Hugging Face's logo

Wonder-Griffin
/
Judge-GPT2

PyTorch
TensorBoard
Safetensors
gpt2
Generated from Trainer
Model card Files Metrics Community
Wonder-Griffin commited on
Commit
eb62f9f
•
1 Parent(s): 4b50879

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. checkpoint-1000/model.safetensors +3 -0
  2. checkpoint-1000/optimizer.pt +3 -0
  3. checkpoint-1000/rng_state.pth +3 -0
  4. checkpoint-1000/scheduler.pt +3 -0
  5. checkpoint-1000/trainer_state.json +733 -0
  6. checkpoint-1000/training_args.bin +3 -0
  7. checkpoint-1500/model.safetensors +3 -0
  8. checkpoint-1500/optimizer.pt +3 -0
  9. checkpoint-1500/rng_state.pth +3 -0
  10. checkpoint-1500/scheduler.pt +3 -0
  11. checkpoint-1500/trainer_state.json +1083 -0
  12. checkpoint-1500/training_args.bin +3 -0
  13. checkpoint-2000/model.safetensors +3 -0
  14. checkpoint-2000/optimizer.pt +3 -0
  15. checkpoint-2000/rng_state.pth +3 -0
  16. checkpoint-2000/scheduler.pt +3 -0
  17. checkpoint-2000/trainer_state.json +1433 -0
  18. checkpoint-2000/training_args.bin +3 -0
  19. config.json +1 -0
  20. merges.txt +0 -0
  21. model.pth +3 -0
  22. runs/Aug01_01-37-45_DadsLapDancer/events.out.tfevents.1722494525.DadsLapDancer.6780.0 +3 -0
  23. runs/Aug01_01-47-58_DadsLapDancer/events.out.tfevents.1722495144.DadsLapDancer.13068.0 +3 -0
  24. runs/Aug01_02-38-07_DadsLapDancer/events.out.tfevents.1722498158.DadsLapDancer.12600.0 +3 -0
  25. runs/Aug01_02-51-57_DadsLapDancer/events.out.tfevents.1722499044.DadsLapDancer.16012.0 +3 -0
  26. runs/Aug01_03-05-35_DadsLapDancer/events.out.tfevents.1722499871.DadsLapDancer.16688.0 +3 -0
  27. runs/Aug01_03-25-55_DadsLapDancer/events.out.tfevents.1722501091.DadsLapDancer.19340.0 +3 -0
  28. runs/Aug01_03-44-45_DadsLapDancer/events.out.tfevents.1722502221.DadsLapDancer.6876.0 +3 -0
  29. runs/Aug01_14-17-09_DadsLapDancer/events.out.tfevents.1722540771.DadsLapDancer.8172.0 +3 -0
  30. runs/Aug01_14-43-24_DadsLapDancer/events.out.tfevents.1722541884.DadsLapDancer.8268.0 +3 -0
  31. runs/Aug01_16-39-05_DadsLapDancer/events.out.tfevents.1722548362.DadsLapDancer.5608.0 +3 -0
  32. runs/Aug01_16-45-01_DadsLapDancer/events.out.tfevents.1722548708.DadsLapDancer.6604.0 +3 -0
  33. runs/Aug01_16-46-45_DadsLapDancer/events.out.tfevents.1722548812.DadsLapDancer.14620.0 +3 -0
  34. runs/Aug01_16-47-33_DadsLapDancer/events.out.tfevents.1722548871.DadsLapDancer.7024.0 +3 -0
  35. runs/Aug01_16-49-14_DadsLapDancer/events.out.tfevents.1722548960.DadsLapDancer.5688.0 +3 -0
  36. runs/Aug01_18-06-27_DadsLapDancer/events.out.tfevents.1722554013.DadsLapDancer.18004.0 +3 -0
  37. runs/Aug01_18-39-26_DadsLapDancer/events.out.tfevents.1722555571.DadsLapDancer.2512.0 +3 -0
  38. runs/Aug01_19-17-43_DadsLapDancer/events.out.tfevents.1722557869.DadsLapDancer.15380.0 +3 -0
  39. runs/Aug01_20-48-10_DadsLapDancer/events.out.tfevents.1722563301.DadsLapDancer.2980.0 +3 -0
  40. runs/Aug01_21-17-07_DadsLapDancer/events.out.tfevents.1722565044.DadsLapDancer.7652.0 +3 -0
  41. runs/Aug01_21-19-14_DadsLapDancer/events.out.tfevents.1722565160.DadsLapDancer.8464.0 +3 -0
  42. runs/Aug01_21-42-11_DadsLapDancer/events.out.tfevents.1722566547.DadsLapDancer.1532.0 +3 -0
  43. runs/Aug01_23-04-07_DadsLapDancer/events.out.tfevents.1722571454.DadsLapDancer.5628.0 +3 -0
  44. runs/Aug01_23-10-54_DadsLapDancer/events.out.tfevents.1722571873.DadsLapDancer.16684.0 +3 -0
  45. runs/Aug01_23-13-15_DadsLapDancer/events.out.tfevents.1722572005.DadsLapDancer.7708.0 +3 -0
  46. runs/Aug01_23-20-59_DadsLapDancer/events.out.tfevents.1722572477.DadsLapDancer.2392.0 +3 -0
  47. runs/Aug01_23-28-28_DadsLapDancer/events.out.tfevents.1722572926.DadsLapDancer.6044.0 +3 -0
  48. runs/Aug01_23-48-43_DadsLapDancer/events.out.tfevents.1722574142.DadsLapDancer.3404.0 +3 -0
  49. runs/Aug02_00-50-38_DadsLapDancer/events.out.tfevents.1722577849.DadsLapDancer.17716.0 +3 -0
  50. runs/Aug02_00-53-03_DadsLapDancer/events.out.tfevents.1722578002.DadsLapDancer.13092.0 +3 -0
checkpoint-1000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b3b08876c224306d8ab098c17163681e7677df4d6d07a7d06b03e492d0ebf8b
3
+ size 903089044
checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6991f85398fe5c16f18c7624e63c69e131d3167ffdcf8abfd8e233fe4dd83cad
3
+ size 1775681497
checkpoint-1000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76a72da01db10d06cc21d3df3f4dec3dc94f60d9ebe28b628f4541c56bb68302
3
+ size 14244
checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee8b965a9c1279372a85c1e8a9d61eb4a18b2d2783ba2c36d2e088d4525253d2
3
+ size 1064
checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,733 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.6397952655150352,
5
+ "eval_steps": 500,
6
+ "global_step": 1000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.006397952655150352,
13
+ "grad_norm": NaN,
14
+ "learning_rate": 4.9893367455747496e-05,
15
+ "loss": 0.0,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.012795905310300703,
20
+ "grad_norm": NaN,
21
+ "learning_rate": 4.978673491149499e-05,
22
+ "loss": 0.0,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.019193857965451054,
27
+ "grad_norm": NaN,
28
+ "learning_rate": 4.968010236724249e-05,
29
+ "loss": 0.0,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.025591810620601407,
34
+ "grad_norm": NaN,
35
+ "learning_rate": 4.9573469822989975e-05,
36
+ "loss": 0.0,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.03198976327575176,
41
+ "grad_norm": NaN,
42
+ "learning_rate": 4.946683727873747e-05,
43
+ "loss": 0.0,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.03838771593090211,
48
+ "grad_norm": NaN,
49
+ "learning_rate": 4.936020473448497e-05,
50
+ "loss": 0.0,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.044785668586052464,
55
+ "grad_norm": NaN,
56
+ "learning_rate": 4.925357219023246e-05,
57
+ "loss": 0.0,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.05118362124120281,
62
+ "grad_norm": NaN,
63
+ "learning_rate": 4.9146939645979955e-05,
64
+ "loss": 0.0,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.05758157389635317,
69
+ "grad_norm": NaN,
70
+ "learning_rate": 4.904030710172745e-05,
71
+ "loss": 0.0,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.06397952655150352,
76
+ "grad_norm": NaN,
77
+ "learning_rate": 4.893367455747495e-05,
78
+ "loss": 0.0,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.07037747920665387,
83
+ "grad_norm": NaN,
84
+ "learning_rate": 4.882704201322244e-05,
85
+ "loss": 0.0,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.07677543186180422,
90
+ "grad_norm": NaN,
91
+ "learning_rate": 4.872040946896993e-05,
92
+ "loss": 0.0,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.08317338451695458,
97
+ "grad_norm": NaN,
98
+ "learning_rate": 4.861377692471742e-05,
99
+ "loss": 0.0,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.08957133717210493,
104
+ "grad_norm": NaN,
105
+ "learning_rate": 4.850714438046492e-05,
106
+ "loss": 0.0,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.09596928982725528,
111
+ "grad_norm": NaN,
112
+ "learning_rate": 4.8400511836212415e-05,
113
+ "loss": 0.0,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.10236724248240563,
118
+ "grad_norm": NaN,
119
+ "learning_rate": 4.829387929195991e-05,
120
+ "loss": 0.0,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.10876519513755598,
125
+ "grad_norm": NaN,
126
+ "learning_rate": 4.81872467477074e-05,
127
+ "loss": 0.0,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.11516314779270634,
132
+ "grad_norm": NaN,
133
+ "learning_rate": 4.80806142034549e-05,
134
+ "loss": 0.0,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.12156110044785669,
139
+ "grad_norm": NaN,
140
+ "learning_rate": 4.797398165920239e-05,
141
+ "loss": 0.0,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.12795905310300704,
146
+ "grad_norm": NaN,
147
+ "learning_rate": 4.786734911494988e-05,
148
+ "loss": 0.0,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.1343570057581574,
153
+ "grad_norm": NaN,
154
+ "learning_rate": 4.7760716570697375e-05,
155
+ "loss": 0.0,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.14075495841330773,
160
+ "grad_norm": NaN,
161
+ "learning_rate": 4.7654084026444875e-05,
162
+ "loss": 0.0,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.1471529110684581,
167
+ "grad_norm": NaN,
168
+ "learning_rate": 4.754745148219237e-05,
169
+ "loss": 0.0,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.15355086372360843,
174
+ "grad_norm": NaN,
175
+ "learning_rate": 4.744081893793986e-05,
176
+ "loss": 0.0,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.1599488163787588,
181
+ "grad_norm": NaN,
182
+ "learning_rate": 4.7334186393687355e-05,
183
+ "loss": 0.0,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.16634676903390916,
188
+ "grad_norm": NaN,
189
+ "learning_rate": 4.7227553849434855e-05,
190
+ "loss": 0.0,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.1727447216890595,
195
+ "grad_norm": NaN,
196
+ "learning_rate": 4.712092130518234e-05,
197
+ "loss": 0.0,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.17914267434420986,
202
+ "grad_norm": NaN,
203
+ "learning_rate": 4.7014288760929834e-05,
204
+ "loss": 0.0,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.1855406269993602,
209
+ "grad_norm": NaN,
210
+ "learning_rate": 4.6907656216677334e-05,
211
+ "loss": 0.0,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.19193857965451055,
216
+ "grad_norm": NaN,
217
+ "learning_rate": 4.680102367242483e-05,
218
+ "loss": 0.0,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.19833653230966092,
223
+ "grad_norm": NaN,
224
+ "learning_rate": 4.669439112817232e-05,
225
+ "loss": 0.0,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.20473448496481125,
230
+ "grad_norm": NaN,
231
+ "learning_rate": 4.6587758583919814e-05,
232
+ "loss": 0.0,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.21113243761996162,
237
+ "grad_norm": NaN,
238
+ "learning_rate": 4.6481126039667314e-05,
239
+ "loss": 0.0,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.21753039027511195,
244
+ "grad_norm": NaN,
245
+ "learning_rate": 4.63744934954148e-05,
246
+ "loss": 0.0,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 0.22392834293026231,
251
+ "grad_norm": NaN,
252
+ "learning_rate": 4.6267860951162294e-05,
253
+ "loss": 0.0,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 0.23032629558541268,
258
+ "grad_norm": NaN,
259
+ "learning_rate": 4.616122840690979e-05,
260
+ "loss": 0.0,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 0.236724248240563,
265
+ "grad_norm": NaN,
266
+ "learning_rate": 4.605459586265729e-05,
267
+ "loss": 0.0,
268
+ "step": 370
269
+ },
270
+ {
271
+ "epoch": 0.24312220089571338,
272
+ "grad_norm": NaN,
273
+ "learning_rate": 4.594796331840478e-05,
274
+ "loss": 0.0,
275
+ "step": 380
276
+ },
277
+ {
278
+ "epoch": 0.2495201535508637,
279
+ "grad_norm": NaN,
280
+ "learning_rate": 4.5841330774152274e-05,
281
+ "loss": 0.0,
282
+ "step": 390
283
+ },
284
+ {
285
+ "epoch": 0.2559181062060141,
286
+ "grad_norm": NaN,
287
+ "learning_rate": 4.573469822989977e-05,
288
+ "loss": 0.0,
289
+ "step": 400
290
+ },
291
+ {
292
+ "epoch": 0.26231605886116444,
293
+ "grad_norm": NaN,
294
+ "learning_rate": 4.562806568564727e-05,
295
+ "loss": 0.0,
296
+ "step": 410
297
+ },
298
+ {
299
+ "epoch": 0.2687140115163148,
300
+ "grad_norm": NaN,
301
+ "learning_rate": 4.5521433141394754e-05,
302
+ "loss": 0.0,
303
+ "step": 420
304
+ },
305
+ {
306
+ "epoch": 0.2751119641714651,
307
+ "grad_norm": NaN,
308
+ "learning_rate": 4.541480059714225e-05,
309
+ "loss": 0.0,
310
+ "step": 430
311
+ },
312
+ {
313
+ "epoch": 0.28150991682661547,
314
+ "grad_norm": NaN,
315
+ "learning_rate": 4.530816805288974e-05,
316
+ "loss": 0.0,
317
+ "step": 440
318
+ },
319
+ {
320
+ "epoch": 0.28790786948176583,
321
+ "grad_norm": NaN,
322
+ "learning_rate": 4.520153550863724e-05,
323
+ "loss": 0.0,
324
+ "step": 450
325
+ },
326
+ {
327
+ "epoch": 0.2943058221369162,
328
+ "grad_norm": NaN,
329
+ "learning_rate": 4.5094902964384734e-05,
330
+ "loss": 0.0,
331
+ "step": 460
332
+ },
333
+ {
334
+ "epoch": 0.30070377479206656,
335
+ "grad_norm": NaN,
336
+ "learning_rate": 4.498827042013223e-05,
337
+ "loss": 0.0,
338
+ "step": 470
339
+ },
340
+ {
341
+ "epoch": 0.30710172744721687,
342
+ "grad_norm": NaN,
343
+ "learning_rate": 4.488163787587972e-05,
344
+ "loss": 0.0,
345
+ "step": 480
346
+ },
347
+ {
348
+ "epoch": 0.31349968010236723,
349
+ "grad_norm": NaN,
350
+ "learning_rate": 4.477500533162721e-05,
351
+ "loss": 0.0,
352
+ "step": 490
353
+ },
354
+ {
355
+ "epoch": 0.3198976327575176,
356
+ "grad_norm": NaN,
357
+ "learning_rate": 4.466837278737471e-05,
358
+ "loss": 0.0,
359
+ "step": 500
360
+ },
361
+ {
362
+ "epoch": 0.32629558541266795,
363
+ "grad_norm": NaN,
364
+ "learning_rate": 4.45617402431222e-05,
365
+ "loss": 0.0,
366
+ "step": 510
367
+ },
368
+ {
369
+ "epoch": 0.3326935380678183,
370
+ "grad_norm": NaN,
371
+ "learning_rate": 4.44551076988697e-05,
372
+ "loss": 0.0,
373
+ "step": 520
374
+ },
375
+ {
376
+ "epoch": 0.3390914907229686,
377
+ "grad_norm": NaN,
378
+ "learning_rate": 4.434847515461719e-05,
379
+ "loss": 0.0,
380
+ "step": 530
381
+ },
382
+ {
383
+ "epoch": 0.345489443378119,
384
+ "grad_norm": NaN,
385
+ "learning_rate": 4.4241842610364687e-05,
386
+ "loss": 0.0,
387
+ "step": 540
388
+ },
389
+ {
390
+ "epoch": 0.35188739603326935,
391
+ "grad_norm": NaN,
392
+ "learning_rate": 4.413521006611218e-05,
393
+ "loss": 0.0,
394
+ "step": 550
395
+ },
396
+ {
397
+ "epoch": 0.3582853486884197,
398
+ "grad_norm": NaN,
399
+ "learning_rate": 4.402857752185967e-05,
400
+ "loss": 0.0,
401
+ "step": 560
402
+ },
403
+ {
404
+ "epoch": 0.3646833013435701,
405
+ "grad_norm": NaN,
406
+ "learning_rate": 4.3921944977607166e-05,
407
+ "loss": 0.0,
408
+ "step": 570
409
+ },
410
+ {
411
+ "epoch": 0.3710812539987204,
412
+ "grad_norm": NaN,
413
+ "learning_rate": 4.381531243335466e-05,
414
+ "loss": 0.0,
415
+ "step": 580
416
+ },
417
+ {
418
+ "epoch": 0.37747920665387075,
419
+ "grad_norm": NaN,
420
+ "learning_rate": 4.370867988910215e-05,
421
+ "loss": 0.0,
422
+ "step": 590
423
+ },
424
+ {
425
+ "epoch": 0.3838771593090211,
426
+ "grad_norm": NaN,
427
+ "learning_rate": 4.360204734484965e-05,
428
+ "loss": 0.0,
429
+ "step": 600
430
+ },
431
+ {
432
+ "epoch": 0.3902751119641715,
433
+ "grad_norm": NaN,
434
+ "learning_rate": 4.3495414800597146e-05,
435
+ "loss": 0.0,
436
+ "step": 610
437
+ },
438
+ {
439
+ "epoch": 0.39667306461932184,
440
+ "grad_norm": NaN,
441
+ "learning_rate": 4.338878225634464e-05,
442
+ "loss": 0.0,
443
+ "step": 620
444
+ },
445
+ {
446
+ "epoch": 0.40307101727447214,
447
+ "grad_norm": NaN,
448
+ "learning_rate": 4.328214971209213e-05,
449
+ "loss": 0.0,
450
+ "step": 630
451
+ },
452
+ {
453
+ "epoch": 0.4094689699296225,
454
+ "grad_norm": NaN,
455
+ "learning_rate": 4.3175517167839626e-05,
456
+ "loss": 0.0,
457
+ "step": 640
458
+ },
459
+ {
460
+ "epoch": 0.41586692258477287,
461
+ "grad_norm": NaN,
462
+ "learning_rate": 4.306888462358712e-05,
463
+ "loss": 0.0,
464
+ "step": 650
465
+ },
466
+ {
467
+ "epoch": 0.42226487523992323,
468
+ "grad_norm": NaN,
469
+ "learning_rate": 4.296225207933461e-05,
470
+ "loss": 0.0,
471
+ "step": 660
472
+ },
473
+ {
474
+ "epoch": 0.4286628278950736,
475
+ "grad_norm": NaN,
476
+ "learning_rate": 4.2855619535082106e-05,
477
+ "loss": 0.0,
478
+ "step": 670
479
+ },
480
+ {
481
+ "epoch": 0.4350607805502239,
482
+ "grad_norm": NaN,
483
+ "learning_rate": 4.2748986990829606e-05,
484
+ "loss": 0.0,
485
+ "step": 680
486
+ },
487
+ {
488
+ "epoch": 0.44145873320537427,
489
+ "grad_norm": NaN,
490
+ "learning_rate": 4.26423544465771e-05,
491
+ "loss": 0.0,
492
+ "step": 690
493
+ },
494
+ {
495
+ "epoch": 0.44785668586052463,
496
+ "grad_norm": NaN,
497
+ "learning_rate": 4.253572190232459e-05,
498
+ "loss": 0.0,
499
+ "step": 700
500
+ },
501
+ {
502
+ "epoch": 0.454254638515675,
503
+ "grad_norm": NaN,
504
+ "learning_rate": 4.2429089358072086e-05,
505
+ "loss": 0.0,
506
+ "step": 710
507
+ },
508
+ {
509
+ "epoch": 0.46065259117082535,
510
+ "grad_norm": NaN,
511
+ "learning_rate": 4.232245681381958e-05,
512
+ "loss": 0.0,
513
+ "step": 720
514
+ },
515
+ {
516
+ "epoch": 0.46705054382597566,
517
+ "grad_norm": NaN,
518
+ "learning_rate": 4.221582426956707e-05,
519
+ "loss": 0.0,
520
+ "step": 730
521
+ },
522
+ {
523
+ "epoch": 0.473448496481126,
524
+ "grad_norm": NaN,
525
+ "learning_rate": 4.2109191725314565e-05,
526
+ "loss": 0.0,
527
+ "step": 740
528
+ },
529
+ {
530
+ "epoch": 0.4798464491362764,
531
+ "grad_norm": NaN,
532
+ "learning_rate": 4.2002559181062066e-05,
533
+ "loss": 0.0,
534
+ "step": 750
535
+ },
536
+ {
537
+ "epoch": 0.48624440179142675,
538
+ "grad_norm": NaN,
539
+ "learning_rate": 4.189592663680956e-05,
540
+ "loss": 0.0,
541
+ "step": 760
542
+ },
543
+ {
544
+ "epoch": 0.4926423544465771,
545
+ "grad_norm": NaN,
546
+ "learning_rate": 4.178929409255705e-05,
547
+ "loss": 0.0,
548
+ "step": 770
549
+ },
550
+ {
551
+ "epoch": 0.4990403071017274,
552
+ "grad_norm": NaN,
553
+ "learning_rate": 4.1682661548304545e-05,
554
+ "loss": 0.0,
555
+ "step": 780
556
+ },
557
+ {
558
+ "epoch": 0.5054382597568778,
559
+ "grad_norm": NaN,
560
+ "learning_rate": 4.157602900405204e-05,
561
+ "loss": 0.0,
562
+ "step": 790
563
+ },
564
+ {
565
+ "epoch": 0.5118362124120281,
566
+ "grad_norm": NaN,
567
+ "learning_rate": 4.146939645979953e-05,
568
+ "loss": 0.0,
569
+ "step": 800
570
+ },
571
+ {
572
+ "epoch": 0.5182341650671785,
573
+ "grad_norm": NaN,
574
+ "learning_rate": 4.1362763915547025e-05,
575
+ "loss": 0.0,
576
+ "step": 810
577
+ },
578
+ {
579
+ "epoch": 0.5246321177223289,
580
+ "grad_norm": NaN,
581
+ "learning_rate": 4.125613137129452e-05,
582
+ "loss": 0.0,
583
+ "step": 820
584
+ },
585
+ {
586
+ "epoch": 0.5310300703774792,
587
+ "grad_norm": NaN,
588
+ "learning_rate": 4.114949882704202e-05,
589
+ "loss": 0.0,
590
+ "step": 830
591
+ },
592
+ {
593
+ "epoch": 0.5374280230326296,
594
+ "grad_norm": NaN,
595
+ "learning_rate": 4.104286628278951e-05,
596
+ "loss": 0.0,
597
+ "step": 840
598
+ },
599
+ {
600
+ "epoch": 0.5438259756877799,
601
+ "grad_norm": NaN,
602
+ "learning_rate": 4.0936233738537005e-05,
603
+ "loss": 0.0,
604
+ "step": 850
605
+ },
606
+ {
607
+ "epoch": 0.5502239283429302,
608
+ "grad_norm": NaN,
609
+ "learning_rate": 4.08296011942845e-05,
610
+ "loss": 0.0,
611
+ "step": 860
612
+ },
613
+ {
614
+ "epoch": 0.5566218809980806,
615
+ "grad_norm": NaN,
616
+ "learning_rate": 4.072296865003199e-05,
617
+ "loss": 0.0,
618
+ "step": 870
619
+ },
620
+ {
621
+ "epoch": 0.5630198336532309,
622
+ "grad_norm": NaN,
623
+ "learning_rate": 4.0616336105779485e-05,
624
+ "loss": 0.0,
625
+ "step": 880
626
+ },
627
+ {
628
+ "epoch": 0.5694177863083814,
629
+ "grad_norm": NaN,
630
+ "learning_rate": 4.050970356152698e-05,
631
+ "loss": 0.0,
632
+ "step": 890
633
+ },
634
+ {
635
+ "epoch": 0.5758157389635317,
636
+ "grad_norm": NaN,
637
+ "learning_rate": 4.040307101727447e-05,
638
+ "loss": 0.0,
639
+ "step": 900
640
+ },
641
+ {
642
+ "epoch": 0.582213691618682,
643
+ "grad_norm": NaN,
644
+ "learning_rate": 4.029643847302197e-05,
645
+ "loss": 0.0,
646
+ "step": 910
647
+ },
648
+ {
649
+ "epoch": 0.5886116442738324,
650
+ "grad_norm": NaN,
651
+ "learning_rate": 4.0189805928769465e-05,
652
+ "loss": 0.0,
653
+ "step": 920
654
+ },
655
+ {
656
+ "epoch": 0.5950095969289827,
657
+ "grad_norm": NaN,
658
+ "learning_rate": 4.008317338451696e-05,
659
+ "loss": 0.0,
660
+ "step": 930
661
+ },
662
+ {
663
+ "epoch": 0.6014075495841331,
664
+ "grad_norm": NaN,
665
+ "learning_rate": 3.997654084026445e-05,
666
+ "loss": 0.0,
667
+ "step": 940
668
+ },
669
+ {
670
+ "epoch": 0.6078055022392834,
671
+ "grad_norm": NaN,
672
+ "learning_rate": 3.9869908296011945e-05,
673
+ "loss": 0.0,
674
+ "step": 950
675
+ },
676
+ {
677
+ "epoch": 0.6142034548944337,
678
+ "grad_norm": NaN,
679
+ "learning_rate": 3.976327575175944e-05,
680
+ "loss": 0.0,
681
+ "step": 960
682
+ },
683
+ {
684
+ "epoch": 0.6206014075495841,
685
+ "grad_norm": NaN,
686
+ "learning_rate": 3.965664320750693e-05,
687
+ "loss": 0.0,
688
+ "step": 970
689
+ },
690
+ {
691
+ "epoch": 0.6269993602047345,
692
+ "grad_norm": NaN,
693
+ "learning_rate": 3.9550010663254424e-05,
694
+ "loss": 0.0,
695
+ "step": 980
696
+ },
697
+ {
698
+ "epoch": 0.6333973128598849,
699
+ "grad_norm": NaN,
700
+ "learning_rate": 3.9443378119001924e-05,
701
+ "loss": 0.0,
702
+ "step": 990
703
+ },
704
+ {
705
+ "epoch": 0.6397952655150352,
706
+ "grad_norm": NaN,
707
+ "learning_rate": 3.933674557474942e-05,
708
+ "loss": 0.0,
709
+ "step": 1000
710
+ }
711
+ ],
712
+ "logging_steps": 10,
713
+ "max_steps": 4689,
714
+ "num_input_tokens_seen": 0,
715
+ "num_train_epochs": 3,
716
+ "save_steps": 500,
717
+ "stateful_callbacks": {
718
+ "TrainerControl": {
719
+ "args": {
720
+ "should_epoch_stop": false,
721
+ "should_evaluate": false,
722
+ "should_log": false,
723
+ "should_save": true,
724
+ "should_training_stop": false
725
+ },
726
+ "attributes": {}
727
+ }
728
+ },
729
+ "total_flos": 0.0,
730
+ "train_batch_size": 16,
731
+ "trial_name": null,
732
+ "trial_params": null
733
+ }
checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ceebcdd4aa5d5e688dc20d333b0c1fc1d54fc83ca77f4049b33fecb99c128b6
3
+ size 5112
checkpoint-1500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b3b08876c224306d8ab098c17163681e7677df4d6d07a7d06b03e492d0ebf8b
3
+ size 903089044
checkpoint-1500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8458250101d468d5a684da9dd04bf1f47f47d04508cb08c1db23742383d9d14b
3
+ size 1775681497
checkpoint-1500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b08ecdd5c584661a30800caeb873f24aebf0cc1bb97324033fd0025242f7d2f
3
+ size 14244
checkpoint-1500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:595e75b1cc83f0fa6a8e63dc3a3505ec5dd465379b9971d537aa837499a29df8
3
+ size 1064
checkpoint-1500/trainer_state.json ADDED
@@ -0,0 +1,1083 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9596928982725528,
5
+ "eval_steps": 500,
6
+ "global_step": 1500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.006397952655150352,
13
+ "grad_norm": NaN,
14
+ "learning_rate": 4.9893367455747496e-05,
15
+ "loss": 0.0,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.012795905310300703,
20
+ "grad_norm": NaN,
21
+ "learning_rate": 4.978673491149499e-05,
22
+ "loss": 0.0,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.019193857965451054,
27
+ "grad_norm": NaN,
28
+ "learning_rate": 4.968010236724249e-05,
29
+ "loss": 0.0,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.025591810620601407,
34
+ "grad_norm": NaN,
35
+ "learning_rate": 4.9573469822989975e-05,
36
+ "loss": 0.0,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.03198976327575176,
41
+ "grad_norm": NaN,
42
+ "learning_rate": 4.946683727873747e-05,
43
+ "loss": 0.0,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.03838771593090211,
48
+ "grad_norm": NaN,
49
+ "learning_rate": 4.936020473448497e-05,
50
+ "loss": 0.0,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.044785668586052464,
55
+ "grad_norm": NaN,
56
+ "learning_rate": 4.925357219023246e-05,
57
+ "loss": 0.0,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.05118362124120281,
62
+ "grad_norm": NaN,
63
+ "learning_rate": 4.9146939645979955e-05,
64
+ "loss": 0.0,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.05758157389635317,
69
+ "grad_norm": NaN,
70
+ "learning_rate": 4.904030710172745e-05,
71
+ "loss": 0.0,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.06397952655150352,
76
+ "grad_norm": NaN,
77
+ "learning_rate": 4.893367455747495e-05,
78
+ "loss": 0.0,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.07037747920665387,
83
+ "grad_norm": NaN,
84
+ "learning_rate": 4.882704201322244e-05,
85
+ "loss": 0.0,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.07677543186180422,
90
+ "grad_norm": NaN,
91
+ "learning_rate": 4.872040946896993e-05,
92
+ "loss": 0.0,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.08317338451695458,
97
+ "grad_norm": NaN,
98
+ "learning_rate": 4.861377692471742e-05,
99
+ "loss": 0.0,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.08957133717210493,
104
+ "grad_norm": NaN,
105
+ "learning_rate": 4.850714438046492e-05,
106
+ "loss": 0.0,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.09596928982725528,
111
+ "grad_norm": NaN,
112
+ "learning_rate": 4.8400511836212415e-05,
113
+ "loss": 0.0,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.10236724248240563,
118
+ "grad_norm": NaN,
119
+ "learning_rate": 4.829387929195991e-05,
120
+ "loss": 0.0,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.10876519513755598,
125
+ "grad_norm": NaN,
126
+ "learning_rate": 4.81872467477074e-05,
127
+ "loss": 0.0,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.11516314779270634,
132
+ "grad_norm": NaN,
133
+ "learning_rate": 4.80806142034549e-05,
134
+ "loss": 0.0,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.12156110044785669,
139
+ "grad_norm": NaN,
140
+ "learning_rate": 4.797398165920239e-05,
141
+ "loss": 0.0,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.12795905310300704,
146
+ "grad_norm": NaN,
147
+ "learning_rate": 4.786734911494988e-05,
148
+ "loss": 0.0,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.1343570057581574,
153
+ "grad_norm": NaN,
154
+ "learning_rate": 4.7760716570697375e-05,
155
+ "loss": 0.0,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.14075495841330773,
160
+ "grad_norm": NaN,
161
+ "learning_rate": 4.7654084026444875e-05,
162
+ "loss": 0.0,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.1471529110684581,
167
+ "grad_norm": NaN,
168
+ "learning_rate": 4.754745148219237e-05,
169
+ "loss": 0.0,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.15355086372360843,
174
+ "grad_norm": NaN,
175
+ "learning_rate": 4.744081893793986e-05,
176
+ "loss": 0.0,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.1599488163787588,
181
+ "grad_norm": NaN,
182
+ "learning_rate": 4.7334186393687355e-05,
183
+ "loss": 0.0,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.16634676903390916,
188
+ "grad_norm": NaN,
189
+ "learning_rate": 4.7227553849434855e-05,
190
+ "loss": 0.0,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.1727447216890595,
195
+ "grad_norm": NaN,
196
+ "learning_rate": 4.712092130518234e-05,
197
+ "loss": 0.0,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.17914267434420986,
202
+ "grad_norm": NaN,
203
+ "learning_rate": 4.7014288760929834e-05,
204
+ "loss": 0.0,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.1855406269993602,
209
+ "grad_norm": NaN,
210
+ "learning_rate": 4.6907656216677334e-05,
211
+ "loss": 0.0,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.19193857965451055,
216
+ "grad_norm": NaN,
217
+ "learning_rate": 4.680102367242483e-05,
218
+ "loss": 0.0,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.19833653230966092,
223
+ "grad_norm": NaN,
224
+ "learning_rate": 4.669439112817232e-05,
225
+ "loss": 0.0,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.20473448496481125,
230
+ "grad_norm": NaN,
231
+ "learning_rate": 4.6587758583919814e-05,
232
+ "loss": 0.0,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.21113243761996162,
237
+ "grad_norm": NaN,
238
+ "learning_rate": 4.6481126039667314e-05,
239
+ "loss": 0.0,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.21753039027511195,
244
+ "grad_norm": NaN,
245
+ "learning_rate": 4.63744934954148e-05,
246
+ "loss": 0.0,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 0.22392834293026231,
251
+ "grad_norm": NaN,
252
+ "learning_rate": 4.6267860951162294e-05,
253
+ "loss": 0.0,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 0.23032629558541268,
258
+ "grad_norm": NaN,
259
+ "learning_rate": 4.616122840690979e-05,
260
+ "loss": 0.0,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 0.236724248240563,
265
+ "grad_norm": NaN,
266
+ "learning_rate": 4.605459586265729e-05,
267
+ "loss": 0.0,
268
+ "step": 370
269
+ },
270
+ {
271
+ "epoch": 0.24312220089571338,
272
+ "grad_norm": NaN,
273
+ "learning_rate": 4.594796331840478e-05,
274
+ "loss": 0.0,
275
+ "step": 380
276
+ },
277
+ {
278
+ "epoch": 0.2495201535508637,
279
+ "grad_norm": NaN,
280
+ "learning_rate": 4.5841330774152274e-05,
281
+ "loss": 0.0,
282
+ "step": 390
283
+ },
284
+ {
285
+ "epoch": 0.2559181062060141,
286
+ "grad_norm": NaN,
287
+ "learning_rate": 4.573469822989977e-05,
288
+ "loss": 0.0,
289
+ "step": 400
290
+ },
291
+ {
292
+ "epoch": 0.26231605886116444,
293
+ "grad_norm": NaN,
294
+ "learning_rate": 4.562806568564727e-05,
295
+ "loss": 0.0,
296
+ "step": 410
297
+ },
298
+ {
299
+ "epoch": 0.2687140115163148,
300
+ "grad_norm": NaN,
301
+ "learning_rate": 4.5521433141394754e-05,
302
+ "loss": 0.0,
303
+ "step": 420
304
+ },
305
+ {
306
+ "epoch": 0.2751119641714651,
307
+ "grad_norm": NaN,
308
+ "learning_rate": 4.541480059714225e-05,
309
+ "loss": 0.0,
310
+ "step": 430
311
+ },
312
+ {
313
+ "epoch": 0.28150991682661547,
314
+ "grad_norm": NaN,
315
+ "learning_rate": 4.530816805288974e-05,
316
+ "loss": 0.0,
317
+ "step": 440
318
+ },
319
+ {
320
+ "epoch": 0.28790786948176583,
321
+ "grad_norm": NaN,
322
+ "learning_rate": 4.520153550863724e-05,
323
+ "loss": 0.0,
324
+ "step": 450
325
+ },
326
+ {
327
+ "epoch": 0.2943058221369162,
328
+ "grad_norm": NaN,
329
+ "learning_rate": 4.5094902964384734e-05,
330
+ "loss": 0.0,
331
+ "step": 460
332
+ },
333
+ {
334
+ "epoch": 0.30070377479206656,
335
+ "grad_norm": NaN,
336
+ "learning_rate": 4.498827042013223e-05,
337
+ "loss": 0.0,
338
+ "step": 470
339
+ },
340
+ {
341
+ "epoch": 0.30710172744721687,
342
+ "grad_norm": NaN,
343
+ "learning_rate": 4.488163787587972e-05,
344
+ "loss": 0.0,
345
+ "step": 480
346
+ },
347
+ {
348
+ "epoch": 0.31349968010236723,
349
+ "grad_norm": NaN,
350
+ "learning_rate": 4.477500533162721e-05,
351
+ "loss": 0.0,
352
+ "step": 490
353
+ },
354
+ {
355
+ "epoch": 0.3198976327575176,
356
+ "grad_norm": NaN,
357
+ "learning_rate": 4.466837278737471e-05,
358
+ "loss": 0.0,
359
+ "step": 500
360
+ },
361
+ {
362
+ "epoch": 0.32629558541266795,
363
+ "grad_norm": NaN,
364
+ "learning_rate": 4.45617402431222e-05,
365
+ "loss": 0.0,
366
+ "step": 510
367
+ },
368
+ {
369
+ "epoch": 0.3326935380678183,
370
+ "grad_norm": NaN,
371
+ "learning_rate": 4.44551076988697e-05,
372
+ "loss": 0.0,
373
+ "step": 520
374
+ },
375
+ {
376
+ "epoch": 0.3390914907229686,
377
+ "grad_norm": NaN,
378
+ "learning_rate": 4.434847515461719e-05,
379
+ "loss": 0.0,
380
+ "step": 530
381
+ },
382
+ {
383
+ "epoch": 0.345489443378119,
384
+ "grad_norm": NaN,
385
+ "learning_rate": 4.4241842610364687e-05,
386
+ "loss": 0.0,
387
+ "step": 540
388
+ },
389
+ {
390
+ "epoch": 0.35188739603326935,
391
+ "grad_norm": NaN,
392
+ "learning_rate": 4.413521006611218e-05,
393
+ "loss": 0.0,
394
+ "step": 550
395
+ },
396
+ {
397
+ "epoch": 0.3582853486884197,
398
+ "grad_norm": NaN,
399
+ "learning_rate": 4.402857752185967e-05,
400
+ "loss": 0.0,
401
+ "step": 560
402
+ },
403
+ {
404
+ "epoch": 0.3646833013435701,
405
+ "grad_norm": NaN,
406
+ "learning_rate": 4.3921944977607166e-05,
407
+ "loss": 0.0,
408
+ "step": 570
409
+ },
410
+ {
411
+ "epoch": 0.3710812539987204,
412
+ "grad_norm": NaN,
413
+ "learning_rate": 4.381531243335466e-05,
414
+ "loss": 0.0,
415
+ "step": 580
416
+ },
417
+ {
418
+ "epoch": 0.37747920665387075,
419
+ "grad_norm": NaN,
420
+ "learning_rate": 4.370867988910215e-05,
421
+ "loss": 0.0,
422
+ "step": 590
423
+ },
424
+ {
425
+ "epoch": 0.3838771593090211,
426
+ "grad_norm": NaN,
427
+ "learning_rate": 4.360204734484965e-05,
428
+ "loss": 0.0,
429
+ "step": 600
430
+ },
431
+ {
432
+ "epoch": 0.3902751119641715,
433
+ "grad_norm": NaN,
434
+ "learning_rate": 4.3495414800597146e-05,
435
+ "loss": 0.0,
436
+ "step": 610
437
+ },
438
+ {
439
+ "epoch": 0.39667306461932184,
440
+ "grad_norm": NaN,
441
+ "learning_rate": 4.338878225634464e-05,
442
+ "loss": 0.0,
443
+ "step": 620
444
+ },
445
+ {
446
+ "epoch": 0.40307101727447214,
447
+ "grad_norm": NaN,
448
+ "learning_rate": 4.328214971209213e-05,
449
+ "loss": 0.0,
450
+ "step": 630
451
+ },
452
+ {
453
+ "epoch": 0.4094689699296225,
454
+ "grad_norm": NaN,
455
+ "learning_rate": 4.3175517167839626e-05,
456
+ "loss": 0.0,
457
+ "step": 640
458
+ },
459
+ {
460
+ "epoch": 0.41586692258477287,
461
+ "grad_norm": NaN,
462
+ "learning_rate": 4.306888462358712e-05,
463
+ "loss": 0.0,
464
+ "step": 650
465
+ },
466
+ {
467
+ "epoch": 0.42226487523992323,
468
+ "grad_norm": NaN,
469
+ "learning_rate": 4.296225207933461e-05,
470
+ "loss": 0.0,
471
+ "step": 660
472
+ },
473
+ {
474
+ "epoch": 0.4286628278950736,
475
+ "grad_norm": NaN,
476
+ "learning_rate": 4.2855619535082106e-05,
477
+ "loss": 0.0,
478
+ "step": 670
479
+ },
480
+ {
481
+ "epoch": 0.4350607805502239,
482
+ "grad_norm": NaN,
483
+ "learning_rate": 4.2748986990829606e-05,
484
+ "loss": 0.0,
485
+ "step": 680
486
+ },
487
+ {
488
+ "epoch": 0.44145873320537427,
489
+ "grad_norm": NaN,
490
+ "learning_rate": 4.26423544465771e-05,
491
+ "loss": 0.0,
492
+ "step": 690
493
+ },
494
+ {
495
+ "epoch": 0.44785668586052463,
496
+ "grad_norm": NaN,
497
+ "learning_rate": 4.253572190232459e-05,
498
+ "loss": 0.0,
499
+ "step": 700
500
+ },
501
+ {
502
+ "epoch": 0.454254638515675,
503
+ "grad_norm": NaN,
504
+ "learning_rate": 4.2429089358072086e-05,
505
+ "loss": 0.0,
506
+ "step": 710
507
+ },
508
+ {
509
+ "epoch": 0.46065259117082535,
510
+ "grad_norm": NaN,
511
+ "learning_rate": 4.232245681381958e-05,
512
+ "loss": 0.0,
513
+ "step": 720
514
+ },
515
+ {
516
+ "epoch": 0.46705054382597566,
517
+ "grad_norm": NaN,
518
+ "learning_rate": 4.221582426956707e-05,
519
+ "loss": 0.0,
520
+ "step": 730
521
+ },
522
+ {
523
+ "epoch": 0.473448496481126,
524
+ "grad_norm": NaN,
525
+ "learning_rate": 4.2109191725314565e-05,
526
+ "loss": 0.0,
527
+ "step": 740
528
+ },
529
+ {
530
+ "epoch": 0.4798464491362764,
531
+ "grad_norm": NaN,
532
+ "learning_rate": 4.2002559181062066e-05,
533
+ "loss": 0.0,
534
+ "step": 750
535
+ },
536
+ {
537
+ "epoch": 0.48624440179142675,
538
+ "grad_norm": NaN,
539
+ "learning_rate": 4.189592663680956e-05,
540
+ "loss": 0.0,
541
+ "step": 760
542
+ },
543
+ {
544
+ "epoch": 0.4926423544465771,
545
+ "grad_norm": NaN,
546
+ "learning_rate": 4.178929409255705e-05,
547
+ "loss": 0.0,
548
+ "step": 770
549
+ },
550
+ {
551
+ "epoch": 0.4990403071017274,
552
+ "grad_norm": NaN,
553
+ "learning_rate": 4.1682661548304545e-05,
554
+ "loss": 0.0,
555
+ "step": 780
556
+ },
557
+ {
558
+ "epoch": 0.5054382597568778,
559
+ "grad_norm": NaN,
560
+ "learning_rate": 4.157602900405204e-05,
561
+ "loss": 0.0,
562
+ "step": 790
563
+ },
564
+ {
565
+ "epoch": 0.5118362124120281,
566
+ "grad_norm": NaN,
567
+ "learning_rate": 4.146939645979953e-05,
568
+ "loss": 0.0,
569
+ "step": 800
570
+ },
571
+ {
572
+ "epoch": 0.5182341650671785,
573
+ "grad_norm": NaN,
574
+ "learning_rate": 4.1362763915547025e-05,
575
+ "loss": 0.0,
576
+ "step": 810
577
+ },
578
+ {
579
+ "epoch": 0.5246321177223289,
580
+ "grad_norm": NaN,
581
+ "learning_rate": 4.125613137129452e-05,
582
+ "loss": 0.0,
583
+ "step": 820
584
+ },
585
+ {
586
+ "epoch": 0.5310300703774792,
587
+ "grad_norm": NaN,
588
+ "learning_rate": 4.114949882704202e-05,
589
+ "loss": 0.0,
590
+ "step": 830
591
+ },
592
+ {
593
+ "epoch": 0.5374280230326296,
594
+ "grad_norm": NaN,
595
+ "learning_rate": 4.104286628278951e-05,
596
+ "loss": 0.0,
597
+ "step": 840
598
+ },
599
+ {
600
+ "epoch": 0.5438259756877799,
601
+ "grad_norm": NaN,
602
+ "learning_rate": 4.0936233738537005e-05,
603
+ "loss": 0.0,
604
+ "step": 850
605
+ },
606
+ {
607
+ "epoch": 0.5502239283429302,
608
+ "grad_norm": NaN,
609
+ "learning_rate": 4.08296011942845e-05,
610
+ "loss": 0.0,
611
+ "step": 860
612
+ },
613
+ {
614
+ "epoch": 0.5566218809980806,
615
+ "grad_norm": NaN,
616
+ "learning_rate": 4.072296865003199e-05,
617
+ "loss": 0.0,
618
+ "step": 870
619
+ },
620
+ {
621
+ "epoch": 0.5630198336532309,
622
+ "grad_norm": NaN,
623
+ "learning_rate": 4.0616336105779485e-05,
624
+ "loss": 0.0,
625
+ "step": 880
626
+ },
627
+ {
628
+ "epoch": 0.5694177863083814,
629
+ "grad_norm": NaN,
630
+ "learning_rate": 4.050970356152698e-05,
631
+ "loss": 0.0,
632
+ "step": 890
633
+ },
634
+ {
635
+ "epoch": 0.5758157389635317,
636
+ "grad_norm": NaN,
637
+ "learning_rate": 4.040307101727447e-05,
638
+ "loss": 0.0,
639
+ "step": 900
640
+ },
641
+ {
642
+ "epoch": 0.582213691618682,
643
+ "grad_norm": NaN,
644
+ "learning_rate": 4.029643847302197e-05,
645
+ "loss": 0.0,
646
+ "step": 910
647
+ },
648
+ {
649
+ "epoch": 0.5886116442738324,
650
+ "grad_norm": NaN,
651
+ "learning_rate": 4.0189805928769465e-05,
652
+ "loss": 0.0,
653
+ "step": 920
654
+ },
655
+ {
656
+ "epoch": 0.5950095969289827,
657
+ "grad_norm": NaN,
658
+ "learning_rate": 4.008317338451696e-05,
659
+ "loss": 0.0,
660
+ "step": 930
661
+ },
662
+ {
663
+ "epoch": 0.6014075495841331,
664
+ "grad_norm": NaN,
665
+ "learning_rate": 3.997654084026445e-05,
666
+ "loss": 0.0,
667
+ "step": 940
668
+ },
669
+ {
670
+ "epoch": 0.6078055022392834,
671
+ "grad_norm": NaN,
672
+ "learning_rate": 3.9869908296011945e-05,
673
+ "loss": 0.0,
674
+ "step": 950
675
+ },
676
+ {
677
+ "epoch": 0.6142034548944337,
678
+ "grad_norm": NaN,
679
+ "learning_rate": 3.976327575175944e-05,
680
+ "loss": 0.0,
681
+ "step": 960
682
+ },
683
+ {
684
+ "epoch": 0.6206014075495841,
685
+ "grad_norm": NaN,
686
+ "learning_rate": 3.965664320750693e-05,
687
+ "loss": 0.0,
688
+ "step": 970
689
+ },
690
+ {
691
+ "epoch": 0.6269993602047345,
692
+ "grad_norm": NaN,
693
+ "learning_rate": 3.9550010663254424e-05,
694
+ "loss": 0.0,
695
+ "step": 980
696
+ },
697
+ {
698
+ "epoch": 0.6333973128598849,
699
+ "grad_norm": NaN,
700
+ "learning_rate": 3.9443378119001924e-05,
701
+ "loss": 0.0,
702
+ "step": 990
703
+ },
704
+ {
705
+ "epoch": 0.6397952655150352,
706
+ "grad_norm": NaN,
707
+ "learning_rate": 3.933674557474942e-05,
708
+ "loss": 0.0,
709
+ "step": 1000
710
+ },
711
+ {
712
+ "epoch": 0.6461932181701855,
713
+ "grad_norm": NaN,
714
+ "learning_rate": 3.923011303049691e-05,
715
+ "loss": 0.0,
716
+ "step": 1010
717
+ },
718
+ {
719
+ "epoch": 0.6525911708253359,
720
+ "grad_norm": NaN,
721
+ "learning_rate": 3.9123480486244404e-05,
722
+ "loss": 0.0,
723
+ "step": 1020
724
+ },
725
+ {
726
+ "epoch": 0.6589891234804862,
727
+ "grad_norm": NaN,
728
+ "learning_rate": 3.90168479419919e-05,
729
+ "loss": 0.0,
730
+ "step": 1030
731
+ },
732
+ {
733
+ "epoch": 0.6653870761356366,
734
+ "grad_norm": NaN,
735
+ "learning_rate": 3.891021539773939e-05,
736
+ "loss": 0.0,
737
+ "step": 1040
738
+ },
739
+ {
740
+ "epoch": 0.6717850287907869,
741
+ "grad_norm": NaN,
742
+ "learning_rate": 3.8803582853486884e-05,
743
+ "loss": 0.0,
744
+ "step": 1050
745
+ },
746
+ {
747
+ "epoch": 0.6781829814459372,
748
+ "grad_norm": NaN,
749
+ "learning_rate": 3.8696950309234384e-05,
750
+ "loss": 0.0,
751
+ "step": 1060
752
+ },
753
+ {
754
+ "epoch": 0.6845809341010877,
755
+ "grad_norm": NaN,
756
+ "learning_rate": 3.859031776498188e-05,
757
+ "loss": 0.0,
758
+ "step": 1070
759
+ },
760
+ {
761
+ "epoch": 0.690978886756238,
762
+ "grad_norm": NaN,
763
+ "learning_rate": 3.848368522072937e-05,
764
+ "loss": 0.0,
765
+ "step": 1080
766
+ },
767
+ {
768
+ "epoch": 0.6973768394113884,
769
+ "grad_norm": NaN,
770
+ "learning_rate": 3.837705267647686e-05,
771
+ "loss": 0.0,
772
+ "step": 1090
773
+ },
774
+ {
775
+ "epoch": 0.7037747920665387,
776
+ "grad_norm": NaN,
777
+ "learning_rate": 3.827042013222436e-05,
778
+ "loss": 0.0,
779
+ "step": 1100
780
+ },
781
+ {
782
+ "epoch": 0.710172744721689,
783
+ "grad_norm": NaN,
784
+ "learning_rate": 3.816378758797185e-05,
785
+ "loss": 0.0,
786
+ "step": 1110
787
+ },
788
+ {
789
+ "epoch": 0.7165706973768394,
790
+ "grad_norm": NaN,
791
+ "learning_rate": 3.8057155043719344e-05,
792
+ "loss": 0.0,
793
+ "step": 1120
794
+ },
795
+ {
796
+ "epoch": 0.7229686500319897,
797
+ "grad_norm": NaN,
798
+ "learning_rate": 3.795052249946684e-05,
799
+ "loss": 0.0,
800
+ "step": 1130
801
+ },
802
+ {
803
+ "epoch": 0.7293666026871402,
804
+ "grad_norm": NaN,
805
+ "learning_rate": 3.784388995521434e-05,
806
+ "loss": 0.0,
807
+ "step": 1140
808
+ },
809
+ {
810
+ "epoch": 0.7357645553422905,
811
+ "grad_norm": NaN,
812
+ "learning_rate": 3.773725741096183e-05,
813
+ "loss": 0.0,
814
+ "step": 1150
815
+ },
816
+ {
817
+ "epoch": 0.7421625079974408,
818
+ "grad_norm": NaN,
819
+ "learning_rate": 3.7630624866709324e-05,
820
+ "loss": 0.0,
821
+ "step": 1160
822
+ },
823
+ {
824
+ "epoch": 0.7485604606525912,
825
+ "grad_norm": NaN,
826
+ "learning_rate": 3.752399232245682e-05,
827
+ "loss": 0.0,
828
+ "step": 1170
829
+ },
830
+ {
831
+ "epoch": 0.7549584133077415,
832
+ "grad_norm": NaN,
833
+ "learning_rate": 3.741735977820431e-05,
834
+ "loss": 0.0,
835
+ "step": 1180
836
+ },
837
+ {
838
+ "epoch": 0.7613563659628919,
839
+ "grad_norm": NaN,
840
+ "learning_rate": 3.73107272339518e-05,
841
+ "loss": 0.0,
842
+ "step": 1190
843
+ },
844
+ {
845
+ "epoch": 0.7677543186180422,
846
+ "grad_norm": NaN,
847
+ "learning_rate": 3.72040946896993e-05,
848
+ "loss": 0.0,
849
+ "step": 1200
850
+ },
851
+ {
852
+ "epoch": 0.7741522712731925,
853
+ "grad_norm": NaN,
854
+ "learning_rate": 3.709746214544679e-05,
855
+ "loss": 0.0,
856
+ "step": 1210
857
+ },
858
+ {
859
+ "epoch": 0.780550223928343,
860
+ "grad_norm": NaN,
861
+ "learning_rate": 3.699082960119429e-05,
862
+ "loss": 0.0,
863
+ "step": 1220
864
+ },
865
+ {
866
+ "epoch": 0.7869481765834933,
867
+ "grad_norm": NaN,
868
+ "learning_rate": 3.688419705694178e-05,
869
+ "loss": 0.0,
870
+ "step": 1230
871
+ },
872
+ {
873
+ "epoch": 0.7933461292386437,
874
+ "grad_norm": NaN,
875
+ "learning_rate": 3.677756451268927e-05,
876
+ "loss": 0.0,
877
+ "step": 1240
878
+ },
879
+ {
880
+ "epoch": 0.799744081893794,
881
+ "grad_norm": NaN,
882
+ "learning_rate": 3.667093196843677e-05,
883
+ "loss": 0.0,
884
+ "step": 1250
885
+ },
886
+ {
887
+ "epoch": 0.8061420345489443,
888
+ "grad_norm": NaN,
889
+ "learning_rate": 3.656429942418426e-05,
890
+ "loss": 0.0,
891
+ "step": 1260
892
+ },
893
+ {
894
+ "epoch": 0.8125399872040947,
895
+ "grad_norm": NaN,
896
+ "learning_rate": 3.6457666879931756e-05,
897
+ "loss": 0.0,
898
+ "step": 1270
899
+ },
900
+ {
901
+ "epoch": 0.818937939859245,
902
+ "grad_norm": NaN,
903
+ "learning_rate": 3.635103433567925e-05,
904
+ "loss": 0.0,
905
+ "step": 1280
906
+ },
907
+ {
908
+ "epoch": 0.8253358925143954,
909
+ "grad_norm": NaN,
910
+ "learning_rate": 3.624440179142675e-05,
911
+ "loss": 0.0,
912
+ "step": 1290
913
+ },
914
+ {
915
+ "epoch": 0.8317338451695457,
916
+ "grad_norm": NaN,
917
+ "learning_rate": 3.613776924717424e-05,
918
+ "loss": 0.0,
919
+ "step": 1300
920
+ },
921
+ {
922
+ "epoch": 0.838131797824696,
923
+ "grad_norm": NaN,
924
+ "learning_rate": 3.6031136702921736e-05,
925
+ "loss": 0.0,
926
+ "step": 1310
927
+ },
928
+ {
929
+ "epoch": 0.8445297504798465,
930
+ "grad_norm": NaN,
931
+ "learning_rate": 3.592450415866922e-05,
932
+ "loss": 0.0,
933
+ "step": 1320
934
+ },
935
+ {
936
+ "epoch": 0.8509277031349968,
937
+ "grad_norm": NaN,
938
+ "learning_rate": 3.581787161441672e-05,
939
+ "loss": 0.0,
940
+ "step": 1330
941
+ },
942
+ {
943
+ "epoch": 0.8573256557901472,
944
+ "grad_norm": NaN,
945
+ "learning_rate": 3.5711239070164216e-05,
946
+ "loss": 0.0,
947
+ "step": 1340
948
+ },
949
+ {
950
+ "epoch": 0.8637236084452975,
951
+ "grad_norm": NaN,
952
+ "learning_rate": 3.560460652591171e-05,
953
+ "loss": 0.0,
954
+ "step": 1350
955
+ },
956
+ {
957
+ "epoch": 0.8701215611004478,
958
+ "grad_norm": NaN,
959
+ "learning_rate": 3.54979739816592e-05,
960
+ "loss": 0.0,
961
+ "step": 1360
962
+ },
963
+ {
964
+ "epoch": 0.8765195137555982,
965
+ "grad_norm": NaN,
966
+ "learning_rate": 3.53913414374067e-05,
967
+ "loss": 0.0,
968
+ "step": 1370
969
+ },
970
+ {
971
+ "epoch": 0.8829174664107485,
972
+ "grad_norm": NaN,
973
+ "learning_rate": 3.5284708893154196e-05,
974
+ "loss": 0.0,
975
+ "step": 1380
976
+ },
977
+ {
978
+ "epoch": 0.889315419065899,
979
+ "grad_norm": NaN,
980
+ "learning_rate": 3.517807634890168e-05,
981
+ "loss": 0.0,
982
+ "step": 1390
983
+ },
984
+ {
985
+ "epoch": 0.8957133717210493,
986
+ "grad_norm": NaN,
987
+ "learning_rate": 3.5071443804649176e-05,
988
+ "loss": 0.0,
989
+ "step": 1400
990
+ },
991
+ {
992
+ "epoch": 0.9021113243761996,
993
+ "grad_norm": NaN,
994
+ "learning_rate": 3.4964811260396676e-05,
995
+ "loss": 0.0,
996
+ "step": 1410
997
+ },
998
+ {
999
+ "epoch": 0.90850927703135,
1000
+ "grad_norm": NaN,
1001
+ "learning_rate": 3.485817871614417e-05,
1002
+ "loss": 0.0,
1003
+ "step": 1420
1004
+ },
1005
+ {
1006
+ "epoch": 0.9149072296865003,
1007
+ "grad_norm": NaN,
1008
+ "learning_rate": 3.475154617189166e-05,
1009
+ "loss": 0.0,
1010
+ "step": 1430
1011
+ },
1012
+ {
1013
+ "epoch": 0.9213051823416507,
1014
+ "grad_norm": NaN,
1015
+ "learning_rate": 3.4644913627639155e-05,
1016
+ "loss": 0.0,
1017
+ "step": 1440
1018
+ },
1019
+ {
1020
+ "epoch": 0.927703134996801,
1021
+ "grad_norm": NaN,
1022
+ "learning_rate": 3.4538281083386656e-05,
1023
+ "loss": 0.0,
1024
+ "step": 1450
1025
+ },
1026
+ {
1027
+ "epoch": 0.9341010876519513,
1028
+ "grad_norm": NaN,
1029
+ "learning_rate": 3.443164853913415e-05,
1030
+ "loss": 0.0,
1031
+ "step": 1460
1032
+ },
1033
+ {
1034
+ "epoch": 0.9404990403071017,
1035
+ "grad_norm": NaN,
1036
+ "learning_rate": 3.4325015994881635e-05,
1037
+ "loss": 0.0,
1038
+ "step": 1470
1039
+ },
1040
+ {
1041
+ "epoch": 0.946896992962252,
1042
+ "grad_norm": NaN,
1043
+ "learning_rate": 3.4218383450629135e-05,
1044
+ "loss": 0.0,
1045
+ "step": 1480
1046
+ },
1047
+ {
1048
+ "epoch": 0.9532949456174025,
1049
+ "grad_norm": NaN,
1050
+ "learning_rate": 3.411175090637663e-05,
1051
+ "loss": 0.0,
1052
+ "step": 1490
1053
+ },
1054
+ {
1055
+ "epoch": 0.9596928982725528,
1056
+ "grad_norm": NaN,
1057
+ "learning_rate": 3.400511836212412e-05,
1058
+ "loss": 0.0,
1059
+ "step": 1500
1060
+ }
1061
+ ],
1062
+ "logging_steps": 10,
1063
+ "max_steps": 4689,
1064
+ "num_input_tokens_seen": 0,
1065
+ "num_train_epochs": 3,
1066
+ "save_steps": 500,
1067
+ "stateful_callbacks": {
1068
+ "TrainerControl": {
1069
+ "args": {
1070
+ "should_epoch_stop": false,
1071
+ "should_evaluate": false,
1072
+ "should_log": false,
1073
+ "should_save": true,
1074
+ "should_training_stop": false
1075
+ },
1076
+ "attributes": {}
1077
+ }
1078
+ },
1079
+ "total_flos": 0.0,
1080
+ "train_batch_size": 16,
1081
+ "trial_name": null,
1082
+ "trial_params": null
1083
+ }
checkpoint-1500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ceebcdd4aa5d5e688dc20d333b0c1fc1d54fc83ca77f4049b33fecb99c128b6
3
+ size 5112
checkpoint-2000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b3b08876c224306d8ab098c17163681e7677df4d6d07a7d06b03e492d0ebf8b
3
+ size 903089044
checkpoint-2000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:abca7179c1432db468aafc040425c8c97a53d6c93f2d80c25a225566dbf8185f
3
+ size 1775681497
checkpoint-2000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45c986df20a5cef4a7a8e12357cbd1adea7882c9a77c9b655f9cee0cc03947c9
3
+ size 14244
checkpoint-2000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac6520a2cc7cc0579869fcd346d4431c4a030bb649bbfff85edd13cedd110447
3
+ size 1064
checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,1433 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.2795905310300704,
5
+ "eval_steps": 500,
6
+ "global_step": 2000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.006397952655150352,
13
+ "grad_norm": NaN,
14
+ "learning_rate": 4.9893367455747496e-05,
15
+ "loss": 0.0,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.012795905310300703,
20
+ "grad_norm": NaN,
21
+ "learning_rate": 4.978673491149499e-05,
22
+ "loss": 0.0,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.019193857965451054,
27
+ "grad_norm": NaN,
28
+ "learning_rate": 4.968010236724249e-05,
29
+ "loss": 0.0,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.025591810620601407,
34
+ "grad_norm": NaN,
35
+ "learning_rate": 4.9573469822989975e-05,
36
+ "loss": 0.0,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.03198976327575176,
41
+ "grad_norm": NaN,
42
+ "learning_rate": 4.946683727873747e-05,
43
+ "loss": 0.0,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.03838771593090211,
48
+ "grad_norm": NaN,
49
+ "learning_rate": 4.936020473448497e-05,
50
+ "loss": 0.0,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.044785668586052464,
55
+ "grad_norm": NaN,
56
+ "learning_rate": 4.925357219023246e-05,
57
+ "loss": 0.0,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.05118362124120281,
62
+ "grad_norm": NaN,
63
+ "learning_rate": 4.9146939645979955e-05,
64
+ "loss": 0.0,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.05758157389635317,
69
+ "grad_norm": NaN,
70
+ "learning_rate": 4.904030710172745e-05,
71
+ "loss": 0.0,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.06397952655150352,
76
+ "grad_norm": NaN,
77
+ "learning_rate": 4.893367455747495e-05,
78
+ "loss": 0.0,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.07037747920665387,
83
+ "grad_norm": NaN,
84
+ "learning_rate": 4.882704201322244e-05,
85
+ "loss": 0.0,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.07677543186180422,
90
+ "grad_norm": NaN,
91
+ "learning_rate": 4.872040946896993e-05,
92
+ "loss": 0.0,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.08317338451695458,
97
+ "grad_norm": NaN,
98
+ "learning_rate": 4.861377692471742e-05,
99
+ "loss": 0.0,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.08957133717210493,
104
+ "grad_norm": NaN,
105
+ "learning_rate": 4.850714438046492e-05,
106
+ "loss": 0.0,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.09596928982725528,
111
+ "grad_norm": NaN,
112
+ "learning_rate": 4.8400511836212415e-05,
113
+ "loss": 0.0,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.10236724248240563,
118
+ "grad_norm": NaN,
119
+ "learning_rate": 4.829387929195991e-05,
120
+ "loss": 0.0,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.10876519513755598,
125
+ "grad_norm": NaN,
126
+ "learning_rate": 4.81872467477074e-05,
127
+ "loss": 0.0,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.11516314779270634,
132
+ "grad_norm": NaN,
133
+ "learning_rate": 4.80806142034549e-05,
134
+ "loss": 0.0,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.12156110044785669,
139
+ "grad_norm": NaN,
140
+ "learning_rate": 4.797398165920239e-05,
141
+ "loss": 0.0,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.12795905310300704,
146
+ "grad_norm": NaN,
147
+ "learning_rate": 4.786734911494988e-05,
148
+ "loss": 0.0,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.1343570057581574,
153
+ "grad_norm": NaN,
154
+ "learning_rate": 4.7760716570697375e-05,
155
+ "loss": 0.0,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.14075495841330773,
160
+ "grad_norm": NaN,
161
+ "learning_rate": 4.7654084026444875e-05,
162
+ "loss": 0.0,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.1471529110684581,
167
+ "grad_norm": NaN,
168
+ "learning_rate": 4.754745148219237e-05,
169
+ "loss": 0.0,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.15355086372360843,
174
+ "grad_norm": NaN,
175
+ "learning_rate": 4.744081893793986e-05,
176
+ "loss": 0.0,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.1599488163787588,
181
+ "grad_norm": NaN,
182
+ "learning_rate": 4.7334186393687355e-05,
183
+ "loss": 0.0,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.16634676903390916,
188
+ "grad_norm": NaN,
189
+ "learning_rate": 4.7227553849434855e-05,
190
+ "loss": 0.0,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.1727447216890595,
195
+ "grad_norm": NaN,
196
+ "learning_rate": 4.712092130518234e-05,
197
+ "loss": 0.0,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.17914267434420986,
202
+ "grad_norm": NaN,
203
+ "learning_rate": 4.7014288760929834e-05,
204
+ "loss": 0.0,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.1855406269993602,
209
+ "grad_norm": NaN,
210
+ "learning_rate": 4.6907656216677334e-05,
211
+ "loss": 0.0,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.19193857965451055,
216
+ "grad_norm": NaN,
217
+ "learning_rate": 4.680102367242483e-05,
218
+ "loss": 0.0,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.19833653230966092,
223
+ "grad_norm": NaN,
224
+ "learning_rate": 4.669439112817232e-05,
225
+ "loss": 0.0,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.20473448496481125,
230
+ "grad_norm": NaN,
231
+ "learning_rate": 4.6587758583919814e-05,
232
+ "loss": 0.0,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.21113243761996162,
237
+ "grad_norm": NaN,
238
+ "learning_rate": 4.6481126039667314e-05,
239
+ "loss": 0.0,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.21753039027511195,
244
+ "grad_norm": NaN,
245
+ "learning_rate": 4.63744934954148e-05,
246
+ "loss": 0.0,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 0.22392834293026231,
251
+ "grad_norm": NaN,
252
+ "learning_rate": 4.6267860951162294e-05,
253
+ "loss": 0.0,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 0.23032629558541268,
258
+ "grad_norm": NaN,
259
+ "learning_rate": 4.616122840690979e-05,
260
+ "loss": 0.0,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 0.236724248240563,
265
+ "grad_norm": NaN,
266
+ "learning_rate": 4.605459586265729e-05,
267
+ "loss": 0.0,
268
+ "step": 370
269
+ },
270
+ {
271
+ "epoch": 0.24312220089571338,
272
+ "grad_norm": NaN,
273
+ "learning_rate": 4.594796331840478e-05,
274
+ "loss": 0.0,
275
+ "step": 380
276
+ },
277
+ {
278
+ "epoch": 0.2495201535508637,
279
+ "grad_norm": NaN,
280
+ "learning_rate": 4.5841330774152274e-05,
281
+ "loss": 0.0,
282
+ "step": 390
283
+ },
284
+ {
285
+ "epoch": 0.2559181062060141,
286
+ "grad_norm": NaN,
287
+ "learning_rate": 4.573469822989977e-05,
288
+ "loss": 0.0,
289
+ "step": 400
290
+ },
291
+ {
292
+ "epoch": 0.26231605886116444,
293
+ "grad_norm": NaN,
294
+ "learning_rate": 4.562806568564727e-05,
295
+ "loss": 0.0,
296
+ "step": 410
297
+ },
298
+ {
299
+ "epoch": 0.2687140115163148,
300
+ "grad_norm": NaN,
301
+ "learning_rate": 4.5521433141394754e-05,
302
+ "loss": 0.0,
303
+ "step": 420
304
+ },
305
+ {
306
+ "epoch": 0.2751119641714651,
307
+ "grad_norm": NaN,
308
+ "learning_rate": 4.541480059714225e-05,
309
+ "loss": 0.0,
310
+ "step": 430
311
+ },
312
+ {
313
+ "epoch": 0.28150991682661547,
314
+ "grad_norm": NaN,
315
+ "learning_rate": 4.530816805288974e-05,
316
+ "loss": 0.0,
317
+ "step": 440
318
+ },
319
+ {
320
+ "epoch": 0.28790786948176583,
321
+ "grad_norm": NaN,
322
+ "learning_rate": 4.520153550863724e-05,
323
+ "loss": 0.0,
324
+ "step": 450
325
+ },
326
+ {
327
+ "epoch": 0.2943058221369162,
328
+ "grad_norm": NaN,
329
+ "learning_rate": 4.5094902964384734e-05,
330
+ "loss": 0.0,
331
+ "step": 460
332
+ },
333
+ {
334
+ "epoch": 0.30070377479206656,
335
+ "grad_norm": NaN,
336
+ "learning_rate": 4.498827042013223e-05,
337
+ "loss": 0.0,
338
+ "step": 470
339
+ },
340
+ {
341
+ "epoch": 0.30710172744721687,
342
+ "grad_norm": NaN,
343
+ "learning_rate": 4.488163787587972e-05,
344
+ "loss": 0.0,
345
+ "step": 480
346
+ },
347
+ {
348
+ "epoch": 0.31349968010236723,
349
+ "grad_norm": NaN,
350
+ "learning_rate": 4.477500533162721e-05,
351
+ "loss": 0.0,
352
+ "step": 490
353
+ },
354
+ {
355
+ "epoch": 0.3198976327575176,
356
+ "grad_norm": NaN,
357
+ "learning_rate": 4.466837278737471e-05,
358
+ "loss": 0.0,
359
+ "step": 500
360
+ },
361
+ {
362
+ "epoch": 0.32629558541266795,
363
+ "grad_norm": NaN,
364
+ "learning_rate": 4.45617402431222e-05,
365
+ "loss": 0.0,
366
+ "step": 510
367
+ },
368
+ {
369
+ "epoch": 0.3326935380678183,
370
+ "grad_norm": NaN,
371
+ "learning_rate": 4.44551076988697e-05,
372
+ "loss": 0.0,
373
+ "step": 520
374
+ },
375
+ {
376
+ "epoch": 0.3390914907229686,
377
+ "grad_norm": NaN,
378
+ "learning_rate": 4.434847515461719e-05,
379
+ "loss": 0.0,
380
+ "step": 530
381
+ },
382
+ {
383
+ "epoch": 0.345489443378119,
384
+ "grad_norm": NaN,
385
+ "learning_rate": 4.4241842610364687e-05,
386
+ "loss": 0.0,
387
+ "step": 540
388
+ },
389
+ {
390
+ "epoch": 0.35188739603326935,
391
+ "grad_norm": NaN,
392
+ "learning_rate": 4.413521006611218e-05,
393
+ "loss": 0.0,
394
+ "step": 550
395
+ },
396
+ {
397
+ "epoch": 0.3582853486884197,
398
+ "grad_norm": NaN,
399
+ "learning_rate": 4.402857752185967e-05,
400
+ "loss": 0.0,
401
+ "step": 560
402
+ },
403
+ {
404
+ "epoch": 0.3646833013435701,
405
+ "grad_norm": NaN,
406
+ "learning_rate": 4.3921944977607166e-05,
407
+ "loss": 0.0,
408
+ "step": 570
409
+ },
410
+ {
411
+ "epoch": 0.3710812539987204,
412
+ "grad_norm": NaN,
413
+ "learning_rate": 4.381531243335466e-05,
414
+ "loss": 0.0,
415
+ "step": 580
416
+ },
417
+ {
418
+ "epoch": 0.37747920665387075,
419
+ "grad_norm": NaN,
420
+ "learning_rate": 4.370867988910215e-05,
421
+ "loss": 0.0,
422
+ "step": 590
423
+ },
424
+ {
425
+ "epoch": 0.3838771593090211,
426
+ "grad_norm": NaN,
427
+ "learning_rate": 4.360204734484965e-05,
428
+ "loss": 0.0,
429
+ "step": 600
430
+ },
431
+ {
432
+ "epoch": 0.3902751119641715,
433
+ "grad_norm": NaN,
434
+ "learning_rate": 4.3495414800597146e-05,
435
+ "loss": 0.0,
436
+ "step": 610
437
+ },
438
+ {
439
+ "epoch": 0.39667306461932184,
440
+ "grad_norm": NaN,
441
+ "learning_rate": 4.338878225634464e-05,
442
+ "loss": 0.0,
443
+ "step": 620
444
+ },
445
+ {
446
+ "epoch": 0.40307101727447214,
447
+ "grad_norm": NaN,
448
+ "learning_rate": 4.328214971209213e-05,
449
+ "loss": 0.0,
450
+ "step": 630
451
+ },
452
+ {
453
+ "epoch": 0.4094689699296225,
454
+ "grad_norm": NaN,
455
+ "learning_rate": 4.3175517167839626e-05,
456
+ "loss": 0.0,
457
+ "step": 640
458
+ },
459
+ {
460
+ "epoch": 0.41586692258477287,
461
+ "grad_norm": NaN,
462
+ "learning_rate": 4.306888462358712e-05,
463
+ "loss": 0.0,
464
+ "step": 650
465
+ },
466
+ {
467
+ "epoch": 0.42226487523992323,
468
+ "grad_norm": NaN,
469
+ "learning_rate": 4.296225207933461e-05,
470
+ "loss": 0.0,
471
+ "step": 660
472
+ },
473
+ {
474
+ "epoch": 0.4286628278950736,
475
+ "grad_norm": NaN,
476
+ "learning_rate": 4.2855619535082106e-05,
477
+ "loss": 0.0,
478
+ "step": 670
479
+ },
480
+ {
481
+ "epoch": 0.4350607805502239,
482
+ "grad_norm": NaN,
483
+ "learning_rate": 4.2748986990829606e-05,
484
+ "loss": 0.0,
485
+ "step": 680
486
+ },
487
+ {
488
+ "epoch": 0.44145873320537427,
489
+ "grad_norm": NaN,
490
+ "learning_rate": 4.26423544465771e-05,
491
+ "loss": 0.0,
492
+ "step": 690
493
+ },
494
+ {
495
+ "epoch": 0.44785668586052463,
496
+ "grad_norm": NaN,
497
+ "learning_rate": 4.253572190232459e-05,
498
+ "loss": 0.0,
499
+ "step": 700
500
+ },
501
+ {
502
+ "epoch": 0.454254638515675,
503
+ "grad_norm": NaN,
504
+ "learning_rate": 4.2429089358072086e-05,
505
+ "loss": 0.0,
506
+ "step": 710
507
+ },
508
+ {
509
+ "epoch": 0.46065259117082535,
510
+ "grad_norm": NaN,
511
+ "learning_rate": 4.232245681381958e-05,
512
+ "loss": 0.0,
513
+ "step": 720
514
+ },
515
+ {
516
+ "epoch": 0.46705054382597566,
517
+ "grad_norm": NaN,
518
+ "learning_rate": 4.221582426956707e-05,
519
+ "loss": 0.0,
520
+ "step": 730
521
+ },
522
+ {
523
+ "epoch": 0.473448496481126,
524
+ "grad_norm": NaN,
525
+ "learning_rate": 4.2109191725314565e-05,
526
+ "loss": 0.0,
527
+ "step": 740
528
+ },
529
+ {
530
+ "epoch": 0.4798464491362764,
531
+ "grad_norm": NaN,
532
+ "learning_rate": 4.2002559181062066e-05,
533
+ "loss": 0.0,
534
+ "step": 750
535
+ },
536
+ {
537
+ "epoch": 0.48624440179142675,
538
+ "grad_norm": NaN,
539
+ "learning_rate": 4.189592663680956e-05,
540
+ "loss": 0.0,
541
+ "step": 760
542
+ },
543
+ {
544
+ "epoch": 0.4926423544465771,
545
+ "grad_norm": NaN,
546
+ "learning_rate": 4.178929409255705e-05,
547
+ "loss": 0.0,
548
+ "step": 770
549
+ },
550
+ {
551
+ "epoch": 0.4990403071017274,
552
+ "grad_norm": NaN,
553
+ "learning_rate": 4.1682661548304545e-05,
554
+ "loss": 0.0,
555
+ "step": 780
556
+ },
557
+ {
558
+ "epoch": 0.5054382597568778,
559
+ "grad_norm": NaN,
560
+ "learning_rate": 4.157602900405204e-05,
561
+ "loss": 0.0,
562
+ "step": 790
563
+ },
564
+ {
565
+ "epoch": 0.5118362124120281,
566
+ "grad_norm": NaN,
567
+ "learning_rate": 4.146939645979953e-05,
568
+ "loss": 0.0,
569
+ "step": 800
570
+ },
571
+ {
572
+ "epoch": 0.5182341650671785,
573
+ "grad_norm": NaN,
574
+ "learning_rate": 4.1362763915547025e-05,
575
+ "loss": 0.0,
576
+ "step": 810
577
+ },
578
+ {
579
+ "epoch": 0.5246321177223289,
580
+ "grad_norm": NaN,
581
+ "learning_rate": 4.125613137129452e-05,
582
+ "loss": 0.0,
583
+ "step": 820
584
+ },
585
+ {
586
+ "epoch": 0.5310300703774792,
587
+ "grad_norm": NaN,
588
+ "learning_rate": 4.114949882704202e-05,
589
+ "loss": 0.0,
590
+ "step": 830
591
+ },
592
+ {
593
+ "epoch": 0.5374280230326296,
594
+ "grad_norm": NaN,
595
+ "learning_rate": 4.104286628278951e-05,
596
+ "loss": 0.0,
597
+ "step": 840
598
+ },
599
+ {
600
+ "epoch": 0.5438259756877799,
601
+ "grad_norm": NaN,
602
+ "learning_rate": 4.0936233738537005e-05,
603
+ "loss": 0.0,
604
+ "step": 850
605
+ },
606
+ {
607
+ "epoch": 0.5502239283429302,
608
+ "grad_norm": NaN,
609
+ "learning_rate": 4.08296011942845e-05,
610
+ "loss": 0.0,
611
+ "step": 860
612
+ },
613
+ {
614
+ "epoch": 0.5566218809980806,
615
+ "grad_norm": NaN,
616
+ "learning_rate": 4.072296865003199e-05,
617
+ "loss": 0.0,
618
+ "step": 870
619
+ },
620
+ {
621
+ "epoch": 0.5630198336532309,
622
+ "grad_norm": NaN,
623
+ "learning_rate": 4.0616336105779485e-05,
624
+ "loss": 0.0,
625
+ "step": 880
626
+ },
627
+ {
628
+ "epoch": 0.5694177863083814,
629
+ "grad_norm": NaN,
630
+ "learning_rate": 4.050970356152698e-05,
631
+ "loss": 0.0,
632
+ "step": 890
633
+ },
634
+ {
635
+ "epoch": 0.5758157389635317,
636
+ "grad_norm": NaN,
637
+ "learning_rate": 4.040307101727447e-05,
638
+ "loss": 0.0,
639
+ "step": 900
640
+ },
641
+ {
642
+ "epoch": 0.582213691618682,
643
+ "grad_norm": NaN,
644
+ "learning_rate": 4.029643847302197e-05,
645
+ "loss": 0.0,
646
+ "step": 910
647
+ },
648
+ {
649
+ "epoch": 0.5886116442738324,
650
+ "grad_norm": NaN,
651
+ "learning_rate": 4.0189805928769465e-05,
652
+ "loss": 0.0,
653
+ "step": 920
654
+ },
655
+ {
656
+ "epoch": 0.5950095969289827,
657
+ "grad_norm": NaN,
658
+ "learning_rate": 4.008317338451696e-05,
659
+ "loss": 0.0,
660
+ "step": 930
661
+ },
662
+ {
663
+ "epoch": 0.6014075495841331,
664
+ "grad_norm": NaN,
665
+ "learning_rate": 3.997654084026445e-05,
666
+ "loss": 0.0,
667
+ "step": 940
668
+ },
669
+ {
670
+ "epoch": 0.6078055022392834,
671
+ "grad_norm": NaN,
672
+ "learning_rate": 3.9869908296011945e-05,
673
+ "loss": 0.0,
674
+ "step": 950
675
+ },
676
+ {
677
+ "epoch": 0.6142034548944337,
678
+ "grad_norm": NaN,
679
+ "learning_rate": 3.976327575175944e-05,
680
+ "loss": 0.0,
681
+ "step": 960
682
+ },
683
+ {
684
+ "epoch": 0.6206014075495841,
685
+ "grad_norm": NaN,
686
+ "learning_rate": 3.965664320750693e-05,
687
+ "loss": 0.0,
688
+ "step": 970
689
+ },
690
+ {
691
+ "epoch": 0.6269993602047345,
692
+ "grad_norm": NaN,
693
+ "learning_rate": 3.9550010663254424e-05,
694
+ "loss": 0.0,
695
+ "step": 980
696
+ },
697
+ {
698
+ "epoch": 0.6333973128598849,
699
+ "grad_norm": NaN,
700
+ "learning_rate": 3.9443378119001924e-05,
701
+ "loss": 0.0,
702
+ "step": 990
703
+ },
704
+ {
705
+ "epoch": 0.6397952655150352,
706
+ "grad_norm": NaN,
707
+ "learning_rate": 3.933674557474942e-05,
708
+ "loss": 0.0,
709
+ "step": 1000
710
+ },
711
+ {
712
+ "epoch": 0.6461932181701855,
713
+ "grad_norm": NaN,
714
+ "learning_rate": 3.923011303049691e-05,
715
+ "loss": 0.0,
716
+ "step": 1010
717
+ },
718
+ {
719
+ "epoch": 0.6525911708253359,
720
+ "grad_norm": NaN,
721
+ "learning_rate": 3.9123480486244404e-05,
722
+ "loss": 0.0,
723
+ "step": 1020
724
+ },
725
+ {
726
+ "epoch": 0.6589891234804862,
727
+ "grad_norm": NaN,
728
+ "learning_rate": 3.90168479419919e-05,
729
+ "loss": 0.0,
730
+ "step": 1030
731
+ },
732
+ {
733
+ "epoch": 0.6653870761356366,
734
+ "grad_norm": NaN,
735
+ "learning_rate": 3.891021539773939e-05,
736
+ "loss": 0.0,
737
+ "step": 1040
738
+ },
739
+ {
740
+ "epoch": 0.6717850287907869,
741
+ "grad_norm": NaN,
742
+ "learning_rate": 3.8803582853486884e-05,
743
+ "loss": 0.0,
744
+ "step": 1050
745
+ },
746
+ {
747
+ "epoch": 0.6781829814459372,
748
+ "grad_norm": NaN,
749
+ "learning_rate": 3.8696950309234384e-05,
750
+ "loss": 0.0,
751
+ "step": 1060
752
+ },
753
+ {
754
+ "epoch": 0.6845809341010877,
755
+ "grad_norm": NaN,
756
+ "learning_rate": 3.859031776498188e-05,
757
+ "loss": 0.0,
758
+ "step": 1070
759
+ },
760
+ {
761
+ "epoch": 0.690978886756238,
762
+ "grad_norm": NaN,
763
+ "learning_rate": 3.848368522072937e-05,
764
+ "loss": 0.0,
765
+ "step": 1080
766
+ },
767
+ {
768
+ "epoch": 0.6973768394113884,
769
+ "grad_norm": NaN,
770
+ "learning_rate": 3.837705267647686e-05,
771
+ "loss": 0.0,
772
+ "step": 1090
773
+ },
774
+ {
775
+ "epoch": 0.7037747920665387,
776
+ "grad_norm": NaN,
777
+ "learning_rate": 3.827042013222436e-05,
778
+ "loss": 0.0,
779
+ "step": 1100
780
+ },
781
+ {
782
+ "epoch": 0.710172744721689,
783
+ "grad_norm": NaN,
784
+ "learning_rate": 3.816378758797185e-05,
785
+ "loss": 0.0,
786
+ "step": 1110
787
+ },
788
+ {
789
+ "epoch": 0.7165706973768394,
790
+ "grad_norm": NaN,
791
+ "learning_rate": 3.8057155043719344e-05,
792
+ "loss": 0.0,
793
+ "step": 1120
794
+ },
795
+ {
796
+ "epoch": 0.7229686500319897,
797
+ "grad_norm": NaN,
798
+ "learning_rate": 3.795052249946684e-05,
799
+ "loss": 0.0,
800
+ "step": 1130
801
+ },
802
+ {
803
+ "epoch": 0.7293666026871402,
804
+ "grad_norm": NaN,
805
+ "learning_rate": 3.784388995521434e-05,
806
+ "loss": 0.0,
807
+ "step": 1140
808
+ },
809
+ {
810
+ "epoch": 0.7357645553422905,
811
+ "grad_norm": NaN,
812
+ "learning_rate": 3.773725741096183e-05,
813
+ "loss": 0.0,
814
+ "step": 1150
815
+ },
816
+ {
817
+ "epoch": 0.7421625079974408,
818
+ "grad_norm": NaN,
819
+ "learning_rate": 3.7630624866709324e-05,
820
+ "loss": 0.0,
821
+ "step": 1160
822
+ },
823
+ {
824
+ "epoch": 0.7485604606525912,
825
+ "grad_norm": NaN,
826
+ "learning_rate": 3.752399232245682e-05,
827
+ "loss": 0.0,
828
+ "step": 1170
829
+ },
830
+ {
831
+ "epoch": 0.7549584133077415,
832
+ "grad_norm": NaN,
833
+ "learning_rate": 3.741735977820431e-05,
834
+ "loss": 0.0,
835
+ "step": 1180
836
+ },
837
+ {
838
+ "epoch": 0.7613563659628919,
839
+ "grad_norm": NaN,
840
+ "learning_rate": 3.73107272339518e-05,
841
+ "loss": 0.0,
842
+ "step": 1190
843
+ },
844
+ {
845
+ "epoch": 0.7677543186180422,
846
+ "grad_norm": NaN,
847
+ "learning_rate": 3.72040946896993e-05,
848
+ "loss": 0.0,
849
+ "step": 1200
850
+ },
851
+ {
852
+ "epoch": 0.7741522712731925,
853
+ "grad_norm": NaN,
854
+ "learning_rate": 3.709746214544679e-05,
855
+ "loss": 0.0,
856
+ "step": 1210
857
+ },
858
+ {
859
+ "epoch": 0.780550223928343,
860
+ "grad_norm": NaN,
861
+ "learning_rate": 3.699082960119429e-05,
862
+ "loss": 0.0,
863
+ "step": 1220
864
+ },
865
+ {
866
+ "epoch": 0.7869481765834933,
867
+ "grad_norm": NaN,
868
+ "learning_rate": 3.688419705694178e-05,
869
+ "loss": 0.0,
870
+ "step": 1230
871
+ },
872
+ {
873
+ "epoch": 0.7933461292386437,
874
+ "grad_norm": NaN,
875
+ "learning_rate": 3.677756451268927e-05,
876
+ "loss": 0.0,
877
+ "step": 1240
878
+ },
879
+ {
880
+ "epoch": 0.799744081893794,
881
+ "grad_norm": NaN,
882
+ "learning_rate": 3.667093196843677e-05,
883
+ "loss": 0.0,
884
+ "step": 1250
885
+ },
886
+ {
887
+ "epoch": 0.8061420345489443,
888
+ "grad_norm": NaN,
889
+ "learning_rate": 3.656429942418426e-05,
890
+ "loss": 0.0,
891
+ "step": 1260
892
+ },
893
+ {
894
+ "epoch": 0.8125399872040947,
895
+ "grad_norm": NaN,
896
+ "learning_rate": 3.6457666879931756e-05,
897
+ "loss": 0.0,
898
+ "step": 1270
899
+ },
900
+ {
901
+ "epoch": 0.818937939859245,
902
+ "grad_norm": NaN,
903
+ "learning_rate": 3.635103433567925e-05,
904
+ "loss": 0.0,
905
+ "step": 1280
906
+ },
907
+ {
908
+ "epoch": 0.8253358925143954,
909
+ "grad_norm": NaN,
910
+ "learning_rate": 3.624440179142675e-05,
911
+ "loss": 0.0,
912
+ "step": 1290
913
+ },
914
+ {
915
+ "epoch": 0.8317338451695457,
916
+ "grad_norm": NaN,
917
+ "learning_rate": 3.613776924717424e-05,
918
+ "loss": 0.0,
919
+ "step": 1300
920
+ },
921
+ {
922
+ "epoch": 0.838131797824696,
923
+ "grad_norm": NaN,
924
+ "learning_rate": 3.6031136702921736e-05,
925
+ "loss": 0.0,
926
+ "step": 1310
927
+ },
928
+ {
929
+ "epoch": 0.8445297504798465,
930
+ "grad_norm": NaN,
931
+ "learning_rate": 3.592450415866922e-05,
932
+ "loss": 0.0,
933
+ "step": 1320
934
+ },
935
+ {
936
+ "epoch": 0.8509277031349968,
937
+ "grad_norm": NaN,
938
+ "learning_rate": 3.581787161441672e-05,
939
+ "loss": 0.0,
940
+ "step": 1330
941
+ },
942
+ {
943
+ "epoch": 0.8573256557901472,
944
+ "grad_norm": NaN,
945
+ "learning_rate": 3.5711239070164216e-05,
946
+ "loss": 0.0,
947
+ "step": 1340
948
+ },
949
+ {
950
+ "epoch": 0.8637236084452975,
951
+ "grad_norm": NaN,
952
+ "learning_rate": 3.560460652591171e-05,
953
+ "loss": 0.0,
954
+ "step": 1350
955
+ },
956
+ {
957
+ "epoch": 0.8701215611004478,
958
+ "grad_norm": NaN,
959
+ "learning_rate": 3.54979739816592e-05,
960
+ "loss": 0.0,
961
+ "step": 1360
962
+ },
963
+ {
964
+ "epoch": 0.8765195137555982,
965
+ "grad_norm": NaN,
966
+ "learning_rate": 3.53913414374067e-05,
967
+ "loss": 0.0,
968
+ "step": 1370
969
+ },
970
+ {
971
+ "epoch": 0.8829174664107485,
972
+ "grad_norm": NaN,
973
+ "learning_rate": 3.5284708893154196e-05,
974
+ "loss": 0.0,
975
+ "step": 1380
976
+ },
977
+ {
978
+ "epoch": 0.889315419065899,
979
+ "grad_norm": NaN,
980
+ "learning_rate": 3.517807634890168e-05,
981
+ "loss": 0.0,
982
+ "step": 1390
983
+ },
984
+ {
985
+ "epoch": 0.8957133717210493,
986
+ "grad_norm": NaN,
987
+ "learning_rate": 3.5071443804649176e-05,
988
+ "loss": 0.0,
989
+ "step": 1400
990
+ },
991
+ {
992
+ "epoch": 0.9021113243761996,
993
+ "grad_norm": NaN,
994
+ "learning_rate": 3.4964811260396676e-05,
995
+ "loss": 0.0,
996
+ "step": 1410
997
+ },
998
+ {
999
+ "epoch": 0.90850927703135,
1000
+ "grad_norm": NaN,
1001
+ "learning_rate": 3.485817871614417e-05,
1002
+ "loss": 0.0,
1003
+ "step": 1420
1004
+ },
1005
+ {
1006
+ "epoch": 0.9149072296865003,
1007
+ "grad_norm": NaN,
1008
+ "learning_rate": 3.475154617189166e-05,
1009
+ "loss": 0.0,
1010
+ "step": 1430
1011
+ },
1012
+ {
1013
+ "epoch": 0.9213051823416507,
1014
+ "grad_norm": NaN,
1015
+ "learning_rate": 3.4644913627639155e-05,
1016
+ "loss": 0.0,
1017
+ "step": 1440
1018
+ },
1019
+ {
1020
+ "epoch": 0.927703134996801,
1021
+ "grad_norm": NaN,
1022
+ "learning_rate": 3.4538281083386656e-05,
1023
+ "loss": 0.0,
1024
+ "step": 1450
1025
+ },
1026
+ {
1027
+ "epoch": 0.9341010876519513,
1028
+ "grad_norm": NaN,
1029
+ "learning_rate": 3.443164853913415e-05,
1030
+ "loss": 0.0,
1031
+ "step": 1460
1032
+ },
1033
+ {
1034
+ "epoch": 0.9404990403071017,
1035
+ "grad_norm": NaN,
1036
+ "learning_rate": 3.4325015994881635e-05,
1037
+ "loss": 0.0,
1038
+ "step": 1470
1039
+ },
1040
+ {
1041
+ "epoch": 0.946896992962252,
1042
+ "grad_norm": NaN,
1043
+ "learning_rate": 3.4218383450629135e-05,
1044
+ "loss": 0.0,
1045
+ "step": 1480
1046
+ },
1047
+ {
1048
+ "epoch": 0.9532949456174025,
1049
+ "grad_norm": NaN,
1050
+ "learning_rate": 3.411175090637663e-05,
1051
+ "loss": 0.0,
1052
+ "step": 1490
1053
+ },
1054
+ {
1055
+ "epoch": 0.9596928982725528,
1056
+ "grad_norm": NaN,
1057
+ "learning_rate": 3.400511836212412e-05,
1058
+ "loss": 0.0,
1059
+ "step": 1500
1060
+ },
1061
+ {
1062
+ "epoch": 0.9660908509277031,
1063
+ "grad_norm": NaN,
1064
+ "learning_rate": 3.3898485817871615e-05,
1065
+ "loss": 0.0,
1066
+ "step": 1510
1067
+ },
1068
+ {
1069
+ "epoch": 0.9724888035828535,
1070
+ "grad_norm": NaN,
1071
+ "learning_rate": 3.379185327361911e-05,
1072
+ "loss": 0.0,
1073
+ "step": 1520
1074
+ },
1075
+ {
1076
+ "epoch": 0.9788867562380038,
1077
+ "grad_norm": NaN,
1078
+ "learning_rate": 3.368522072936661e-05,
1079
+ "loss": 0.0,
1080
+ "step": 1530
1081
+ },
1082
+ {
1083
+ "epoch": 0.9852847088931542,
1084
+ "grad_norm": NaN,
1085
+ "learning_rate": 3.3578588185114095e-05,
1086
+ "loss": 0.0,
1087
+ "step": 1540
1088
+ },
1089
+ {
1090
+ "epoch": 0.9916826615483045,
1091
+ "grad_norm": NaN,
1092
+ "learning_rate": 3.347195564086159e-05,
1093
+ "loss": 0.0,
1094
+ "step": 1550
1095
+ },
1096
+ {
1097
+ "epoch": 0.9980806142034548,
1098
+ "grad_norm": NaN,
1099
+ "learning_rate": 3.336532309660909e-05,
1100
+ "loss": 0.0,
1101
+ "step": 1560
1102
+ },
1103
+ {
1104
+ "epoch": 1.0044785668586051,
1105
+ "grad_norm": NaN,
1106
+ "learning_rate": 3.325869055235658e-05,
1107
+ "loss": 0.0,
1108
+ "step": 1570
1109
+ },
1110
+ {
1111
+ "epoch": 1.0108765195137557,
1112
+ "grad_norm": NaN,
1113
+ "learning_rate": 3.3152058008104075e-05,
1114
+ "loss": 0.0,
1115
+ "step": 1580
1116
+ },
1117
+ {
1118
+ "epoch": 1.017274472168906,
1119
+ "grad_norm": NaN,
1120
+ "learning_rate": 3.304542546385157e-05,
1121
+ "loss": 0.0,
1122
+ "step": 1590
1123
+ },
1124
+ {
1125
+ "epoch": 1.0236724248240563,
1126
+ "grad_norm": NaN,
1127
+ "learning_rate": 3.293879291959907e-05,
1128
+ "loss": 0.0,
1129
+ "step": 1600
1130
+ },
1131
+ {
1132
+ "epoch": 1.0300703774792066,
1133
+ "grad_norm": NaN,
1134
+ "learning_rate": 3.283216037534656e-05,
1135
+ "loss": 0.0,
1136
+ "step": 1610
1137
+ },
1138
+ {
1139
+ "epoch": 1.036468330134357,
1140
+ "grad_norm": NaN,
1141
+ "learning_rate": 3.272552783109405e-05,
1142
+ "loss": 0.0,
1143
+ "step": 1620
1144
+ },
1145
+ {
1146
+ "epoch": 1.0428662827895074,
1147
+ "grad_norm": NaN,
1148
+ "learning_rate": 3.261889528684154e-05,
1149
+ "loss": 0.0,
1150
+ "step": 1630
1151
+ },
1152
+ {
1153
+ "epoch": 1.0492642354446577,
1154
+ "grad_norm": NaN,
1155
+ "learning_rate": 3.251226274258904e-05,
1156
+ "loss": 0.0,
1157
+ "step": 1640
1158
+ },
1159
+ {
1160
+ "epoch": 1.055662188099808,
1161
+ "grad_norm": NaN,
1162
+ "learning_rate": 3.2405630198336535e-05,
1163
+ "loss": 0.0,
1164
+ "step": 1650
1165
+ },
1166
+ {
1167
+ "epoch": 1.0620601407549584,
1168
+ "grad_norm": NaN,
1169
+ "learning_rate": 3.229899765408403e-05,
1170
+ "loss": 0.0,
1171
+ "step": 1660
1172
+ },
1173
+ {
1174
+ "epoch": 1.0684580934101087,
1175
+ "grad_norm": NaN,
1176
+ "learning_rate": 3.219236510983152e-05,
1177
+ "loss": 0.0,
1178
+ "step": 1670
1179
+ },
1180
+ {
1181
+ "epoch": 1.0748560460652592,
1182
+ "grad_norm": NaN,
1183
+ "learning_rate": 3.208573256557902e-05,
1184
+ "loss": 0.0,
1185
+ "step": 1680
1186
+ },
1187
+ {
1188
+ "epoch": 1.0812539987204095,
1189
+ "grad_norm": NaN,
1190
+ "learning_rate": 3.197910002132651e-05,
1191
+ "loss": 0.0,
1192
+ "step": 1690
1193
+ },
1194
+ {
1195
+ "epoch": 1.0876519513755598,
1196
+ "grad_norm": NaN,
1197
+ "learning_rate": 3.1872467477074e-05,
1198
+ "loss": 0.0,
1199
+ "step": 1700
1200
+ },
1201
+ {
1202
+ "epoch": 1.0940499040307101,
1203
+ "grad_norm": NaN,
1204
+ "learning_rate": 3.17658349328215e-05,
1205
+ "loss": 0.0,
1206
+ "step": 1710
1207
+ },
1208
+ {
1209
+ "epoch": 1.1004478566858604,
1210
+ "grad_norm": NaN,
1211
+ "learning_rate": 3.1659202388568994e-05,
1212
+ "loss": 0.0,
1213
+ "step": 1720
1214
+ },
1215
+ {
1216
+ "epoch": 1.106845809341011,
1217
+ "grad_norm": NaN,
1218
+ "learning_rate": 3.155256984431649e-05,
1219
+ "loss": 0.0,
1220
+ "step": 1730
1221
+ },
1222
+ {
1223
+ "epoch": 1.1132437619961613,
1224
+ "grad_norm": NaN,
1225
+ "learning_rate": 3.144593730006398e-05,
1226
+ "loss": 0.0,
1227
+ "step": 1740
1228
+ },
1229
+ {
1230
+ "epoch": 1.1196417146513116,
1231
+ "grad_norm": NaN,
1232
+ "learning_rate": 3.1339304755811474e-05,
1233
+ "loss": 0.0,
1234
+ "step": 1750
1235
+ },
1236
+ {
1237
+ "epoch": 1.1260396673064619,
1238
+ "grad_norm": NaN,
1239
+ "learning_rate": 3.1232672211558974e-05,
1240
+ "loss": 0.0,
1241
+ "step": 1760
1242
+ },
1243
+ {
1244
+ "epoch": 1.1324376199616122,
1245
+ "grad_norm": NaN,
1246
+ "learning_rate": 3.112603966730646e-05,
1247
+ "loss": 0.0,
1248
+ "step": 1770
1249
+ },
1250
+ {
1251
+ "epoch": 1.1388355726167627,
1252
+ "grad_norm": NaN,
1253
+ "learning_rate": 3.1019407123053954e-05,
1254
+ "loss": 0.0,
1255
+ "step": 1780
1256
+ },
1257
+ {
1258
+ "epoch": 1.145233525271913,
1259
+ "grad_norm": NaN,
1260
+ "learning_rate": 3.0912774578801454e-05,
1261
+ "loss": 0.0,
1262
+ "step": 1790
1263
+ },
1264
+ {
1265
+ "epoch": 1.1516314779270633,
1266
+ "grad_norm": NaN,
1267
+ "learning_rate": 3.080614203454895e-05,
1268
+ "loss": 0.0,
1269
+ "step": 1800
1270
+ },
1271
+ {
1272
+ "epoch": 1.1580294305822136,
1273
+ "grad_norm": NaN,
1274
+ "learning_rate": 3.069950949029644e-05,
1275
+ "loss": 0.0,
1276
+ "step": 1810
1277
+ },
1278
+ {
1279
+ "epoch": 1.164427383237364,
1280
+ "grad_norm": NaN,
1281
+ "learning_rate": 3.0592876946043934e-05,
1282
+ "loss": 0.0,
1283
+ "step": 1820
1284
+ },
1285
+ {
1286
+ "epoch": 1.1708253358925145,
1287
+ "grad_norm": NaN,
1288
+ "learning_rate": 3.048624440179143e-05,
1289
+ "loss": 0.0,
1290
+ "step": 1830
1291
+ },
1292
+ {
1293
+ "epoch": 1.1772232885476648,
1294
+ "grad_norm": NaN,
1295
+ "learning_rate": 3.037961185753892e-05,
1296
+ "loss": 0.0,
1297
+ "step": 1840
1298
+ },
1299
+ {
1300
+ "epoch": 1.183621241202815,
1301
+ "grad_norm": NaN,
1302
+ "learning_rate": 3.0272979313286414e-05,
1303
+ "loss": 0.0,
1304
+ "step": 1850
1305
+ },
1306
+ {
1307
+ "epoch": 1.1900191938579654,
1308
+ "grad_norm": NaN,
1309
+ "learning_rate": 3.016634676903391e-05,
1310
+ "loss": 0.0,
1311
+ "step": 1860
1312
+ },
1313
+ {
1314
+ "epoch": 1.1964171465131157,
1315
+ "grad_norm": NaN,
1316
+ "learning_rate": 3.0059714224781403e-05,
1317
+ "loss": 0.0,
1318
+ "step": 1870
1319
+ },
1320
+ {
1321
+ "epoch": 1.2028150991682662,
1322
+ "grad_norm": NaN,
1323
+ "learning_rate": 2.99530816805289e-05,
1324
+ "loss": 0.0,
1325
+ "step": 1880
1326
+ },
1327
+ {
1328
+ "epoch": 1.2092130518234165,
1329
+ "grad_norm": NaN,
1330
+ "learning_rate": 2.9846449136276393e-05,
1331
+ "loss": 0.0,
1332
+ "step": 1890
1333
+ },
1334
+ {
1335
+ "epoch": 1.2156110044785668,
1336
+ "grad_norm": NaN,
1337
+ "learning_rate": 2.973981659202389e-05,
1338
+ "loss": 0.0,
1339
+ "step": 1900
1340
+ },
1341
+ {
1342
+ "epoch": 1.2220089571337172,
1343
+ "grad_norm": NaN,
1344
+ "learning_rate": 2.9633184047771383e-05,
1345
+ "loss": 0.0,
1346
+ "step": 1910
1347
+ },
1348
+ {
1349
+ "epoch": 1.2284069097888675,
1350
+ "grad_norm": NaN,
1351
+ "learning_rate": 2.9526551503518873e-05,
1352
+ "loss": 0.0,
1353
+ "step": 1920
1354
+ },
1355
+ {
1356
+ "epoch": 1.234804862444018,
1357
+ "grad_norm": NaN,
1358
+ "learning_rate": 2.941991895926637e-05,
1359
+ "loss": 0.0,
1360
+ "step": 1930
1361
+ },
1362
+ {
1363
+ "epoch": 1.2412028150991683,
1364
+ "grad_norm": NaN,
1365
+ "learning_rate": 2.9313286415013863e-05,
1366
+ "loss": 0.0,
1367
+ "step": 1940
1368
+ },
1369
+ {
1370
+ "epoch": 1.2476007677543186,
1371
+ "grad_norm": NaN,
1372
+ "learning_rate": 2.9206653870761356e-05,
1373
+ "loss": 0.0,
1374
+ "step": 1950
1375
+ },
1376
+ {
1377
+ "epoch": 1.253998720409469,
1378
+ "grad_norm": NaN,
1379
+ "learning_rate": 2.9100021326508853e-05,
1380
+ "loss": 0.0,
1381
+ "step": 1960
1382
+ },
1383
+ {
1384
+ "epoch": 1.2603966730646192,
1385
+ "grad_norm": NaN,
1386
+ "learning_rate": 2.8993388782256346e-05,
1387
+ "loss": 0.0,
1388
+ "step": 1970
1389
+ },
1390
+ {
1391
+ "epoch": 1.2667946257197698,
1392
+ "grad_norm": NaN,
1393
+ "learning_rate": 2.8886756238003843e-05,
1394
+ "loss": 0.0,
1395
+ "step": 1980
1396
+ },
1397
+ {
1398
+ "epoch": 1.27319257837492,
1399
+ "grad_norm": NaN,
1400
+ "learning_rate": 2.8780123693751333e-05,
1401
+ "loss": 0.0,
1402
+ "step": 1990
1403
+ },
1404
+ {
1405
+ "epoch": 1.2795905310300704,
1406
+ "grad_norm": NaN,
1407
+ "learning_rate": 2.8673491149498826e-05,
1408
+ "loss": 0.0,
1409
+ "step": 2000
1410
+ }
1411
+ ],
1412
+ "logging_steps": 10,
1413
+ "max_steps": 4689,
1414
+ "num_input_tokens_seen": 0,
1415
+ "num_train_epochs": 3,
1416
+ "save_steps": 500,
1417
+ "stateful_callbacks": {
1418
+ "TrainerControl": {
1419
+ "args": {
1420
+ "should_epoch_stop": false,
1421
+ "should_evaluate": false,
1422
+ "should_log": false,
1423
+ "should_save": true,
1424
+ "should_training_stop": false
1425
+ },
1426
+ "attributes": {}
1427
+ }
1428
+ },
1429
+ "total_flos": 0.0,
1430
+ "train_batch_size": 16,
1431
+ "trial_name": null,
1432
+ "trial_params": null
1433
+ }
checkpoint-2000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ceebcdd4aa5d5e688dc20d333b0c1fc1d54fc83ca77f4049b33fecb99c128b6
3
+ size 5112
config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"return_dict": true, "output_hidden_states": false, "output_attentions": false, "torchscript": false, "torch_dtype": null, "use_bfloat16": false, "tf_legacy_loss": false, "pruned_heads": {}, "tie_word_embeddings": true, "chunk_size_feed_forward": 0, "is_encoder_decoder": false, "is_decoder": false, "cross_attention_hidden_size": null, "add_cross_attention": false, "tie_encoder_decoder": false, "max_length": 20, "min_length": 0, "do_sample": false, "early_stopping": false, "num_beams": 1, "num_beam_groups": 1, "diversity_penalty": 0.0, "temperature": 1.0, "top_k": 50, "top_p": 1.0, "typical_p": 1.0, "repetition_penalty": 1.0, "length_penalty": 1.0, "no_repeat_ngram_size": 0, "encoder_no_repeat_ngram_size": 0, "bad_words_ids": null, "num_return_sequences": 1, "output_scores": false, "return_dict_in_generate": false, "forced_bos_token_id": null, "forced_eos_token_id": null, "remove_invalid_values": false, "exponential_decay_length_penalty": null, "suppress_tokens": null, "begin_suppress_tokens": null, "architectures": null, "finetuning_task": null, "id2label": {"0": "LABEL_0", "1": "LABEL_1"}, "label2id": {"LABEL_0": 0, "LABEL_1": 1}, "tokenizer_class": null, "prefix": null, "bos_token_id": null, "pad_token_id": null, "eos_token_id": null, "sep_token_id": null, "decoder_start_token_id": null, "task_specific_params": null, "problem_type": null, "_name_or_path": "", "transformers_version": "4.43.3", "vocab_size": 50257, "n_layer": 12, "n_head": 12, "n_embd": 768, "output_dir": "./results", "block_size": 512, "hidden_size": 768, "dropout": 0.1, "bias": true, "ff_expansion_factor": 4, "label_smoothing": 0.1, "model_type": "judge-gpt"}
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ad7ff53135b3a3cb270495374fc00d8d4748b20ab6b9710bbbcfbdb7fdc77cd
3
+ size 623813362
runs/Aug01_01-37-45_DadsLapDancer/events.out.tfevents.1722494525.DadsLapDancer.6780.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2d12c1c4718203d4ae6470378d8195c9eff820b5473a76f120e9a7e956cb654
3
+ size 4150
runs/Aug01_01-47-58_DadsLapDancer/events.out.tfevents.1722495144.DadsLapDancer.13068.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a13b363d44287339160dfa2387259a8b3801194664543ea11501e364475960fe
3
+ size 4150
runs/Aug01_02-38-07_DadsLapDancer/events.out.tfevents.1722498158.DadsLapDancer.12600.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0147250e5fda8b28fe795f3cf9d6fbf91edf77fdd4fa7e66f818ed1495e8fb63
3
+ size 4150
runs/Aug01_02-51-57_DadsLapDancer/events.out.tfevents.1722499044.DadsLapDancer.16012.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd0bb26f560052a0e0e18cf31f52a543b8c1d2523d708354b8dc63b8008cddf7
3
+ size 4150
runs/Aug01_03-05-35_DadsLapDancer/events.out.tfevents.1722499871.DadsLapDancer.16688.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e53711cba9ec31d5a08e0b85e9f2d0c318291979a1a76eca392d761fcc7c3ac1
3
+ size 4150
runs/Aug01_03-25-55_DadsLapDancer/events.out.tfevents.1722501091.DadsLapDancer.19340.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db147844dc5b1576e5963f39d570402359c6d7121c26a7c52841ee9ca02e42e3
3
+ size 4150
runs/Aug01_03-44-45_DadsLapDancer/events.out.tfevents.1722502221.DadsLapDancer.6876.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2e7435219b858d8c5319f62300a38c7532875650fb8964e57f580ac6b1f138c
3
+ size 4150
runs/Aug01_14-17-09_DadsLapDancer/events.out.tfevents.1722540771.DadsLapDancer.8172.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db704a7ec1990c540a8ad58bea3b5b17e3303f79f5605e464d9d4bf654e94cd7
3
+ size 4150
runs/Aug01_14-43-24_DadsLapDancer/events.out.tfevents.1722541884.DadsLapDancer.8268.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd6f1bf820e1072d6b85e6df376351288fe9fe2f9d059c66f748e47aa8372b33
3
+ size 4150
runs/Aug01_16-39-05_DadsLapDancer/events.out.tfevents.1722548362.DadsLapDancer.5608.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d96e84a5441274be598841b411a3083827301b599bd8542e49ba221ae59cfda3
3
+ size 4150
runs/Aug01_16-45-01_DadsLapDancer/events.out.tfevents.1722548708.DadsLapDancer.6604.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a79913760583ac052b2d4688abdae090f86124119cda3742c509e465a17293a
3
+ size 4150
runs/Aug01_16-46-45_DadsLapDancer/events.out.tfevents.1722548812.DadsLapDancer.14620.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca6d512d9182b30b7f84f7bbd1f4c3f126fd2ba42d679c1d6b54b67366f6c790
3
+ size 4150
runs/Aug01_16-47-33_DadsLapDancer/events.out.tfevents.1722548871.DadsLapDancer.7024.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c7fd50f1369b44f2c5d72949977627d5af940182e4dfa27ce73761713eb997a
3
+ size 4150
runs/Aug01_16-49-14_DadsLapDancer/events.out.tfevents.1722548960.DadsLapDancer.5688.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30be1e415443b35f43ee116f252cc1deaeae3243191a779852780d564a58176e
3
+ size 4150
runs/Aug01_18-06-27_DadsLapDancer/events.out.tfevents.1722554013.DadsLapDancer.18004.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:98066535b36ec56a8c5c169934987fe842eb65b7c79e9c7e9b5856836be8fe84
3
+ size 4150
runs/Aug01_18-39-26_DadsLapDancer/events.out.tfevents.1722555571.DadsLapDancer.2512.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41ea1928f846efda5df50953c54225db5de91f62c932f5de88659f4e6474eda0
3
+ size 4150
runs/Aug01_19-17-43_DadsLapDancer/events.out.tfevents.1722557869.DadsLapDancer.15380.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a8b5c7e2b6dd385836634c0fa7a9986bdcd8a019c7a5dbb78aad2c086a112d5
3
+ size 4150
runs/Aug01_20-48-10_DadsLapDancer/events.out.tfevents.1722563301.DadsLapDancer.2980.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0c6c36db22de9ffe36f6d5d4f58cdd5d2eeb1578838d851edf8fe32f58b6b824
3
+ size 4150
runs/Aug01_21-17-07_DadsLapDancer/events.out.tfevents.1722565044.DadsLapDancer.7652.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6038dfe10b88275c2d99d8ae2be6da247bdfb6a6243c675541409bbfa69d67f3
3
+ size 4150
runs/Aug01_21-19-14_DadsLapDancer/events.out.tfevents.1722565160.DadsLapDancer.8464.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:90bf80a52e7496cadac66284a5917c70897dd4740b2f3111d119a3793d5ec7d7
3
+ size 4150
runs/Aug01_21-42-11_DadsLapDancer/events.out.tfevents.1722566547.DadsLapDancer.1532.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2977f82710ea871247d3ce7bbae2f1dab6e3919fca8ee27528c5a87314869bc
3
+ size 4150
runs/Aug01_23-04-07_DadsLapDancer/events.out.tfevents.1722571454.DadsLapDancer.5628.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c81bfd4b067488531a4ff0911778f08d8c0d475d861454a883e162e1ae602d12
3
+ size 4150
runs/Aug01_23-10-54_DadsLapDancer/events.out.tfevents.1722571873.DadsLapDancer.16684.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c512e417c8065f27e4c17a76e8bf88c7c19f36fc42ccbfd2775b77d5600364fe
3
+ size 4150
runs/Aug01_23-13-15_DadsLapDancer/events.out.tfevents.1722572005.DadsLapDancer.7708.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e82727e4f9f06c2670675b6a6be4ec5b99d0caa5b6c2cffd258445bfd4d89f52
3
+ size 4150
runs/Aug01_23-20-59_DadsLapDancer/events.out.tfevents.1722572477.DadsLapDancer.2392.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ebfd5987dd7ab13009cb3d77427beeb2ed769912ae7dd08ef7e4f81cf09719e
3
+ size 4150
runs/Aug01_23-28-28_DadsLapDancer/events.out.tfevents.1722572926.DadsLapDancer.6044.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c05dcb47b30226957fc12e0f8bffc177b69803b0ad2c4584cf773d8a811f8f86
3
+ size 4150
runs/Aug01_23-48-43_DadsLapDancer/events.out.tfevents.1722574142.DadsLapDancer.3404.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f8c71c022ec7c613b0fc311bd3634bf99a93e2b9301ba428a59725c0d876f80
3
+ size 4150
runs/Aug02_00-50-38_DadsLapDancer/events.out.tfevents.1722577849.DadsLapDancer.17716.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f1962ef4b222cc535829bbd9220a67f368613d546ffcd9ebb72259bc71a0d8b
3
+ size 4150
runs/Aug02_00-53-03_DadsLapDancer/events.out.tfevents.1722578002.DadsLapDancer.13092.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6e797a93e9865b09f91e93c8e5e9ea34b5d07419865e848a65863c7fe374d34
3
+ size 4150