
352
+ "loss": 1.6757,
353
+ "step": 235
354
+ },
355
+ {
356
+ "epoch": 0.35767511177347244,
357
+ "grad_norm": 0.30070470151226314,
358
+ "learning_rate": 8.123264544194934e-06,
359
+ "loss": 1.6443,
360
+ "step": 240
361
+ },
362
+ {
363
+ "epoch": 0.3651266766020864,
364
+ "grad_norm": 0.3090720148911324,
365
+ "learning_rate": 8.02050506075718e-06,
366
+ "loss": 1.6689,
367
+ "step": 245
368
+ },
369
+ {
370
+ "epoch": 0.37257824143070045,
371
+ "grad_norm": 0.29743017001679256,
372
+ "learning_rate": 7.91569601967025e-06,
373
+ "loss": 1.6278,
374
+ "step": 250
375
+ },
376
+ {
377
+ "epoch": 0.38002980625931443,
378
+ "grad_norm": 0.297851200782397,
379
+ "learning_rate": 7.808908538898703e-06,
380
+ "loss": 1.6155,
381
+ "step": 255
382
+ },
383
+ {
384
+ "epoch": 0.38748137108792846,
385
+ "grad_norm": 0.3435439482120138,
386
+ "learning_rate": 7.70021507887338e-06,
387
+ "loss": 1.5611,
388
+ "step": 260
389
+ },
390
+ {
391
+ "epoch": 0.3949329359165425,
392
+ "grad_norm": 0.32733272749783104,
393
+ "learning_rate": 7.5896893933235135e-06,
394
+ "loss": 1.6055,
395
+ "step": 265
396
+ },
397
+ {
398
+ "epoch": 0.40238450074515647,
399
+ "grad_norm": 0.3056544624446995,
400
+ "learning_rate": 7.477406479231299e-06,
401
+ "loss": 1.678,
402
+ "step": 270
403
+ },
404
+ {
405
+ "epoch": 0.4098360655737705,
406
+ "grad_norm": 0.31368790810083264,
407
+ "learning_rate": 7.363442525942827e-06,
408
+ "loss": 1.5912,
409
+ "step": 275
410
+ },
411
+ {
412
+ "epoch": 0.4172876304023845,
413
+ "grad_norm": 0.2993808857251287,
414
+ "learning_rate": 7.247874863469964e-06,
415
+ "loss": 1.6412,
416
+ "step": 280
417
+ },
418
+ {
419
+ "epoch": 0.4247391952309985,
420
+ "grad_norm": 0.32056752037833786,
421
+ "learning_rate": 7.1307819100182275e-06,
422
+ "loss": 1.6949,
423
+ "step": 285
424
+ },
425
+ {
426
+ "epoch": 0.43219076005961254,
427
+ "grad_norm": 0.3361900057942298,
428
+ "learning_rate": 7.01224311877627e-06,
429
+ "loss": 1.6513,
430
+ "step": 290
431
+ },
432
+ {
433
+ "epoch": 0.4396423248882265,
434
+ "grad_norm": 0.3373143541448011,
435
+ "learning_rate": 6.892338924003068e-06,
436
+ "loss": 1.6699,
437
+ "step": 295
438
+ },
439
+ {
440
+ "epoch": 0.44709388971684055,
441
+ "grad_norm": 0.30913581638324406,
442
+ "learning_rate": 6.771150686449436e-06,
443
+ "loss": 1.5889,
444
+ "step": 300
445
+ },
446
+ {
447
+ "epoch": 0.45454545454545453,
448
+ "grad_norm": 0.3324140439223907,
449
+ "learning_rate": 6.648760638150833e-06,
450
+ "loss": 1.6559,
451
+ "step": 305
452
+ },
453
+ {
454
+ "epoch": 0.46199701937406856,
455
+ "grad_norm": 0.3143801425022107,
456
+ "learning_rate": 6.525251826628991e-06,
457
+ "loss": 1.6297,
458
+ "step": 310
459
+ },
460
+ {
461
+ "epoch": 0.46944858420268254,
462
+ "grad_norm": 0.32690734523122295,
463
+ "learning_rate": 6.400708058540182e-06,
464
+ "loss": 1.6407,
465
+ "step": 315
466
+ },
467
+ {
468
+ "epoch": 0.47690014903129657,
469
+ "grad_norm": 0.31268911619063705,
470
+ "learning_rate": 6.275213842808383e-06,
471
+ "loss": 1.6245,
472
+ "step": 320
473
+ },
474
+ {
475
+ "epoch": 0.4843517138599106,
476
+ "grad_norm": 0.36149919350714527,
477
+ "learning_rate": 6.148854333281905e-06,
478
+ "loss": 1.6476,
479
+ "step": 325
480
+ },
481
+ {
482
+ "epoch": 0.4918032786885246,
483
+ "grad_norm": 0.3590585354206088,
484
+ "learning_rate": 6.021715270952435e-06,
485
+ "loss": 1.6452,
486
+ "step": 330
487
+ },
488
+ {
489
+ "epoch": 0.4992548435171386,
490
+ "grad_norm": 0.3039096750446457,
491
+ "learning_rate": 5.893882925775648e-06,
492
+ "loss": 1.6415,
493
+ "step": 335
494
+ },
495
+ {
496
+ "epoch": 0.5067064083457526,
497
+ "grad_norm": 0.31141048527235143,
498
+ "learning_rate": 5.765444038132901e-06,
499
+ "loss": 1.6571,
500
+ "step": 340
501
+ },
502
+ {
503
+ "epoch": 0.5141579731743666,
504
+ "grad_norm": 0.3173169163707792,
505
+ "learning_rate": 5.636485759973729e-06,
506
+ "loss": 1.5806,
507
+ "step": 345
508
+ },
509
+ {
510
+ "epoch": 0.5216095380029806,
511
+ "grad_norm": 0.3363924031410881,
512
+ "learning_rate": 5.507095595679059e-06,
513
+ "loss": 1.6871,
514
+ "step": 350
515
+ },
516
+ {
517
+ "epoch": 0.5290611028315947,
518
+ "grad_norm": 0.3307680686760043,
519
+ "learning_rate": 5.377361342685287e-06,
520
+ "loss": 1.6933,
521
+ "step": 355
522
+ },
523
+ {
524
+ "epoch": 0.5365126676602087,
525
+ "grad_norm": 0.32876995727805525,
526
+ "learning_rate": 5.2473710319095054e-06,
527
+ "loss": 1.6201,
528
+ "step": 360
529
+ },
530
+ {
531
+ "epoch": 0.5439642324888226,
532
+ "grad_norm": 0.3452442083785589,
533
+ "learning_rate": 5.117212868016303e-06,
534
+ "loss": 1.7146,
535
+ "step": 365
536
+ },
537
+ {
538
+ "epoch": 0.5514157973174366,
539
+ "grad_norm": 0.3659924599969553,
540
+ "learning_rate": 4.9869751695666615e-06,
541
+ "loss": 1.6215,
542
+ "step": 370
543
+ },
544
+ {
545
+ "epoch": 0.5588673621460507,
546
+ "grad_norm": 0.3233317882165812,
547
+ "learning_rate": 4.856746309089582e-06,
548
+ "loss": 1.6315,
549
+ "step": 375
550
+ },
551
+ {
552
+ "epoch": 0.5663189269746647,
553
+ "grad_norm": 0.30217314493419994,
554
+ "learning_rate": 4.726614653117071e-06,
555
+ "loss": 1.6945,
556
+ "step": 380
557
+ },
558
+ {
559
+ "epoch": 0.5737704918032787,
560
+ "grad_norm": 0.35844882665530153,
561
+ "learning_rate": 4.596668502223214e-06,
562
+ "loss": 1.6924,
563
+ "step": 385
564
+ },
565
+ {
566
+ "epoch": 0.5812220566318927,
567
+ "grad_norm": 0.34041157047813575,
568
+ "learning_rate": 4.466996031108004e-06,
569
+ "loss": 1.629,
570
+ "step": 390
571
+ },
572
+ {
573
+ "epoch": 0.5886736214605067,
574
+ "grad_norm": 0.314658238073181,
575
+ "learning_rate": 4.337685228766561e-06,
576
+ "loss": 1.6797,
577
+ "step": 395
578
+ },
579
+ {
580
+ "epoch": 0.5961251862891207,
581
+ "grad_norm": 0.32647573861445145,
582
+ "learning_rate": 4.208823838784387e-06,
583
+ "loss": 1.6447,
584
+ "step": 400
585
+ },
586
+ {
587
+ "epoch": 0.5961251862891207,
588
+ "eval_loss": 1.6408501863479614,
589
+ "eval_runtime": 580.0543,
590
+ "eval_samples_per_second": 4.113,
591
+ "eval_steps_per_second": 0.259,
592
+ "step": 400
593
+ },
594
+ {
595
+ "epoch": 0.6035767511177347,
596
+ "grad_norm": 0.37909958848060366,
597
+ "learning_rate": 4.0804992997991335e-06,
598
+ "loss": 1.6574,
599
+ "step": 405
600
+ },
601
+ {
602
+ "epoch": 0.6110283159463488,
603
+ "grad_norm": 0.3790204664396362,
604
+ "learning_rate": 3.952798686169279e-06,
605
+ "loss": 1.7235,
606
+ "step": 410
607
+ },
608
+ {
609
+ "epoch": 0.6184798807749627,
610
+ "grad_norm": 0.3703026855695823,
611
+ "learning_rate": 3.825808648890005e-06,
612
+ "loss": 1.675,
613
+ "step": 415
614
+ },
615
+ {
616
+ "epoch": 0.6259314456035767,
617
+ "grad_norm": 0.3426042362005175,
618
+ "learning_rate": 3.699615356796342e-06,
619
+ "loss": 1.5951,
620
+ "step": 420
621
+ },
622
+ {
623
+ "epoch": 0.6333830104321908,
624
+ "grad_norm": 0.3472991803298781,
625
+ "learning_rate": 3.5743044380934655e-06,
626
+ "loss": 1.7072,
627
+ "step": 425
628
+ },
629
+ {
630
+ "epoch": 0.6408345752608048,
631
+ "grad_norm": 0.3645949602782747,
632
+ "learning_rate": 3.449960922253858e-06,
633
+ "loss": 1.6207,
634
+ "step": 430
635
+ },
636
+ {
637
+ "epoch": 0.6482861400894188,
638
+ "grad_norm": 0.3314569171608062,
639
+ "learning_rate": 3.326669182320736e-06,
640
+ "loss": 1.6281,
641
+ "step": 435
642
+ },
643
+ {
644
+ "epoch": 0.6557377049180327,
645
+ "grad_norm": 0.31811370472714595,
646
+ "learning_rate": 3.2045128776568783e-06,
647
+ "loss": 1.6572,
648
+ "step": 440
649
+ },
650
+ {
651
+ "epoch": 0.6631892697466468,
652
+ "grad_norm": 0.3377222335896338,
653
+ "learning_rate": 3.0835748971777413e-06,
654
+ "loss": 1.6651,
655
+ "step": 445
656
+ },
657
+ {
658
+ "epoch": 0.6706408345752608,
659
+ "grad_norm": 0.331361220355902,
660
+ "learning_rate": 2.9639373031073525e-06,
661
+ "loss": 1.6878,
662
+ "step": 450
663
+ },
664
+ {
665
+ "epoch": 0.6780923994038748,
666
+ "grad_norm": 0.30006578695394276,
667
+ "learning_rate": 2.8456812752951483e-06,
668
+ "loss": 1.6887,
669
+ "step": 455
670
+ },
671
+ {
672
+ "epoch": 0.6855439642324889,
673
+ "grad_norm": 0.34387137915082105,
674
+ "learning_rate": 2.728887056131553e-06,
675
+ "loss": 1.6273,
676
+ "step": 460
677
+ },
678
+ {
679
+ "epoch": 0.6929955290611028,
680
+ "grad_norm": 0.3426777102093575,
681
+ "learning_rate": 2.6136338960996665e-06,
682
+ "loss": 1.6867,
683
+ "step": 465
684
+ },
685
+ {
686
+ "epoch": 0.7004470938897168,
687
+ "grad_norm": 0.3047030104059201,
688
+ "learning_rate": 2.5000000000000015e-06,
689
+ "loss": 1.6514,
690
+ "step": 470
691
+ },
692
+ {
693
+ "epoch": 0.7078986587183308,
694
+ "grad_norm": 0.3194123549278598,
695
+ "learning_rate": 2.3880624738847837e-06,
696
+ "loss": 1.6494,
697
+ "step": 475
698
+ },
699
+ {
700
+ "epoch": 0.7153502235469449,
701
+ "grad_norm": 0.28354706856772705,
702
+ "learning_rate": 2.277897272737787e-06,
703
+ "loss": 1.5821,
704
+ "step": 480
705
+ },
706
+ {
707
+ "epoch": 0.7228017883755589,
708
+ "grad_norm": 0.36693040405683097,
709
+ "learning_rate": 2.1695791489352346e-06,
710
+ "loss": 1.6627,
711
+ "step": 485
712
+ },
713
+ {
714
+ "epoch": 0.7302533532041728,
715
+ "grad_norm": 0.3268289701429299,
716
+ "learning_rate": 2.063181601522722e-06,
717
+ "loss": 1.543,
718
+ "step": 490
719
+ },
720
+ {
721
+ "epoch": 0.7377049180327869,
722
+ "grad_norm": 0.337638547503993,
723
+ "learning_rate": 1.9587768263425886e-06,
724
+ "loss": 1.6527,
725
+ "step": 495
726
+ },
727
+ {
728
+ "epoch": 0.7451564828614009,
729
+ "grad_norm": 0.2879745114712397,
730
+ "learning_rate": 1.856435667045577e-06,
731
+ "loss": 1.5879,
732
+ "step": 500
733
+ },
734
+ {
735
+ "epoch": 0.7526080476900149,
736
+ "grad_norm": 0.3081841268702886,
737
+ "learning_rate": 1.7562275670200041e-06,
738
+ "loss": 1.5423,
739
+ "step": 505
740
+ },
741
+ {
742
+ "epoch": 0.7600596125186289,
743
+ "grad_norm": 0.3237369195299089,
744
+ "learning_rate": 1.6582205222711051e-06,
745
+ "loss": 1.6172,
746
+ "step": 510
747
+ },
748
+ {
749
+ "epoch": 0.767511177347243,
750
+ "grad_norm": 0.3484298930902378,
751
+ "learning_rate": 1.5624810352824709e-06,
752
+ "loss": 1.6456,
753
+ "step": 515
754
+ },
755
+ {
756
+ "epoch": 0.7749627421758569,
757
+ "grad_norm": 0.3160389330113336,
758
+ "learning_rate": 1.4690740698909223e-06,
759
+ "loss": 1.5929,
760
+ "step": 520
761
+ },
762
+ {
763
+ "epoch": 0.7824143070044709,
764
+ "grad_norm": 0.36108945655051267,
765
+ "learning_rate": 1.3780630072054313e-06,
766
+ "loss": 1.6718,
767
+ "step": 525
768
+ },
769
+ {
770
+ "epoch": 0.789865871833085,
771
+ "grad_norm": 0.32927194714240243,
772
+ "learning_rate": 1.289509602599996e-06,
773
+ "loss": 1.6058,
774
+ "step": 530
775
+ },
776
+ {
777
+ "epoch": 0.797317436661699,
778
+ "grad_norm": 0.3769043230835878,
779
+ "learning_rate": 1.203473943809651e-06,
780
+ "loss": 1.6295,
781
+ "step": 535
782
+ },
783
+ {
784
+ "epoch": 0.8047690014903129,
785
+ "grad_norm": 0.33941548710215963,
786
+ "learning_rate": 1.1200144101580634e-06,
787
+ "loss": 1.6792,
788
+ "step": 540
789
+ },
790
+ {
791
+ "epoch": 0.812220566318927,
792
+ "grad_norm": 0.3706526425603311,
793
+ "learning_rate": 1.0391876329443534e-06,
794
+ "loss": 1.6146,
795
+ "step": 545
796
+ },
797
+ {
798
+ "epoch": 0.819672131147541,
799
+ "grad_norm": 0.35398755740858545,
800
+ "learning_rate": 9.610484570160444e-07,
801
+ "loss": 1.7103,
802
+ "step": 550
803
+ },
804
+ {
805
+ "epoch": 0.827123695976155,
806
+ "grad_norm": 0.35561387409815454,
807
+ "learning_rate": 8.856499035541972e-07,
808
+ "loss": 1.6163,
809
+ "step": 555
810
+ },
811
+ {
812
+ "epoch": 0.834575260804769,
813
+ "grad_norm": 0.348602818005502,
814
+ "learning_rate": 8.130431340959982e-07,
815
+ "loss": 1.6807,
816
+ "step": 560
817
+ },
818
+ {
819
+ "epoch": 0.842026825633383,
820
+ "grad_norm": 0.3675149781602246,
821
+ "learning_rate": 7.432774158191946e-07,
822
+ "loss": 1.6201,
823
+ "step": 565
824
+ },
825
+ {
826
+ "epoch": 0.849478390461997,
827
+ "grad_norm": 0.35625514274912495,
828
+ "learning_rate": 6.764000881119631e-07,
829
+ "loss": 1.5743,
830
+ "step": 570
831
+ },
832
+ {
833
+ "epoch": 0.856929955290611,
834
+ "grad_norm": 0.3186585069526685,
835
+ "learning_rate": 6.12456530450844e-07,
836
+ "loss": 1.6272,
837
+ "step": 575
838
+ },
839
+ {
840
+ "epoch": 0.8643815201192251,
841
+ "grad_norm": 0.3347624139888847,
842
+ "learning_rate": 5.514901316086058e-07,
843
+ "loss": 1.6818,
844
+ "step": 580
845
+ },
846
+ {
847
+ "epoch": 0.8718330849478391,
848
+ "grad_norm": 0.3821665104781894,
849
+ "learning_rate": 4.935422602128698e-07,
850
+ "loss": 1.7344,
851
+ "step": 585
852
+ },
853
+ {
854
+ "epoch": 0.879284649776453,
855
+ "grad_norm": 0.31160320897543975,
856
+ "learning_rate": 4.386522366755169e-07,
857
+ "loss": 1.5668,
858
+ "step": 590
859
+ },
860
+ {
861
+ "epoch": 0.886736214605067,
862
+ "grad_norm": 0.32657433748759934,
863
+ "learning_rate": 3.868573065118936e-07,
864
+ "loss": 1.6732,
865
+ "step": 595
866
+ },
867
+ {
868
+ "epoch": 0.8941877794336811,
869
+ "grad_norm": 0.34252424286072203,
870
+ "learning_rate": 3.381926150679543e-07,
871
+ "loss": 1.6245,
872
+ "step": 600
873
+ },
874
+ {
875
+ "epoch": 0.8941877794336811,
876
+ "eval_loss": 1.6357628107070923,
877
+ "eval_runtime": 580.6663,
878
+ "eval_samples_per_second": 4.109,
879
+ "eval_steps_per_second": 0.258,
880
+ "step": 600
881
+ },
882
+ {
883
+ "epoch": 0.9016393442622951,
884
+ "grad_norm": 0.3432161875223241,
885
+ "learning_rate": 2.9269118367244385e-07,
886
+ "loss": 1.6999,
887
+ "step": 605
888
+ },
889
+ {
890
+ "epoch": 0.9090909090909091,
891
+ "grad_norm": 0.3458124106576635,
892
+ "learning_rate": 2.5038388723034935e-07,
893
+ "loss": 1.6037,
894
+ "step": 610
895
+ },
896
+ {
897
+ "epoch": 0.9165424739195231,
898
+ "grad_norm": 0.34778709529194757,
899
+ "learning_rate": 2.1129943327279522e-07,
900
+ "loss": 1.7079,
901
+ "step": 615
902
+ },
903
+ {
904
+ "epoch": 0.9239940387481371,
905
+ "grad_norm": 0.3605158717479098,
906
+ "learning_rate": 1.7546434247760147e-07,
907
+ "loss": 1.6404,
908
+ "step": 620
909
+ },
910
+ {
911
+ "epoch": 0.9314456035767511,
912
+ "grad_norm": 0.29865645458342877,
913
+ "learning_rate": 1.429029306737345e-07,
914
+ "loss": 1.6086,
915
+ "step": 625
916
+ },
917
+ {
918
+ "epoch": 0.9388971684053651,
919
+ "grad_norm": 0.3290659895478686,
920
+ "learning_rate": 1.1363729234184828e-07,
921
+ "loss": 1.596,
922
+ "step": 630
923
+ },
924
+ {
925
+ "epoch": 0.9463487332339792,
926
+ "grad_norm": 0.32311961294020886,
927
+ "learning_rate": 8.768728562211948e-08,
928
+ "loss": 1.5774,
929
+ "step": 635
930
+ },
931
+ {
932
+ "epoch": 0.9538002980625931,
933
+ "grad_norm": 0.3628568702485729,
934
+ "learning_rate": 6.507051883954618e-08,
935
+ "loss": 1.6421,
936
+ "step": 640
937
+ },
938
+ {
939
+ "epoch": 0.9612518628912071,
940
+ "grad_norm": 0.35678638324651946,
941
+ "learning_rate": 4.580233855585425e-08,
942
+ "loss": 1.5815,
943
+ "step": 645
944
+ },
945
+ {
946
+ "epoch": 0.9687034277198212,
947
+ "grad_norm": 0.32862520535733186,
948
+ "learning_rate": 2.989581915611994e-08,
949
+ "loss": 1.5756,
950
+ "step": 650
951
+ },
952
+ {
953
+ "epoch": 0.9761549925484352,
954
+ "grad_norm": 0.36012152823713983,
955
+ "learning_rate": 1.7361753977169214e-08,
956
+ "loss": 1.6405,
957
+ "step": 655
958
+ },
959
+ {
960
+ "epoch": 0.9836065573770492,
961
+ "grad_norm": 0.3640507189017552,
962
+ "learning_rate": 8.208647983782847e-09,
963
+ "loss": 1.62,
964
+ "step": 660
965
+ },
966
+ {
967
+ "epoch": 0.9910581222056631,
968
+ "grad_norm": 0.33586265256557224,
969
+ "learning_rate": 2.442711997670544e-09,
970
+ "loss": 1.6375,
971
+ "step": 665
972
+ },
973
+ {
974
+ "epoch": 0.9985096870342772,
975
+ "grad_norm": 0.3189391910196833,
976
+ "learning_rate": 6.785848312707011e-11,
977
+ "loss": 1.6954,
978
+ "step": 670
979
+ },
980
+ {
981
+ "epoch": 1.0,
982
+ "step": 671,
983
+ "total_flos": 919616052789248.0,
984
+ "train_loss": 1.6831368206154926,
985
+ "train_runtime": 18035.2779,
986
+ "train_samples_per_second": 1.191,
987
+ "train_steps_per_second": 0.037
988
+ }
989
+ ],
990
+ "logging_steps": 5,
991
+ "max_steps": 671,
992
+ "num_input_tokens_seen": 0,
993
+ "num_train_epochs": 1,
994
+ "save_steps": 200,
995
+ "stateful_callbacks": {
996
+ "TrainerControl": {
997
+ "args": {
998
+ "should_epoch_stop": false,
999
+ "should_evaluate": false,
1000
+ "should_log": false,
1001
+ "should_save": true,
1002
+ "should_training_stop": false
1003
+ },
1004
+ "attributes": {}
1005
+ }
1006
+ },
1007
+ "total_flos": 919616052789248.0,
1008
+ "train_batch_size": 4,
1009
+ "trial_name": null,
1010
+ "trial_params": null
1011
+ }