tejaskamtam commited on
Commit
97121e7
1 Parent(s): bd2c029

End of training

Browse files
Files changed (5) hide show
  1. README.md +16 -4
  2. all_results.json +15 -0
  3. eval_results.json +10 -0
  4. train_results.json +8 -0
  5. trainer_state.json +645 -0
README.md CHANGED
@@ -3,11 +3,23 @@ license: apache-2.0
3
  base_model: google/electra-base-generator
4
  tags:
5
  - generated_from_trainer
 
 
6
  metrics:
7
  - accuracy
8
  model-index:
9
  - name: electra-base-finetuned-xe_ey_fae
10
- results: []
 
 
 
 
 
 
 
 
 
 
11
  ---
12
 
13
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -15,10 +27,10 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # electra-base-finetuned-xe_ey_fae
17
 
18
- This model is a fine-tuned version of [google/electra-base-generator](https://huggingface.co/google/electra-base-generator) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
- - Loss: 1.7175
21
- - Accuracy: 0.6678
22
 
23
  ## Model description
24
 
 
3
  base_model: google/electra-base-generator
4
  tags:
5
  - generated_from_trainer
6
+ datasets:
7
+ - datasets/all_binary_and_xe_ey_fae_counterfactual
8
  metrics:
9
  - accuracy
10
  model-index:
11
  - name: electra-base-finetuned-xe_ey_fae
12
+ results:
13
+ - task:
14
+ name: Masked Language Modeling
15
+ type: fill-mask
16
+ dataset:
17
+ name: datasets/all_binary_and_xe_ey_fae_counterfactual
18
+ type: datasets/all_binary_and_xe_ey_fae_counterfactual
19
+ metrics:
20
+ - name: Accuracy
21
+ type: accuracy
22
+ value: 0.667333329363415
23
  ---
24
 
25
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
27
 
28
  # electra-base-finetuned-xe_ey_fae
29
 
30
+ This model is a fine-tuned version of [google/electra-base-generator](https://huggingface.co/google/electra-base-generator) on the datasets/all_binary_and_xe_ey_fae_counterfactual dataset.
31
  It achieves the following results on the evaluation set:
32
+ - Loss: 1.7211
33
+ - Accuracy: 0.6673
34
 
35
  ## Model description
36
 
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.64,
3
+ "eval_accuracy": 0.667333329363415,
4
+ "eval_loss": 1.7211331129074097,
5
+ "eval_runtime": 36.3378,
6
+ "eval_samples": 15525,
7
+ "eval_samples_per_second": 427.241,
8
+ "eval_steps_per_second": 53.415,
9
+ "perplexity": 5.59085995202853,
10
+ "train_loss": 1.9300706578696647,
11
+ "train_runtime": 3479.9405,
12
+ "train_samples": 124124,
13
+ "train_samples_per_second": 107.005,
14
+ "train_steps_per_second": 6.688
15
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.64,
3
+ "eval_accuracy": 0.667333329363415,
4
+ "eval_loss": 1.7211331129074097,
5
+ "eval_runtime": 36.3378,
6
+ "eval_samples": 15525,
7
+ "eval_samples_per_second": 427.241,
8
+ "eval_steps_per_second": 53.415,
9
+ "perplexity": 5.59085995202853
10
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.64,
3
+ "train_loss": 1.9300706578696647,
4
+ "train_runtime": 3479.9405,
5
+ "train_samples": 124124,
6
+ "train_samples_per_second": 107.005,
7
+ "train_steps_per_second": 6.688
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,645 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.7163910865783691,
3
+ "best_model_checkpoint": "finetuning/output/electra-base-finetuned_xe_ey_fae/checkpoint-19000",
4
+ "epoch": 2.642433616911575,
5
+ "eval_steps": 500,
6
+ "global_step": 20500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.06,
13
+ "learning_rate": 9.786027326630576e-06,
14
+ "loss": 2.5359,
15
+ "step": 500
16
+ },
17
+ {
18
+ "epoch": 0.06,
19
+ "eval_accuracy": 0.6227738650589344,
20
+ "eval_loss": 2.0696377754211426,
21
+ "eval_runtime": 35.9348,
22
+ "eval_samples_per_second": 432.033,
23
+ "eval_steps_per_second": 54.015,
24
+ "step": 500
25
+ },
26
+ {
27
+ "epoch": 0.13,
28
+ "learning_rate": 9.571195325255651e-06,
29
+ "loss": 2.1807,
30
+ "step": 1000
31
+ },
32
+ {
33
+ "epoch": 0.13,
34
+ "eval_accuracy": 0.6352025430222344,
35
+ "eval_loss": 1.9677125215530396,
36
+ "eval_runtime": 36.0518,
37
+ "eval_samples_per_second": 430.631,
38
+ "eval_steps_per_second": 53.839,
39
+ "step": 1000
40
+ },
41
+ {
42
+ "epoch": 0.19,
43
+ "learning_rate": 9.356363323880726e-06,
44
+ "loss": 2.1028,
45
+ "step": 1500
46
+ },
47
+ {
48
+ "epoch": 0.19,
49
+ "eval_accuracy": 0.641511887420089,
50
+ "eval_loss": 1.9191973209381104,
51
+ "eval_runtime": 36.3057,
52
+ "eval_samples_per_second": 427.619,
53
+ "eval_steps_per_second": 53.463,
54
+ "step": 1500
55
+ },
56
+ {
57
+ "epoch": 0.26,
58
+ "learning_rate": 9.141531322505801e-06,
59
+ "loss": 2.0658,
60
+ "step": 2000
61
+ },
62
+ {
63
+ "epoch": 0.26,
64
+ "eval_accuracy": 0.6450855805600152,
65
+ "eval_loss": 1.892332673072815,
66
+ "eval_runtime": 36.0414,
67
+ "eval_samples_per_second": 430.754,
68
+ "eval_steps_per_second": 53.855,
69
+ "step": 2000
70
+ },
71
+ {
72
+ "epoch": 0.32,
73
+ "learning_rate": 8.926699321130876e-06,
74
+ "loss": 2.0426,
75
+ "step": 2500
76
+ },
77
+ {
78
+ "epoch": 0.32,
79
+ "eval_accuracy": 0.6478244526689617,
80
+ "eval_loss": 1.8699322938919067,
81
+ "eval_runtime": 36.0808,
82
+ "eval_samples_per_second": 430.284,
83
+ "eval_steps_per_second": 53.796,
84
+ "step": 2500
85
+ },
86
+ {
87
+ "epoch": 0.39,
88
+ "learning_rate": 8.71186731975595e-06,
89
+ "loss": 2.0133,
90
+ "step": 3000
91
+ },
92
+ {
93
+ "epoch": 0.39,
94
+ "eval_accuracy": 0.6489956025492812,
95
+ "eval_loss": 1.8580025434494019,
96
+ "eval_runtime": 36.4103,
97
+ "eval_samples_per_second": 426.391,
98
+ "eval_steps_per_second": 53.309,
99
+ "step": 3000
100
+ },
101
+ {
102
+ "epoch": 0.45,
103
+ "learning_rate": 8.497464982383777e-06,
104
+ "loss": 1.9978,
105
+ "step": 3500
106
+ },
107
+ {
108
+ "epoch": 0.45,
109
+ "eval_accuracy": 0.6506941121373793,
110
+ "eval_loss": 1.8410626649856567,
111
+ "eval_runtime": 36.1001,
112
+ "eval_samples_per_second": 430.054,
113
+ "eval_steps_per_second": 53.767,
114
+ "step": 3500
115
+ },
116
+ {
117
+ "epoch": 0.52,
118
+ "learning_rate": 8.282632981008852e-06,
119
+ "loss": 1.9862,
120
+ "step": 4000
121
+ },
122
+ {
123
+ "epoch": 0.52,
124
+ "eval_accuracy": 0.6524157728010056,
125
+ "eval_loss": 1.8297162055969238,
126
+ "eval_runtime": 36.1019,
127
+ "eval_samples_per_second": 430.032,
128
+ "eval_steps_per_second": 53.764,
129
+ "step": 4000
130
+ },
131
+ {
132
+ "epoch": 0.58,
133
+ "learning_rate": 8.068230643636676e-06,
134
+ "loss": 1.9745,
135
+ "step": 4500
136
+ },
137
+ {
138
+ "epoch": 0.58,
139
+ "eval_accuracy": 0.6545309828179512,
140
+ "eval_loss": 1.8154131174087524,
141
+ "eval_runtime": 36.0473,
142
+ "eval_samples_per_second": 430.684,
143
+ "eval_steps_per_second": 53.846,
144
+ "step": 4500
145
+ },
146
+ {
147
+ "epoch": 0.64,
148
+ "learning_rate": 7.853398642261751e-06,
149
+ "loss": 1.9606,
150
+ "step": 5000
151
+ },
152
+ {
153
+ "epoch": 0.64,
154
+ "eval_accuracy": 0.6556562172935413,
155
+ "eval_loss": 1.8056122064590454,
156
+ "eval_runtime": 36.2735,
157
+ "eval_samples_per_second": 427.999,
158
+ "eval_steps_per_second": 53.51,
159
+ "step": 5000
160
+ },
161
+ {
162
+ "epoch": 0.71,
163
+ "learning_rate": 7.638996304889577e-06,
164
+ "loss": 1.9486,
165
+ "step": 5500
166
+ },
167
+ {
168
+ "epoch": 0.71,
169
+ "eval_accuracy": 0.6560147022088998,
170
+ "eval_loss": 1.8032631874084473,
171
+ "eval_runtime": 36.0743,
172
+ "eval_samples_per_second": 430.362,
173
+ "eval_steps_per_second": 53.806,
174
+ "step": 5500
175
+ },
176
+ {
177
+ "epoch": 0.77,
178
+ "learning_rate": 7.424164303514653e-06,
179
+ "loss": 1.9416,
180
+ "step": 6000
181
+ },
182
+ {
183
+ "epoch": 0.77,
184
+ "eval_accuracy": 0.6580551701728226,
185
+ "eval_loss": 1.7894020080566406,
186
+ "eval_runtime": 36.1654,
187
+ "eval_samples_per_second": 429.278,
188
+ "eval_steps_per_second": 53.67,
189
+ "step": 6000
190
+ },
191
+ {
192
+ "epoch": 0.84,
193
+ "learning_rate": 7.209332302139728e-06,
194
+ "loss": 1.9279,
195
+ "step": 6500
196
+ },
197
+ {
198
+ "epoch": 0.84,
199
+ "eval_accuracy": 0.658183904138693,
200
+ "eval_loss": 1.7848395109176636,
201
+ "eval_runtime": 36.1173,
202
+ "eval_samples_per_second": 429.849,
203
+ "eval_steps_per_second": 53.742,
204
+ "step": 6500
205
+ },
206
+ {
207
+ "epoch": 0.9,
208
+ "learning_rate": 6.9945003007648025e-06,
209
+ "loss": 1.9196,
210
+ "step": 7000
211
+ },
212
+ {
213
+ "epoch": 0.9,
214
+ "eval_accuracy": 0.6592785508757635,
215
+ "eval_loss": 1.7786365747451782,
216
+ "eval_runtime": 36.2739,
217
+ "eval_samples_per_second": 427.994,
218
+ "eval_steps_per_second": 53.51,
219
+ "step": 7000
220
+ },
221
+ {
222
+ "epoch": 0.97,
223
+ "learning_rate": 6.779668299389877e-06,
224
+ "loss": 1.9168,
225
+ "step": 7500
226
+ },
227
+ {
228
+ "epoch": 0.97,
229
+ "eval_accuracy": 0.6591822827938671,
230
+ "eval_loss": 1.7761502265930176,
231
+ "eval_runtime": 36.6269,
232
+ "eval_samples_per_second": 423.868,
233
+ "eval_steps_per_second": 52.994,
234
+ "step": 7500
235
+ },
236
+ {
237
+ "epoch": 1.03,
238
+ "learning_rate": 6.564836298014953e-06,
239
+ "loss": 1.9123,
240
+ "step": 8000
241
+ },
242
+ {
243
+ "epoch": 1.03,
244
+ "eval_accuracy": 0.6596853436378691,
245
+ "eval_loss": 1.7743586301803589,
246
+ "eval_runtime": 36.1389,
247
+ "eval_samples_per_second": 429.592,
248
+ "eval_steps_per_second": 53.709,
249
+ "step": 8000
250
+ },
251
+ {
252
+ "epoch": 1.1,
253
+ "learning_rate": 6.350004296640028e-06,
254
+ "loss": 1.8942,
255
+ "step": 8500
256
+ },
257
+ {
258
+ "epoch": 1.1,
259
+ "eval_accuracy": 0.6610733402069573,
260
+ "eval_loss": 1.7624884843826294,
261
+ "eval_runtime": 36.0335,
262
+ "eval_samples_per_second": 430.849,
263
+ "eval_steps_per_second": 53.867,
264
+ "step": 8500
265
+ },
266
+ {
267
+ "epoch": 1.16,
268
+ "learning_rate": 6.135172295265103e-06,
269
+ "loss": 1.9053,
270
+ "step": 9000
271
+ },
272
+ {
273
+ "epoch": 1.16,
274
+ "eval_accuracy": 0.662326418448169,
275
+ "eval_loss": 1.7575763463974,
276
+ "eval_runtime": 36.357,
277
+ "eval_samples_per_second": 427.016,
278
+ "eval_steps_per_second": 53.387,
279
+ "step": 9000
280
+ },
281
+ {
282
+ "epoch": 1.22,
283
+ "learning_rate": 5.9203402938901785e-06,
284
+ "loss": 1.898,
285
+ "step": 9500
286
+ },
287
+ {
288
+ "epoch": 1.22,
289
+ "eval_accuracy": 0.6620202516286527,
290
+ "eval_loss": 1.758821725845337,
291
+ "eval_runtime": 36.1788,
292
+ "eval_samples_per_second": 429.118,
293
+ "eval_steps_per_second": 53.65,
294
+ "step": 9500
295
+ },
296
+ {
297
+ "epoch": 1.29,
298
+ "learning_rate": 5.705508292515254e-06,
299
+ "loss": 1.8896,
300
+ "step": 10000
301
+ },
302
+ {
303
+ "epoch": 1.29,
304
+ "eval_accuracy": 0.6625110635175566,
305
+ "eval_loss": 1.7518248558044434,
306
+ "eval_runtime": 36.1554,
307
+ "eval_samples_per_second": 429.396,
308
+ "eval_steps_per_second": 53.685,
309
+ "step": 10000
310
+ },
311
+ {
312
+ "epoch": 1.35,
313
+ "learning_rate": 5.490676291140329e-06,
314
+ "loss": 1.8796,
315
+ "step": 10500
316
+ },
317
+ {
318
+ "epoch": 1.35,
319
+ "eval_accuracy": 0.661861605044167,
320
+ "eval_loss": 1.755669116973877,
321
+ "eval_runtime": 36.1342,
322
+ "eval_samples_per_second": 429.648,
323
+ "eval_steps_per_second": 53.716,
324
+ "step": 10500
325
+ },
326
+ {
327
+ "epoch": 1.42,
328
+ "learning_rate": 5.275844289765404e-06,
329
+ "loss": 1.8838,
330
+ "step": 11000
331
+ },
332
+ {
333
+ "epoch": 1.42,
334
+ "eval_accuracy": 0.6628265417860324,
335
+ "eval_loss": 1.7511305809020996,
336
+ "eval_runtime": 36.0117,
337
+ "eval_samples_per_second": 431.11,
338
+ "eval_steps_per_second": 53.899,
339
+ "step": 11000
340
+ },
341
+ {
342
+ "epoch": 1.48,
343
+ "learning_rate": 5.061441952393229e-06,
344
+ "loss": 1.8869,
345
+ "step": 11500
346
+ },
347
+ {
348
+ "epoch": 1.48,
349
+ "eval_accuracy": 0.6639589859082099,
350
+ "eval_loss": 1.7436553239822388,
351
+ "eval_runtime": 36.239,
352
+ "eval_samples_per_second": 428.406,
353
+ "eval_steps_per_second": 53.561,
354
+ "step": 11500
355
+ },
356
+ {
357
+ "epoch": 1.55,
358
+ "learning_rate": 4.846609951018304e-06,
359
+ "loss": 1.8756,
360
+ "step": 12000
361
+ },
362
+ {
363
+ "epoch": 1.55,
364
+ "eval_accuracy": 0.6641049700653768,
365
+ "eval_loss": 1.742509126663208,
366
+ "eval_runtime": 36.1208,
367
+ "eval_samples_per_second": 429.808,
368
+ "eval_steps_per_second": 53.736,
369
+ "step": 12000
370
+ },
371
+ {
372
+ "epoch": 1.61,
373
+ "learning_rate": 4.631777949643379e-06,
374
+ "loss": 1.8775,
375
+ "step": 12500
376
+ },
377
+ {
378
+ "epoch": 1.61,
379
+ "eval_accuracy": 0.6640769398921977,
380
+ "eval_loss": 1.7409285306930542,
381
+ "eval_runtime": 36.1893,
382
+ "eval_samples_per_second": 428.994,
383
+ "eval_steps_per_second": 53.635,
384
+ "step": 12500
385
+ },
386
+ {
387
+ "epoch": 1.68,
388
+ "learning_rate": 4.416945948268455e-06,
389
+ "loss": 1.8757,
390
+ "step": 13000
391
+ },
392
+ {
393
+ "epoch": 1.68,
394
+ "eval_accuracy": 0.664925807451965,
395
+ "eval_loss": 1.7372323274612427,
396
+ "eval_runtime": 36.0287,
397
+ "eval_samples_per_second": 430.906,
398
+ "eval_steps_per_second": 53.874,
399
+ "step": 13000
400
+ },
401
+ {
402
+ "epoch": 1.74,
403
+ "learning_rate": 4.20254361089628e-06,
404
+ "loss": 1.8616,
405
+ "step": 13500
406
+ },
407
+ {
408
+ "epoch": 1.74,
409
+ "eval_accuracy": 0.6645522086560093,
410
+ "eval_loss": 1.7387374639511108,
411
+ "eval_runtime": 36.2158,
412
+ "eval_samples_per_second": 428.68,
413
+ "eval_steps_per_second": 53.595,
414
+ "step": 13500
415
+ },
416
+ {
417
+ "epoch": 1.8,
418
+ "learning_rate": 3.987711609521355e-06,
419
+ "loss": 1.8675,
420
+ "step": 14000
421
+ },
422
+ {
423
+ "epoch": 1.8,
424
+ "eval_accuracy": 0.6648440628084251,
425
+ "eval_loss": 1.7335091829299927,
426
+ "eval_runtime": 36.1693,
427
+ "eval_samples_per_second": 429.231,
428
+ "eval_steps_per_second": 53.664,
429
+ "step": 14000
430
+ },
431
+ {
432
+ "epoch": 1.87,
433
+ "learning_rate": 3.7728796081464296e-06,
434
+ "loss": 1.8725,
435
+ "step": 14500
436
+ },
437
+ {
438
+ "epoch": 1.87,
439
+ "eval_accuracy": 0.6660341443052158,
440
+ "eval_loss": 1.728769302368164,
441
+ "eval_runtime": 36.1159,
442
+ "eval_samples_per_second": 429.866,
443
+ "eval_steps_per_second": 53.744,
444
+ "step": 14500
445
+ },
446
+ {
447
+ "epoch": 1.93,
448
+ "learning_rate": 3.558047606771505e-06,
449
+ "loss": 1.8678,
450
+ "step": 15000
451
+ },
452
+ {
453
+ "epoch": 1.93,
454
+ "eval_accuracy": 0.66591998939469,
455
+ "eval_loss": 1.730508804321289,
456
+ "eval_runtime": 36.0446,
457
+ "eval_samples_per_second": 430.716,
458
+ "eval_steps_per_second": 53.85,
459
+ "step": 15000
460
+ },
461
+ {
462
+ "epoch": 2.0,
463
+ "learning_rate": 3.34321560539658e-06,
464
+ "loss": 1.8611,
465
+ "step": 15500
466
+ },
467
+ {
468
+ "epoch": 2.0,
469
+ "eval_accuracy": 0.6666247127717294,
470
+ "eval_loss": 1.7255862951278687,
471
+ "eval_runtime": 36.3664,
472
+ "eval_samples_per_second": 426.905,
473
+ "eval_steps_per_second": 53.373,
474
+ "step": 15500
475
+ },
476
+ {
477
+ "epoch": 2.06,
478
+ "learning_rate": 3.1288132680244054e-06,
479
+ "loss": 1.853,
480
+ "step": 16000
481
+ },
482
+ {
483
+ "epoch": 2.06,
484
+ "eval_accuracy": 0.6661196617167527,
485
+ "eval_loss": 1.7286032438278198,
486
+ "eval_runtime": 36.0258,
487
+ "eval_samples_per_second": 430.941,
488
+ "eval_steps_per_second": 53.878,
489
+ "step": 16000
490
+ },
491
+ {
492
+ "epoch": 2.13,
493
+ "learning_rate": 2.9139812666494803e-06,
494
+ "loss": 1.8487,
495
+ "step": 16500
496
+ },
497
+ {
498
+ "epoch": 2.13,
499
+ "eval_accuracy": 0.6658712718524595,
500
+ "eval_loss": 1.7284834384918213,
501
+ "eval_runtime": 36.2843,
502
+ "eval_samples_per_second": 427.871,
503
+ "eval_steps_per_second": 53.494,
504
+ "step": 16500
505
+ },
506
+ {
507
+ "epoch": 2.19,
508
+ "learning_rate": 2.6991492652745556e-06,
509
+ "loss": 1.8543,
510
+ "step": 17000
511
+ },
512
+ {
513
+ "epoch": 2.19,
514
+ "eval_accuracy": 0.666799617645458,
515
+ "eval_loss": 1.7229472398757935,
516
+ "eval_runtime": 36.1382,
517
+ "eval_samples_per_second": 429.601,
518
+ "eval_steps_per_second": 53.71,
519
+ "step": 17000
520
+ },
521
+ {
522
+ "epoch": 2.26,
523
+ "learning_rate": 2.484317263899631e-06,
524
+ "loss": 1.8519,
525
+ "step": 17500
526
+ },
527
+ {
528
+ "epoch": 2.26,
529
+ "eval_accuracy": 0.6669869788832046,
530
+ "eval_loss": 1.7240232229232788,
531
+ "eval_runtime": 36.2928,
532
+ "eval_samples_per_second": 427.771,
533
+ "eval_steps_per_second": 53.482,
534
+ "step": 17500
535
+ },
536
+ {
537
+ "epoch": 2.32,
538
+ "learning_rate": 2.2694852625247057e-06,
539
+ "loss": 1.851,
540
+ "step": 18000
541
+ },
542
+ {
543
+ "epoch": 2.32,
544
+ "eval_accuracy": 0.6662479933850755,
545
+ "eval_loss": 1.7275055646896362,
546
+ "eval_runtime": 36.0625,
547
+ "eval_samples_per_second": 430.502,
548
+ "eval_steps_per_second": 53.823,
549
+ "step": 18000
550
+ },
551
+ {
552
+ "epoch": 2.38,
553
+ "learning_rate": 2.054653261149781e-06,
554
+ "loss": 1.8547,
555
+ "step": 18500
556
+ },
557
+ {
558
+ "epoch": 2.38,
559
+ "eval_accuracy": 0.6672813284171724,
560
+ "eval_loss": 1.7197449207305908,
561
+ "eval_runtime": 36.3297,
562
+ "eval_samples_per_second": 427.337,
563
+ "eval_steps_per_second": 53.427,
564
+ "step": 18500
565
+ },
566
+ {
567
+ "epoch": 2.45,
568
+ "learning_rate": 1.8398212597748563e-06,
569
+ "loss": 1.8476,
570
+ "step": 19000
571
+ },
572
+ {
573
+ "epoch": 2.45,
574
+ "eval_accuracy": 0.6674510477353122,
575
+ "eval_loss": 1.7163910865783691,
576
+ "eval_runtime": 36.0727,
577
+ "eval_samples_per_second": 430.38,
578
+ "eval_steps_per_second": 53.808,
579
+ "step": 19000
580
+ },
581
+ {
582
+ "epoch": 2.51,
583
+ "learning_rate": 1.6249892583999314e-06,
584
+ "loss": 1.8444,
585
+ "step": 19500
586
+ },
587
+ {
588
+ "epoch": 2.51,
589
+ "eval_accuracy": 0.667601267022319,
590
+ "eval_loss": 1.7213865518569946,
591
+ "eval_runtime": 36.1639,
592
+ "eval_samples_per_second": 429.296,
593
+ "eval_steps_per_second": 53.672,
594
+ "step": 19500
595
+ },
596
+ {
597
+ "epoch": 2.58,
598
+ "learning_rate": 1.4101572570250067e-06,
599
+ "loss": 1.8544,
600
+ "step": 20000
601
+ },
602
+ {
603
+ "epoch": 2.58,
604
+ "eval_accuracy": 0.6668017942219797,
605
+ "eval_loss": 1.7217011451721191,
606
+ "eval_runtime": 36.3821,
607
+ "eval_samples_per_second": 426.721,
608
+ "eval_steps_per_second": 53.35,
609
+ "step": 20000
610
+ },
611
+ {
612
+ "epoch": 2.64,
613
+ "learning_rate": 1.1953252556500817e-06,
614
+ "loss": 1.8491,
615
+ "step": 20500
616
+ },
617
+ {
618
+ "epoch": 2.64,
619
+ "eval_accuracy": 0.6678062285854136,
620
+ "eval_loss": 1.717513918876648,
621
+ "eval_runtime": 36.0279,
622
+ "eval_samples_per_second": 430.916,
623
+ "eval_steps_per_second": 53.875,
624
+ "step": 20500
625
+ },
626
+ {
627
+ "epoch": 2.64,
628
+ "step": 20500,
629
+ "total_flos": 9980146705514496.0,
630
+ "train_loss": 1.9300706578696647,
631
+ "train_runtime": 3479.9405,
632
+ "train_samples_per_second": 107.005,
633
+ "train_steps_per_second": 6.688
634
+ }
635
+ ],
636
+ "logging_steps": 500,
637
+ "max_steps": 23274,
638
+ "num_input_tokens_seen": 0,
639
+ "num_train_epochs": 3,
640
+ "save_steps": 500,
641
+ "total_flos": 9980146705514496.0,
642
+ "train_batch_size": 8,
643
+ "trial_name": null,
644
+ "trial_params": null
645
+ }