sedrickkeh commited on
Commit
8ba46b5
·
verified ·
1 Parent(s): ac4e1f8

End of training

Browse files
README.md CHANGED
@@ -4,6 +4,7 @@ license: gemma
4
  base_model: google/gemma-2-9b
5
  tags:
6
  - llama-factory
 
7
  - generated_from_trainer
8
  model-index:
9
  - name: hp_ablations_gemma_epoch4
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # hp_ablations_gemma_epoch4
17
 
18
- This model is a fine-tuned version of [google/gemma-2-9b](https://huggingface.co/google/gemma-2-9b) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
  - Loss: 0.6362
21
 
 
4
  base_model: google/gemma-2-9b
5
  tags:
6
  - llama-factory
7
+ - full
8
  - generated_from_trainer
9
  model-index:
10
  - name: hp_ablations_gemma_epoch4
 
16
 
17
  # hp_ablations_gemma_epoch4
18
 
19
+ This model is a fine-tuned version of [google/gemma-2-9b](https://huggingface.co/google/gemma-2-9b) on the mlfoundations-dev/oh-dcft-v3.1-gpt-4o-mini dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.6362
22
 
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.9988716502115658,
3
+ "eval_loss": 0.6362190246582031,
4
+ "eval_runtime": 361.6713,
5
+ "eval_samples_per_second": 33.016,
6
+ "eval_steps_per_second": 0.517,
7
+ "total_flos": 6752101308825600.0,
8
+ "train_loss": 0.5187085666788351,
9
+ "train_runtime": 82722.7317,
10
+ "train_samples_per_second": 10.97,
11
+ "train_steps_per_second": 0.021
12
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.9988716502115658,
3
+ "eval_loss": 0.6362190246582031,
4
+ "eval_runtime": 361.6713,
5
+ "eval_samples_per_second": 33.016,
6
+ "eval_steps_per_second": 0.517
7
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.9988716502115658,
3
+ "total_flos": 6752101308825600.0,
4
+ "train_loss": 0.5187085666788351,
5
+ "train_runtime": 82722.7317,
6
+ "train_samples_per_second": 10.97,
7
+ "train_steps_per_second": 0.021
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1313 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.9988716502115658,
5
+ "eval_steps": 500,
6
+ "global_step": 1772,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.022566995768688293,
13
+ "grad_norm": 1.1533478871584986,
14
+ "learning_rate": 5e-06,
15
+ "loss": 0.744,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.045133991537376586,
20
+ "grad_norm": 0.964708516920586,
21
+ "learning_rate": 5e-06,
22
+ "loss": 0.6804,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.06770098730606489,
27
+ "grad_norm": 0.7414689176853874,
28
+ "learning_rate": 5e-06,
29
+ "loss": 0.6629,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.09026798307475317,
34
+ "grad_norm": 0.6402562297823468,
35
+ "learning_rate": 5e-06,
36
+ "loss": 0.6403,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.11283497884344147,
41
+ "grad_norm": 0.6438158764026781,
42
+ "learning_rate": 5e-06,
43
+ "loss": 0.6334,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.13540197461212977,
48
+ "grad_norm": 0.6086347979873902,
49
+ "learning_rate": 5e-06,
50
+ "loss": 0.6331,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.15796897038081806,
55
+ "grad_norm": 0.637737303544012,
56
+ "learning_rate": 5e-06,
57
+ "loss": 0.6285,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.18053596614950634,
62
+ "grad_norm": 0.6180779913246882,
63
+ "learning_rate": 5e-06,
64
+ "loss": 0.6309,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.20310296191819463,
69
+ "grad_norm": 0.7628906889304371,
70
+ "learning_rate": 5e-06,
71
+ "loss": 0.6285,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.22566995768688294,
76
+ "grad_norm": 0.6560412532230508,
77
+ "learning_rate": 5e-06,
78
+ "loss": 0.6273,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.24823695345557123,
83
+ "grad_norm": 0.5971484737001814,
84
+ "learning_rate": 5e-06,
85
+ "loss": 0.6183,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.27080394922425954,
90
+ "grad_norm": 0.62454415866273,
91
+ "learning_rate": 5e-06,
92
+ "loss": 0.6169,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.2933709449929478,
97
+ "grad_norm": 0.9358718838187522,
98
+ "learning_rate": 5e-06,
99
+ "loss": 0.6184,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.3159379407616361,
104
+ "grad_norm": 0.6707162435124239,
105
+ "learning_rate": 5e-06,
106
+ "loss": 0.6134,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.3385049365303244,
111
+ "grad_norm": 0.607779817667474,
112
+ "learning_rate": 5e-06,
113
+ "loss": 0.6135,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.3610719322990127,
118
+ "grad_norm": 0.5866694919013473,
119
+ "learning_rate": 5e-06,
120
+ "loss": 0.6163,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.383638928067701,
125
+ "grad_norm": 0.6395663339457631,
126
+ "learning_rate": 5e-06,
127
+ "loss": 0.6124,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.40620592383638926,
132
+ "grad_norm": 0.5802311797909204,
133
+ "learning_rate": 5e-06,
134
+ "loss": 0.6086,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.4287729196050776,
139
+ "grad_norm": 0.5655529914842923,
140
+ "learning_rate": 5e-06,
141
+ "loss": 0.6034,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.4513399153737659,
146
+ "grad_norm": 0.5978170679918469,
147
+ "learning_rate": 5e-06,
148
+ "loss": 0.6094,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.47390691114245415,
153
+ "grad_norm": 0.5598729087663762,
154
+ "learning_rate": 5e-06,
155
+ "loss": 0.606,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.49647390691114246,
160
+ "grad_norm": 0.6135515117029436,
161
+ "learning_rate": 5e-06,
162
+ "loss": 0.5985,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.5190409026798307,
167
+ "grad_norm": 0.5982246973373736,
168
+ "learning_rate": 5e-06,
169
+ "loss": 0.6074,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.5416078984485191,
174
+ "grad_norm": 0.634679350959857,
175
+ "learning_rate": 5e-06,
176
+ "loss": 0.6056,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.5641748942172073,
181
+ "grad_norm": 0.5848371303406198,
182
+ "learning_rate": 5e-06,
183
+ "loss": 0.5998,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.5867418899858956,
188
+ "grad_norm": 0.7237357300883238,
189
+ "learning_rate": 5e-06,
190
+ "loss": 0.6029,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.609308885754584,
195
+ "grad_norm": 0.5758396058045636,
196
+ "learning_rate": 5e-06,
197
+ "loss": 0.5982,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.6318758815232722,
202
+ "grad_norm": 0.551647922547292,
203
+ "learning_rate": 5e-06,
204
+ "loss": 0.6043,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.6544428772919605,
209
+ "grad_norm": 0.6070055411621582,
210
+ "learning_rate": 5e-06,
211
+ "loss": 0.5979,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.6770098730606487,
216
+ "grad_norm": 0.6071346856787038,
217
+ "learning_rate": 5e-06,
218
+ "loss": 0.6042,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.6995768688293371,
223
+ "grad_norm": 0.5694826541205662,
224
+ "learning_rate": 5e-06,
225
+ "loss": 0.5972,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.7221438645980254,
230
+ "grad_norm": 0.5219595834784061,
231
+ "learning_rate": 5e-06,
232
+ "loss": 0.6006,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.7447108603667136,
237
+ "grad_norm": 0.5425184695486164,
238
+ "learning_rate": 5e-06,
239
+ "loss": 0.5962,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.767277856135402,
244
+ "grad_norm": 0.5849183234601607,
245
+ "learning_rate": 5e-06,
246
+ "loss": 0.5956,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 0.7898448519040903,
251
+ "grad_norm": 0.6577639097966262,
252
+ "learning_rate": 5e-06,
253
+ "loss": 0.5968,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 0.8124118476727785,
258
+ "grad_norm": 0.6147316484890941,
259
+ "learning_rate": 5e-06,
260
+ "loss": 0.596,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 0.8349788434414669,
265
+ "grad_norm": 0.5747506292979353,
266
+ "learning_rate": 5e-06,
267
+ "loss": 0.6027,
268
+ "step": 370
269
+ },
270
+ {
271
+ "epoch": 0.8575458392101551,
272
+ "grad_norm": 0.5687682085546433,
273
+ "learning_rate": 5e-06,
274
+ "loss": 0.5999,
275
+ "step": 380
276
+ },
277
+ {
278
+ "epoch": 0.8801128349788434,
279
+ "grad_norm": 0.616324958523922,
280
+ "learning_rate": 5e-06,
281
+ "loss": 0.5941,
282
+ "step": 390
283
+ },
284
+ {
285
+ "epoch": 0.9026798307475318,
286
+ "grad_norm": 0.5900329665323057,
287
+ "learning_rate": 5e-06,
288
+ "loss": 0.5962,
289
+ "step": 400
290
+ },
291
+ {
292
+ "epoch": 0.92524682651622,
293
+ "grad_norm": 0.5841528129448796,
294
+ "learning_rate": 5e-06,
295
+ "loss": 0.5939,
296
+ "step": 410
297
+ },
298
+ {
299
+ "epoch": 0.9478138222849083,
300
+ "grad_norm": 0.586906708055177,
301
+ "learning_rate": 5e-06,
302
+ "loss": 0.5944,
303
+ "step": 420
304
+ },
305
+ {
306
+ "epoch": 0.9703808180535967,
307
+ "grad_norm": 0.5824744589920343,
308
+ "learning_rate": 5e-06,
309
+ "loss": 0.5887,
310
+ "step": 430
311
+ },
312
+ {
313
+ "epoch": 0.9929478138222849,
314
+ "grad_norm": 0.7143042613195987,
315
+ "learning_rate": 5e-06,
316
+ "loss": 0.5935,
317
+ "step": 440
318
+ },
319
+ {
320
+ "epoch": 0.9997179125528914,
321
+ "eval_loss": 0.5913009643554688,
322
+ "eval_runtime": 359.8373,
323
+ "eval_samples_per_second": 33.184,
324
+ "eval_steps_per_second": 0.52,
325
+ "step": 443
326
+ },
327
+ {
328
+ "epoch": 1.0155148095909732,
329
+ "grad_norm": 0.7349947461827919,
330
+ "learning_rate": 5e-06,
331
+ "loss": 0.6021,
332
+ "step": 450
333
+ },
334
+ {
335
+ "epoch": 1.0380818053596614,
336
+ "grad_norm": 0.5967862073767953,
337
+ "learning_rate": 5e-06,
338
+ "loss": 0.528,
339
+ "step": 460
340
+ },
341
+ {
342
+ "epoch": 1.0606488011283497,
343
+ "grad_norm": 0.5908248473188383,
344
+ "learning_rate": 5e-06,
345
+ "loss": 0.543,
346
+ "step": 470
347
+ },
348
+ {
349
+ "epoch": 1.0832157968970382,
350
+ "grad_norm": 0.5783595460572618,
351
+ "learning_rate": 5e-06,
352
+ "loss": 0.5306,
353
+ "step": 480
354
+ },
355
+ {
356
+ "epoch": 1.1057827926657264,
357
+ "grad_norm": 0.6027833555237775,
358
+ "learning_rate": 5e-06,
359
+ "loss": 0.541,
360
+ "step": 490
361
+ },
362
+ {
363
+ "epoch": 1.1283497884344147,
364
+ "grad_norm": 0.5779961111728702,
365
+ "learning_rate": 5e-06,
366
+ "loss": 0.5393,
367
+ "step": 500
368
+ },
369
+ {
370
+ "epoch": 1.150916784203103,
371
+ "grad_norm": 0.609691227257698,
372
+ "learning_rate": 5e-06,
373
+ "loss": 0.5409,
374
+ "step": 510
375
+ },
376
+ {
377
+ "epoch": 1.1734837799717912,
378
+ "grad_norm": 0.6288262969758429,
379
+ "learning_rate": 5e-06,
380
+ "loss": 0.5336,
381
+ "step": 520
382
+ },
383
+ {
384
+ "epoch": 1.1960507757404795,
385
+ "grad_norm": 0.6441493312512533,
386
+ "learning_rate": 5e-06,
387
+ "loss": 0.5393,
388
+ "step": 530
389
+ },
390
+ {
391
+ "epoch": 1.2186177715091677,
392
+ "grad_norm": 0.5801582242360297,
393
+ "learning_rate": 5e-06,
394
+ "loss": 0.5371,
395
+ "step": 540
396
+ },
397
+ {
398
+ "epoch": 1.2411847672778562,
399
+ "grad_norm": 0.6407989788015263,
400
+ "learning_rate": 5e-06,
401
+ "loss": 0.5422,
402
+ "step": 550
403
+ },
404
+ {
405
+ "epoch": 1.2637517630465445,
406
+ "grad_norm": 0.6145932063398675,
407
+ "learning_rate": 5e-06,
408
+ "loss": 0.5438,
409
+ "step": 560
410
+ },
411
+ {
412
+ "epoch": 1.2863187588152327,
413
+ "grad_norm": 0.5894609854660963,
414
+ "learning_rate": 5e-06,
415
+ "loss": 0.5415,
416
+ "step": 570
417
+ },
418
+ {
419
+ "epoch": 1.308885754583921,
420
+ "grad_norm": 0.5712716247350705,
421
+ "learning_rate": 5e-06,
422
+ "loss": 0.5409,
423
+ "step": 580
424
+ },
425
+ {
426
+ "epoch": 1.3314527503526092,
427
+ "grad_norm": 0.5545979807457848,
428
+ "learning_rate": 5e-06,
429
+ "loss": 0.5409,
430
+ "step": 590
431
+ },
432
+ {
433
+ "epoch": 1.3540197461212977,
434
+ "grad_norm": 0.6116638284685687,
435
+ "learning_rate": 5e-06,
436
+ "loss": 0.5463,
437
+ "step": 600
438
+ },
439
+ {
440
+ "epoch": 1.376586741889986,
441
+ "grad_norm": 0.603741984926436,
442
+ "learning_rate": 5e-06,
443
+ "loss": 0.539,
444
+ "step": 610
445
+ },
446
+ {
447
+ "epoch": 1.3991537376586742,
448
+ "grad_norm": 0.5576981790641173,
449
+ "learning_rate": 5e-06,
450
+ "loss": 0.543,
451
+ "step": 620
452
+ },
453
+ {
454
+ "epoch": 1.4217207334273625,
455
+ "grad_norm": 0.5585863324687254,
456
+ "learning_rate": 5e-06,
457
+ "loss": 0.5411,
458
+ "step": 630
459
+ },
460
+ {
461
+ "epoch": 1.4442877291960508,
462
+ "grad_norm": 0.6685966084888875,
463
+ "learning_rate": 5e-06,
464
+ "loss": 0.5407,
465
+ "step": 640
466
+ },
467
+ {
468
+ "epoch": 1.466854724964739,
469
+ "grad_norm": 0.563715043671093,
470
+ "learning_rate": 5e-06,
471
+ "loss": 0.5418,
472
+ "step": 650
473
+ },
474
+ {
475
+ "epoch": 1.4894217207334273,
476
+ "grad_norm": 0.664496256459769,
477
+ "learning_rate": 5e-06,
478
+ "loss": 0.5456,
479
+ "step": 660
480
+ },
481
+ {
482
+ "epoch": 1.5119887165021155,
483
+ "grad_norm": 0.5712447353649974,
484
+ "learning_rate": 5e-06,
485
+ "loss": 0.5405,
486
+ "step": 670
487
+ },
488
+ {
489
+ "epoch": 1.5345557122708038,
490
+ "grad_norm": 0.5486833946787922,
491
+ "learning_rate": 5e-06,
492
+ "loss": 0.5345,
493
+ "step": 680
494
+ },
495
+ {
496
+ "epoch": 1.5571227080394923,
497
+ "grad_norm": 0.6092638196410558,
498
+ "learning_rate": 5e-06,
499
+ "loss": 0.537,
500
+ "step": 690
501
+ },
502
+ {
503
+ "epoch": 1.5796897038081805,
504
+ "grad_norm": 0.5716085259389777,
505
+ "learning_rate": 5e-06,
506
+ "loss": 0.5407,
507
+ "step": 700
508
+ },
509
+ {
510
+ "epoch": 1.6022566995768688,
511
+ "grad_norm": 0.5372993001292306,
512
+ "learning_rate": 5e-06,
513
+ "loss": 0.5437,
514
+ "step": 710
515
+ },
516
+ {
517
+ "epoch": 1.6248236953455573,
518
+ "grad_norm": 0.5950554022209111,
519
+ "learning_rate": 5e-06,
520
+ "loss": 0.5516,
521
+ "step": 720
522
+ },
523
+ {
524
+ "epoch": 1.6473906911142455,
525
+ "grad_norm": 0.5644998290854271,
526
+ "learning_rate": 5e-06,
527
+ "loss": 0.5347,
528
+ "step": 730
529
+ },
530
+ {
531
+ "epoch": 1.6699576868829338,
532
+ "grad_norm": 0.7194589295267496,
533
+ "learning_rate": 5e-06,
534
+ "loss": 0.5506,
535
+ "step": 740
536
+ },
537
+ {
538
+ "epoch": 1.692524682651622,
539
+ "grad_norm": 0.654147586643372,
540
+ "learning_rate": 5e-06,
541
+ "loss": 0.5424,
542
+ "step": 750
543
+ },
544
+ {
545
+ "epoch": 1.7150916784203103,
546
+ "grad_norm": 0.5858691889312596,
547
+ "learning_rate": 5e-06,
548
+ "loss": 0.5412,
549
+ "step": 760
550
+ },
551
+ {
552
+ "epoch": 1.7376586741889986,
553
+ "grad_norm": 0.6698089700161991,
554
+ "learning_rate": 5e-06,
555
+ "loss": 0.5414,
556
+ "step": 770
557
+ },
558
+ {
559
+ "epoch": 1.7602256699576868,
560
+ "grad_norm": 0.5539764143704193,
561
+ "learning_rate": 5e-06,
562
+ "loss": 0.5372,
563
+ "step": 780
564
+ },
565
+ {
566
+ "epoch": 1.782792665726375,
567
+ "grad_norm": 0.5902260571431798,
568
+ "learning_rate": 5e-06,
569
+ "loss": 0.5411,
570
+ "step": 790
571
+ },
572
+ {
573
+ "epoch": 1.8053596614950633,
574
+ "grad_norm": 0.6653827196776052,
575
+ "learning_rate": 5e-06,
576
+ "loss": 0.5364,
577
+ "step": 800
578
+ },
579
+ {
580
+ "epoch": 1.8279266572637518,
581
+ "grad_norm": 0.5629278214932536,
582
+ "learning_rate": 5e-06,
583
+ "loss": 0.5471,
584
+ "step": 810
585
+ },
586
+ {
587
+ "epoch": 1.85049365303244,
588
+ "grad_norm": 0.716393148987963,
589
+ "learning_rate": 5e-06,
590
+ "loss": 0.5343,
591
+ "step": 820
592
+ },
593
+ {
594
+ "epoch": 1.8730606488011283,
595
+ "grad_norm": 0.5869857580731778,
596
+ "learning_rate": 5e-06,
597
+ "loss": 0.546,
598
+ "step": 830
599
+ },
600
+ {
601
+ "epoch": 1.8956276445698168,
602
+ "grad_norm": 0.5932511634770244,
603
+ "learning_rate": 5e-06,
604
+ "loss": 0.5461,
605
+ "step": 840
606
+ },
607
+ {
608
+ "epoch": 1.918194640338505,
609
+ "grad_norm": 0.5548454392387719,
610
+ "learning_rate": 5e-06,
611
+ "loss": 0.5342,
612
+ "step": 850
613
+ },
614
+ {
615
+ "epoch": 1.9407616361071933,
616
+ "grad_norm": 0.6292606426061244,
617
+ "learning_rate": 5e-06,
618
+ "loss": 0.5391,
619
+ "step": 860
620
+ },
621
+ {
622
+ "epoch": 1.9633286318758816,
623
+ "grad_norm": 0.6964512767642522,
624
+ "learning_rate": 5e-06,
625
+ "loss": 0.5457,
626
+ "step": 870
627
+ },
628
+ {
629
+ "epoch": 1.9858956276445698,
630
+ "grad_norm": 0.597052250032102,
631
+ "learning_rate": 5e-06,
632
+ "loss": 0.5435,
633
+ "step": 880
634
+ },
635
+ {
636
+ "epoch": 1.9994358251057829,
637
+ "eval_loss": 0.5871431231498718,
638
+ "eval_runtime": 359.7238,
639
+ "eval_samples_per_second": 33.195,
640
+ "eval_steps_per_second": 0.52,
641
+ "step": 886
642
+ },
643
+ {
644
+ "epoch": 2.008462623413258,
645
+ "grad_norm": 1.0307272915735466,
646
+ "learning_rate": 5e-06,
647
+ "loss": 0.5672,
648
+ "step": 890
649
+ },
650
+ {
651
+ "epoch": 2.0310296191819464,
652
+ "grad_norm": 0.7570634539626596,
653
+ "learning_rate": 5e-06,
654
+ "loss": 0.4779,
655
+ "step": 900
656
+ },
657
+ {
658
+ "epoch": 2.0535966149506346,
659
+ "grad_norm": 0.7952962689694523,
660
+ "learning_rate": 5e-06,
661
+ "loss": 0.4796,
662
+ "step": 910
663
+ },
664
+ {
665
+ "epoch": 2.076163610719323,
666
+ "grad_norm": 0.6214258578581712,
667
+ "learning_rate": 5e-06,
668
+ "loss": 0.472,
669
+ "step": 920
670
+ },
671
+ {
672
+ "epoch": 2.098730606488011,
673
+ "grad_norm": 0.6061682325391837,
674
+ "learning_rate": 5e-06,
675
+ "loss": 0.4792,
676
+ "step": 930
677
+ },
678
+ {
679
+ "epoch": 2.1212976022566994,
680
+ "grad_norm": 0.5893935434044923,
681
+ "learning_rate": 5e-06,
682
+ "loss": 0.4838,
683
+ "step": 940
684
+ },
685
+ {
686
+ "epoch": 2.143864598025388,
687
+ "grad_norm": 0.5983033582412839,
688
+ "learning_rate": 5e-06,
689
+ "loss": 0.4855,
690
+ "step": 950
691
+ },
692
+ {
693
+ "epoch": 2.1664315937940763,
694
+ "grad_norm": 0.6061301651629116,
695
+ "learning_rate": 5e-06,
696
+ "loss": 0.482,
697
+ "step": 960
698
+ },
699
+ {
700
+ "epoch": 2.1889985895627646,
701
+ "grad_norm": 0.7208585763313551,
702
+ "learning_rate": 5e-06,
703
+ "loss": 0.48,
704
+ "step": 970
705
+ },
706
+ {
707
+ "epoch": 2.211565585331453,
708
+ "grad_norm": 0.6130737469375042,
709
+ "learning_rate": 5e-06,
710
+ "loss": 0.475,
711
+ "step": 980
712
+ },
713
+ {
714
+ "epoch": 2.234132581100141,
715
+ "grad_norm": 0.5687127566641503,
716
+ "learning_rate": 5e-06,
717
+ "loss": 0.4748,
718
+ "step": 990
719
+ },
720
+ {
721
+ "epoch": 2.2566995768688294,
722
+ "grad_norm": 0.7034322954282024,
723
+ "learning_rate": 5e-06,
724
+ "loss": 0.4839,
725
+ "step": 1000
726
+ },
727
+ {
728
+ "epoch": 2.2792665726375176,
729
+ "grad_norm": 0.5935224377076785,
730
+ "learning_rate": 5e-06,
731
+ "loss": 0.4907,
732
+ "step": 1010
733
+ },
734
+ {
735
+ "epoch": 2.301833568406206,
736
+ "grad_norm": 0.7990706799966594,
737
+ "learning_rate": 5e-06,
738
+ "loss": 0.485,
739
+ "step": 1020
740
+ },
741
+ {
742
+ "epoch": 2.324400564174894,
743
+ "grad_norm": 0.6024785296050031,
744
+ "learning_rate": 5e-06,
745
+ "loss": 0.489,
746
+ "step": 1030
747
+ },
748
+ {
749
+ "epoch": 2.3469675599435824,
750
+ "grad_norm": 0.6533399713336523,
751
+ "learning_rate": 5e-06,
752
+ "loss": 0.487,
753
+ "step": 1040
754
+ },
755
+ {
756
+ "epoch": 2.3695345557122707,
757
+ "grad_norm": 0.6341103214680793,
758
+ "learning_rate": 5e-06,
759
+ "loss": 0.4849,
760
+ "step": 1050
761
+ },
762
+ {
763
+ "epoch": 2.392101551480959,
764
+ "grad_norm": 0.6145632264645421,
765
+ "learning_rate": 5e-06,
766
+ "loss": 0.4844,
767
+ "step": 1060
768
+ },
769
+ {
770
+ "epoch": 2.414668547249647,
771
+ "grad_norm": 0.7153893920420706,
772
+ "learning_rate": 5e-06,
773
+ "loss": 0.4848,
774
+ "step": 1070
775
+ },
776
+ {
777
+ "epoch": 2.4372355430183354,
778
+ "grad_norm": 0.7079536947244722,
779
+ "learning_rate": 5e-06,
780
+ "loss": 0.4883,
781
+ "step": 1080
782
+ },
783
+ {
784
+ "epoch": 2.459802538787024,
785
+ "grad_norm": 0.5750394845236195,
786
+ "learning_rate": 5e-06,
787
+ "loss": 0.4863,
788
+ "step": 1090
789
+ },
790
+ {
791
+ "epoch": 2.4823695345557124,
792
+ "grad_norm": 0.5940313086667759,
793
+ "learning_rate": 5e-06,
794
+ "loss": 0.4896,
795
+ "step": 1100
796
+ },
797
+ {
798
+ "epoch": 2.5049365303244007,
799
+ "grad_norm": 0.5915949703706999,
800
+ "learning_rate": 5e-06,
801
+ "loss": 0.4922,
802
+ "step": 1110
803
+ },
804
+ {
805
+ "epoch": 2.527503526093089,
806
+ "grad_norm": 0.6599431703449483,
807
+ "learning_rate": 5e-06,
808
+ "loss": 0.4838,
809
+ "step": 1120
810
+ },
811
+ {
812
+ "epoch": 2.550070521861777,
813
+ "grad_norm": 0.5987887803366096,
814
+ "learning_rate": 5e-06,
815
+ "loss": 0.4883,
816
+ "step": 1130
817
+ },
818
+ {
819
+ "epoch": 2.5726375176304654,
820
+ "grad_norm": 0.79487550387113,
821
+ "learning_rate": 5e-06,
822
+ "loss": 0.4871,
823
+ "step": 1140
824
+ },
825
+ {
826
+ "epoch": 2.5952045133991537,
827
+ "grad_norm": 0.712422302635071,
828
+ "learning_rate": 5e-06,
829
+ "loss": 0.4877,
830
+ "step": 1150
831
+ },
832
+ {
833
+ "epoch": 2.617771509167842,
834
+ "grad_norm": 0.6372686005869476,
835
+ "learning_rate": 5e-06,
836
+ "loss": 0.4915,
837
+ "step": 1160
838
+ },
839
+ {
840
+ "epoch": 2.64033850493653,
841
+ "grad_norm": 0.6258750499046419,
842
+ "learning_rate": 5e-06,
843
+ "loss": 0.4809,
844
+ "step": 1170
845
+ },
846
+ {
847
+ "epoch": 2.6629055007052185,
848
+ "grad_norm": 0.6254081356701218,
849
+ "learning_rate": 5e-06,
850
+ "loss": 0.4913,
851
+ "step": 1180
852
+ },
853
+ {
854
+ "epoch": 2.685472496473907,
855
+ "grad_norm": 0.5772656389234755,
856
+ "learning_rate": 5e-06,
857
+ "loss": 0.4919,
858
+ "step": 1190
859
+ },
860
+ {
861
+ "epoch": 2.7080394922425954,
862
+ "grad_norm": 0.5958356234239867,
863
+ "learning_rate": 5e-06,
864
+ "loss": 0.4901,
865
+ "step": 1200
866
+ },
867
+ {
868
+ "epoch": 2.7306064880112837,
869
+ "grad_norm": 0.5845924683927192,
870
+ "learning_rate": 5e-06,
871
+ "loss": 0.4913,
872
+ "step": 1210
873
+ },
874
+ {
875
+ "epoch": 2.753173483779972,
876
+ "grad_norm": 0.5627157674386412,
877
+ "learning_rate": 5e-06,
878
+ "loss": 0.4884,
879
+ "step": 1220
880
+ },
881
+ {
882
+ "epoch": 2.77574047954866,
883
+ "grad_norm": 0.5821830745956523,
884
+ "learning_rate": 5e-06,
885
+ "loss": 0.4945,
886
+ "step": 1230
887
+ },
888
+ {
889
+ "epoch": 2.7983074753173485,
890
+ "grad_norm": 0.6229685304395023,
891
+ "learning_rate": 5e-06,
892
+ "loss": 0.4934,
893
+ "step": 1240
894
+ },
895
+ {
896
+ "epoch": 2.8208744710860367,
897
+ "grad_norm": 0.5845914499963646,
898
+ "learning_rate": 5e-06,
899
+ "loss": 0.4933,
900
+ "step": 1250
901
+ },
902
+ {
903
+ "epoch": 2.843441466854725,
904
+ "grad_norm": 0.5574457219985247,
905
+ "learning_rate": 5e-06,
906
+ "loss": 0.495,
907
+ "step": 1260
908
+ },
909
+ {
910
+ "epoch": 2.8660084626234132,
911
+ "grad_norm": 0.5996150224624698,
912
+ "learning_rate": 5e-06,
913
+ "loss": 0.4919,
914
+ "step": 1270
915
+ },
916
+ {
917
+ "epoch": 2.8885754583921015,
918
+ "grad_norm": 0.5752771355797817,
919
+ "learning_rate": 5e-06,
920
+ "loss": 0.4901,
921
+ "step": 1280
922
+ },
923
+ {
924
+ "epoch": 2.9111424541607898,
925
+ "grad_norm": 0.6240736677369982,
926
+ "learning_rate": 5e-06,
927
+ "loss": 0.4938,
928
+ "step": 1290
929
+ },
930
+ {
931
+ "epoch": 2.933709449929478,
932
+ "grad_norm": 0.614687551114313,
933
+ "learning_rate": 5e-06,
934
+ "loss": 0.4889,
935
+ "step": 1300
936
+ },
937
+ {
938
+ "epoch": 2.9562764456981663,
939
+ "grad_norm": 0.6546354143056629,
940
+ "learning_rate": 5e-06,
941
+ "loss": 0.493,
942
+ "step": 1310
943
+ },
944
+ {
945
+ "epoch": 2.9788434414668545,
946
+ "grad_norm": 0.5981369023208123,
947
+ "learning_rate": 5e-06,
948
+ "loss": 0.4899,
949
+ "step": 1320
950
+ },
951
+ {
952
+ "epoch": 2.9991537376586743,
953
+ "eval_loss": 0.6006776094436646,
954
+ "eval_runtime": 359.5275,
955
+ "eval_samples_per_second": 33.213,
956
+ "eval_steps_per_second": 0.52,
957
+ "step": 1329
958
+ },
959
+ {
960
+ "epoch": 3.0014104372355432,
961
+ "grad_norm": 1.7818284423885802,
962
+ "learning_rate": 5e-06,
963
+ "loss": 0.5325,
964
+ "step": 1330
965
+ },
966
+ {
967
+ "epoch": 3.0239774330042315,
968
+ "grad_norm": 0.7669244086926804,
969
+ "learning_rate": 5e-06,
970
+ "loss": 0.4154,
971
+ "step": 1340
972
+ },
973
+ {
974
+ "epoch": 3.0465444287729198,
975
+ "grad_norm": 0.6707220051250778,
976
+ "learning_rate": 5e-06,
977
+ "loss": 0.4178,
978
+ "step": 1350
979
+ },
980
+ {
981
+ "epoch": 3.069111424541608,
982
+ "grad_norm": 0.6712211926030488,
983
+ "learning_rate": 5e-06,
984
+ "loss": 0.4179,
985
+ "step": 1360
986
+ },
987
+ {
988
+ "epoch": 3.0916784203102963,
989
+ "grad_norm": 0.6852453014581671,
990
+ "learning_rate": 5e-06,
991
+ "loss": 0.4199,
992
+ "step": 1370
993
+ },
994
+ {
995
+ "epoch": 3.1142454160789845,
996
+ "grad_norm": 0.6952511595354455,
997
+ "learning_rate": 5e-06,
998
+ "loss": 0.4172,
999
+ "step": 1380
1000
+ },
1001
+ {
1002
+ "epoch": 3.136812411847673,
1003
+ "grad_norm": 0.6739341752053783,
1004
+ "learning_rate": 5e-06,
1005
+ "loss": 0.4196,
1006
+ "step": 1390
1007
+ },
1008
+ {
1009
+ "epoch": 3.159379407616361,
1010
+ "grad_norm": 0.6810240432279709,
1011
+ "learning_rate": 5e-06,
1012
+ "loss": 0.4217,
1013
+ "step": 1400
1014
+ },
1015
+ {
1016
+ "epoch": 3.1819464033850493,
1017
+ "grad_norm": 0.6451044056426664,
1018
+ "learning_rate": 5e-06,
1019
+ "loss": 0.4228,
1020
+ "step": 1410
1021
+ },
1022
+ {
1023
+ "epoch": 3.2045133991537376,
1024
+ "grad_norm": 0.7365176770943456,
1025
+ "learning_rate": 5e-06,
1026
+ "loss": 0.4264,
1027
+ "step": 1420
1028
+ },
1029
+ {
1030
+ "epoch": 3.227080394922426,
1031
+ "grad_norm": 0.6889861565739961,
1032
+ "learning_rate": 5e-06,
1033
+ "loss": 0.4236,
1034
+ "step": 1430
1035
+ },
1036
+ {
1037
+ "epoch": 3.249647390691114,
1038
+ "grad_norm": 0.6619155809062343,
1039
+ "learning_rate": 5e-06,
1040
+ "loss": 0.429,
1041
+ "step": 1440
1042
+ },
1043
+ {
1044
+ "epoch": 3.272214386459803,
1045
+ "grad_norm": 0.6938859033729257,
1046
+ "learning_rate": 5e-06,
1047
+ "loss": 0.4315,
1048
+ "step": 1450
1049
+ },
1050
+ {
1051
+ "epoch": 3.294781382228491,
1052
+ "grad_norm": 0.6269637473408832,
1053
+ "learning_rate": 5e-06,
1054
+ "loss": 0.4311,
1055
+ "step": 1460
1056
+ },
1057
+ {
1058
+ "epoch": 3.3173483779971793,
1059
+ "grad_norm": 0.6871689545622378,
1060
+ "learning_rate": 5e-06,
1061
+ "loss": 0.429,
1062
+ "step": 1470
1063
+ },
1064
+ {
1065
+ "epoch": 3.3399153737658676,
1066
+ "grad_norm": 0.6431121820988837,
1067
+ "learning_rate": 5e-06,
1068
+ "loss": 0.4268,
1069
+ "step": 1480
1070
+ },
1071
+ {
1072
+ "epoch": 3.362482369534556,
1073
+ "grad_norm": 0.6227148520808465,
1074
+ "learning_rate": 5e-06,
1075
+ "loss": 0.4291,
1076
+ "step": 1490
1077
+ },
1078
+ {
1079
+ "epoch": 3.385049365303244,
1080
+ "grad_norm": 0.6926627432900843,
1081
+ "learning_rate": 5e-06,
1082
+ "loss": 0.4369,
1083
+ "step": 1500
1084
+ },
1085
+ {
1086
+ "epoch": 3.4076163610719323,
1087
+ "grad_norm": 0.6753898611704962,
1088
+ "learning_rate": 5e-06,
1089
+ "loss": 0.4289,
1090
+ "step": 1510
1091
+ },
1092
+ {
1093
+ "epoch": 3.4301833568406206,
1094
+ "grad_norm": 0.6235946356953773,
1095
+ "learning_rate": 5e-06,
1096
+ "loss": 0.4285,
1097
+ "step": 1520
1098
+ },
1099
+ {
1100
+ "epoch": 3.452750352609309,
1101
+ "grad_norm": 0.6452950491570345,
1102
+ "learning_rate": 5e-06,
1103
+ "loss": 0.4339,
1104
+ "step": 1530
1105
+ },
1106
+ {
1107
+ "epoch": 3.475317348377997,
1108
+ "grad_norm": 0.6466567653039634,
1109
+ "learning_rate": 5e-06,
1110
+ "loss": 0.4298,
1111
+ "step": 1540
1112
+ },
1113
+ {
1114
+ "epoch": 3.4978843441466854,
1115
+ "grad_norm": 0.7157977999949227,
1116
+ "learning_rate": 5e-06,
1117
+ "loss": 0.4367,
1118
+ "step": 1550
1119
+ },
1120
+ {
1121
+ "epoch": 3.5204513399153736,
1122
+ "grad_norm": 0.746147622100308,
1123
+ "learning_rate": 5e-06,
1124
+ "loss": 0.4333,
1125
+ "step": 1560
1126
+ },
1127
+ {
1128
+ "epoch": 3.543018335684062,
1129
+ "grad_norm": 0.63252883620392,
1130
+ "learning_rate": 5e-06,
1131
+ "loss": 0.4381,
1132
+ "step": 1570
1133
+ },
1134
+ {
1135
+ "epoch": 3.56558533145275,
1136
+ "grad_norm": 0.645023562996112,
1137
+ "learning_rate": 5e-06,
1138
+ "loss": 0.4287,
1139
+ "step": 1580
1140
+ },
1141
+ {
1142
+ "epoch": 3.5881523272214384,
1143
+ "grad_norm": 0.6438099493606649,
1144
+ "learning_rate": 5e-06,
1145
+ "loss": 0.4288,
1146
+ "step": 1590
1147
+ },
1148
+ {
1149
+ "epoch": 3.610719322990127,
1150
+ "grad_norm": 0.6919344965122289,
1151
+ "learning_rate": 5e-06,
1152
+ "loss": 0.4322,
1153
+ "step": 1600
1154
+ },
1155
+ {
1156
+ "epoch": 3.6332863187588154,
1157
+ "grad_norm": 0.6663827038922323,
1158
+ "learning_rate": 5e-06,
1159
+ "loss": 0.4353,
1160
+ "step": 1610
1161
+ },
1162
+ {
1163
+ "epoch": 3.6558533145275036,
1164
+ "grad_norm": 0.6633293647142574,
1165
+ "learning_rate": 5e-06,
1166
+ "loss": 0.4321,
1167
+ "step": 1620
1168
+ },
1169
+ {
1170
+ "epoch": 3.678420310296192,
1171
+ "grad_norm": 0.6568725046800021,
1172
+ "learning_rate": 5e-06,
1173
+ "loss": 0.4352,
1174
+ "step": 1630
1175
+ },
1176
+ {
1177
+ "epoch": 3.70098730606488,
1178
+ "grad_norm": 0.673099496711102,
1179
+ "learning_rate": 5e-06,
1180
+ "loss": 0.436,
1181
+ "step": 1640
1182
+ },
1183
+ {
1184
+ "epoch": 3.7235543018335684,
1185
+ "grad_norm": 0.6438662463205753,
1186
+ "learning_rate": 5e-06,
1187
+ "loss": 0.4321,
1188
+ "step": 1650
1189
+ },
1190
+ {
1191
+ "epoch": 3.7461212976022567,
1192
+ "grad_norm": 0.6523537779817469,
1193
+ "learning_rate": 5e-06,
1194
+ "loss": 0.4308,
1195
+ "step": 1660
1196
+ },
1197
+ {
1198
+ "epoch": 3.768688293370945,
1199
+ "grad_norm": 0.6785421362215835,
1200
+ "learning_rate": 5e-06,
1201
+ "loss": 0.4363,
1202
+ "step": 1670
1203
+ },
1204
+ {
1205
+ "epoch": 3.791255289139633,
1206
+ "grad_norm": 0.6620925087943027,
1207
+ "learning_rate": 5e-06,
1208
+ "loss": 0.4421,
1209
+ "step": 1680
1210
+ },
1211
+ {
1212
+ "epoch": 3.8138222849083214,
1213
+ "grad_norm": 0.6148218995999806,
1214
+ "learning_rate": 5e-06,
1215
+ "loss": 0.4332,
1216
+ "step": 1690
1217
+ },
1218
+ {
1219
+ "epoch": 3.83638928067701,
1220
+ "grad_norm": 0.7122084725034734,
1221
+ "learning_rate": 5e-06,
1222
+ "loss": 0.431,
1223
+ "step": 1700
1224
+ },
1225
+ {
1226
+ "epoch": 3.8589562764456984,
1227
+ "grad_norm": 0.6454980786635541,
1228
+ "learning_rate": 5e-06,
1229
+ "loss": 0.4398,
1230
+ "step": 1710
1231
+ },
1232
+ {
1233
+ "epoch": 3.8815232722143866,
1234
+ "grad_norm": 0.6274368423659055,
1235
+ "learning_rate": 5e-06,
1236
+ "loss": 0.4387,
1237
+ "step": 1720
1238
+ },
1239
+ {
1240
+ "epoch": 3.904090267983075,
1241
+ "grad_norm": 0.6324679361432074,
1242
+ "learning_rate": 5e-06,
1243
+ "loss": 0.4365,
1244
+ "step": 1730
1245
+ },
1246
+ {
1247
+ "epoch": 3.926657263751763,
1248
+ "grad_norm": 0.6304335216014254,
1249
+ "learning_rate": 5e-06,
1250
+ "loss": 0.4373,
1251
+ "step": 1740
1252
+ },
1253
+ {
1254
+ "epoch": 3.9492242595204514,
1255
+ "grad_norm": 0.6460344821987493,
1256
+ "learning_rate": 5e-06,
1257
+ "loss": 0.4366,
1258
+ "step": 1750
1259
+ },
1260
+ {
1261
+ "epoch": 3.9717912552891397,
1262
+ "grad_norm": 0.6439249065834609,
1263
+ "learning_rate": 5e-06,
1264
+ "loss": 0.4362,
1265
+ "step": 1760
1266
+ },
1267
+ {
1268
+ "epoch": 3.994358251057828,
1269
+ "grad_norm": 0.6241101368641281,
1270
+ "learning_rate": 5e-06,
1271
+ "loss": 0.4359,
1272
+ "step": 1770
1273
+ },
1274
+ {
1275
+ "epoch": 3.9988716502115658,
1276
+ "eval_loss": 0.6362190246582031,
1277
+ "eval_runtime": 359.9366,
1278
+ "eval_samples_per_second": 33.175,
1279
+ "eval_steps_per_second": 0.52,
1280
+ "step": 1772
1281
+ },
1282
+ {
1283
+ "epoch": 3.9988716502115658,
1284
+ "step": 1772,
1285
+ "total_flos": 6752101308825600.0,
1286
+ "train_loss": 0.5187085666788351,
1287
+ "train_runtime": 82722.7317,
1288
+ "train_samples_per_second": 10.97,
1289
+ "train_steps_per_second": 0.021
1290
+ }
1291
+ ],
1292
+ "logging_steps": 10,
1293
+ "max_steps": 1772,
1294
+ "num_input_tokens_seen": 0,
1295
+ "num_train_epochs": 4,
1296
+ "save_steps": 500,
1297
+ "stateful_callbacks": {
1298
+ "TrainerControl": {
1299
+ "args": {
1300
+ "should_epoch_stop": false,
1301
+ "should_evaluate": false,
1302
+ "should_log": false,
1303
+ "should_save": true,
1304
+ "should_training_stop": true
1305
+ },
1306
+ "attributes": {}
1307
+ }
1308
+ },
1309
+ "total_flos": 6752101308825600.0,
1310
+ "train_batch_size": 4,
1311
+ "trial_name": null,
1312
+ "trial_params": null
1313
+ }
training_eval_loss.png ADDED
training_loss.png ADDED