pszemraj commited on
Commit
c8bf38b
1 Parent(s): 228609b

End of training

Browse files
Files changed (4) hide show
  1. all_results.json +14 -0
  2. eval_results.json +9 -0
  3. train_results.json +8 -0
  4. trainer_state.json +1594 -0
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.94,
3
+ "eval_f1": 0.5395372489193999,
4
+ "eval_loss": 0.2657696008682251,
5
+ "eval_runtime": 30.9809,
6
+ "eval_samples": 989,
7
+ "eval_samples_per_second": 31.923,
8
+ "eval_steps_per_second": 8.005,
9
+ "train_loss": 47546.01565268776,
10
+ "train_runtime": 9264.1071,
11
+ "train_samples": 7914,
12
+ "train_samples_per_second": 8.543,
13
+ "train_steps_per_second": 0.133
14
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.94,
3
+ "eval_f1": 0.5395372489193999,
4
+ "eval_loss": 0.2657696008682251,
5
+ "eval_runtime": 30.9809,
6
+ "eval_samples": 989,
7
+ "eval_samples_per_second": 31.923,
8
+ "eval_steps_per_second": 8.005
9
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.94,
3
+ "train_loss": 47546.01565268776,
4
+ "train_runtime": 9264.1071,
5
+ "train_samples": 7914,
6
+ "train_samples_per_second": 8.543,
7
+ "train_steps_per_second": 0.133
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1594 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.944416371905003,
5
+ "eval_steps": 500,
6
+ "global_step": 1230,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.04,
13
+ "learning_rate": 4.0540540540540545e-06,
14
+ "loss": 2998113.6,
15
+ "step": 5
16
+ },
17
+ {
18
+ "epoch": 0.08,
19
+ "learning_rate": 8.108108108108109e-06,
20
+ "loss": 2797824.6,
21
+ "step": 10
22
+ },
23
+ {
24
+ "epoch": 0.12,
25
+ "learning_rate": 1.2162162162162164e-05,
26
+ "loss": 2452371.6,
27
+ "step": 15
28
+ },
29
+ {
30
+ "epoch": 0.16,
31
+ "learning_rate": 1.6216216216216218e-05,
32
+ "loss": 1871456.8,
33
+ "step": 20
34
+ },
35
+ {
36
+ "epoch": 0.2,
37
+ "learning_rate": 2.027027027027027e-05,
38
+ "loss": 1119156.9,
39
+ "step": 25
40
+ },
41
+ {
42
+ "epoch": 0.24,
43
+ "learning_rate": 2.4324324324324327e-05,
44
+ "loss": 420413.4,
45
+ "step": 30
46
+ },
47
+ {
48
+ "epoch": 0.28,
49
+ "learning_rate": 2.8378378378378378e-05,
50
+ "loss": 36048.9469,
51
+ "step": 35
52
+ },
53
+ {
54
+ "epoch": 0.32,
55
+ "learning_rate": 2.9924559932942165e-05,
56
+ "loss": 866.2308,
57
+ "step": 40
58
+ },
59
+ {
60
+ "epoch": 0.36,
61
+ "learning_rate": 2.9798826487845768e-05,
62
+ "loss": 6.5443,
63
+ "step": 45
64
+ },
65
+ {
66
+ "epoch": 0.4,
67
+ "learning_rate": 2.9673093042749373e-05,
68
+ "loss": 0.5418,
69
+ "step": 50
70
+ },
71
+ {
72
+ "epoch": 0.44,
73
+ "learning_rate": 2.9547359597652975e-05,
74
+ "loss": 0.4141,
75
+ "step": 55
76
+ },
77
+ {
78
+ "epoch": 0.49,
79
+ "learning_rate": 2.942162615255658e-05,
80
+ "loss": 0.3744,
81
+ "step": 60
82
+ },
83
+ {
84
+ "epoch": 0.53,
85
+ "learning_rate": 2.9295892707460186e-05,
86
+ "loss": 0.3621,
87
+ "step": 65
88
+ },
89
+ {
90
+ "epoch": 0.57,
91
+ "learning_rate": 2.917015926236379e-05,
92
+ "loss": 0.3423,
93
+ "step": 70
94
+ },
95
+ {
96
+ "epoch": 0.61,
97
+ "learning_rate": 2.9044425817267394e-05,
98
+ "loss": 0.3333,
99
+ "step": 75
100
+ },
101
+ {
102
+ "epoch": 0.65,
103
+ "learning_rate": 2.8918692372171e-05,
104
+ "loss": 0.3359,
105
+ "step": 80
106
+ },
107
+ {
108
+ "epoch": 0.69,
109
+ "learning_rate": 2.87929589270746e-05,
110
+ "loss": 0.3301,
111
+ "step": 85
112
+ },
113
+ {
114
+ "epoch": 0.73,
115
+ "learning_rate": 2.8667225481978207e-05,
116
+ "loss": 0.3238,
117
+ "step": 90
118
+ },
119
+ {
120
+ "epoch": 0.77,
121
+ "learning_rate": 2.854149203688181e-05,
122
+ "loss": 0.3221,
123
+ "step": 95
124
+ },
125
+ {
126
+ "epoch": 0.81,
127
+ "learning_rate": 2.8415758591785418e-05,
128
+ "loss": 0.3132,
129
+ "step": 100
130
+ },
131
+ {
132
+ "epoch": 0.85,
133
+ "learning_rate": 2.829002514668902e-05,
134
+ "loss": 0.3141,
135
+ "step": 105
136
+ },
137
+ {
138
+ "epoch": 0.89,
139
+ "learning_rate": 2.8164291701592623e-05,
140
+ "loss": 0.3019,
141
+ "step": 110
142
+ },
143
+ {
144
+ "epoch": 0.93,
145
+ "learning_rate": 2.8038558256496228e-05,
146
+ "loss": 0.324,
147
+ "step": 115
148
+ },
149
+ {
150
+ "epoch": 0.97,
151
+ "learning_rate": 2.7912824811399834e-05,
152
+ "loss": 0.3118,
153
+ "step": 120
154
+ },
155
+ {
156
+ "epoch": 0.99,
157
+ "eval_f1": 0.3004154682007031,
158
+ "eval_loss": 0.2937778830528259,
159
+ "eval_runtime": 30.7677,
160
+ "eval_samples_per_second": 32.144,
161
+ "eval_steps_per_second": 8.06,
162
+ "step": 123
163
+ },
164
+ {
165
+ "epoch": 1.01,
166
+ "learning_rate": 2.778709136630344e-05,
167
+ "loss": 0.3159,
168
+ "step": 125
169
+ },
170
+ {
171
+ "epoch": 1.05,
172
+ "learning_rate": 2.766135792120704e-05,
173
+ "loss": 0.2949,
174
+ "step": 130
175
+ },
176
+ {
177
+ "epoch": 1.09,
178
+ "learning_rate": 2.7535624476110643e-05,
179
+ "loss": 0.3043,
180
+ "step": 135
181
+ },
182
+ {
183
+ "epoch": 1.13,
184
+ "learning_rate": 2.7409891031014252e-05,
185
+ "loss": 0.3076,
186
+ "step": 140
187
+ },
188
+ {
189
+ "epoch": 1.17,
190
+ "learning_rate": 2.7284157585917854e-05,
191
+ "loss": 0.301,
192
+ "step": 145
193
+ },
194
+ {
195
+ "epoch": 1.21,
196
+ "learning_rate": 2.715842414082146e-05,
197
+ "loss": 0.3006,
198
+ "step": 150
199
+ },
200
+ {
201
+ "epoch": 1.25,
202
+ "learning_rate": 2.7032690695725062e-05,
203
+ "loss": 0.2929,
204
+ "step": 155
205
+ },
206
+ {
207
+ "epoch": 1.29,
208
+ "learning_rate": 2.6906957250628668e-05,
209
+ "loss": 0.3107,
210
+ "step": 160
211
+ },
212
+ {
213
+ "epoch": 1.33,
214
+ "learning_rate": 2.6781223805532273e-05,
215
+ "loss": 0.2988,
216
+ "step": 165
217
+ },
218
+ {
219
+ "epoch": 1.37,
220
+ "learning_rate": 2.6655490360435875e-05,
221
+ "loss": 0.3097,
222
+ "step": 170
223
+ },
224
+ {
225
+ "epoch": 1.41,
226
+ "learning_rate": 2.6529756915339484e-05,
227
+ "loss": 0.2967,
228
+ "step": 175
229
+ },
230
+ {
231
+ "epoch": 1.46,
232
+ "learning_rate": 2.6404023470243086e-05,
233
+ "loss": 0.2918,
234
+ "step": 180
235
+ },
236
+ {
237
+ "epoch": 1.5,
238
+ "learning_rate": 2.627829002514669e-05,
239
+ "loss": 0.287,
240
+ "step": 185
241
+ },
242
+ {
243
+ "epoch": 1.54,
244
+ "learning_rate": 2.6152556580050294e-05,
245
+ "loss": 0.3104,
246
+ "step": 190
247
+ },
248
+ {
249
+ "epoch": 1.58,
250
+ "learning_rate": 2.60268231349539e-05,
251
+ "loss": 0.2859,
252
+ "step": 195
253
+ },
254
+ {
255
+ "epoch": 1.62,
256
+ "learning_rate": 2.59010896898575e-05,
257
+ "loss": 0.2938,
258
+ "step": 200
259
+ },
260
+ {
261
+ "epoch": 1.66,
262
+ "learning_rate": 2.5775356244761107e-05,
263
+ "loss": 0.2884,
264
+ "step": 205
265
+ },
266
+ {
267
+ "epoch": 1.7,
268
+ "learning_rate": 2.564962279966471e-05,
269
+ "loss": 0.2858,
270
+ "step": 210
271
+ },
272
+ {
273
+ "epoch": 1.74,
274
+ "learning_rate": 2.5523889354568318e-05,
275
+ "loss": 0.2956,
276
+ "step": 215
277
+ },
278
+ {
279
+ "epoch": 1.78,
280
+ "learning_rate": 2.539815590947192e-05,
281
+ "loss": 0.2916,
282
+ "step": 220
283
+ },
284
+ {
285
+ "epoch": 1.82,
286
+ "learning_rate": 2.5272422464375523e-05,
287
+ "loss": 0.2956,
288
+ "step": 225
289
+ },
290
+ {
291
+ "epoch": 1.86,
292
+ "learning_rate": 2.5146689019279128e-05,
293
+ "loss": 0.2905,
294
+ "step": 230
295
+ },
296
+ {
297
+ "epoch": 1.9,
298
+ "learning_rate": 2.5020955574182734e-05,
299
+ "loss": 0.2843,
300
+ "step": 235
301
+ },
302
+ {
303
+ "epoch": 1.94,
304
+ "learning_rate": 2.489522212908634e-05,
305
+ "loss": 0.2878,
306
+ "step": 240
307
+ },
308
+ {
309
+ "epoch": 1.98,
310
+ "learning_rate": 2.476948868398994e-05,
311
+ "loss": 0.2732,
312
+ "step": 245
313
+ },
314
+ {
315
+ "epoch": 2.0,
316
+ "eval_f1": 0.46014293567894443,
317
+ "eval_loss": 0.27736958861351013,
318
+ "eval_runtime": 34.3798,
319
+ "eval_samples_per_second": 28.767,
320
+ "eval_steps_per_second": 7.214,
321
+ "step": 247
322
+ },
323
+ {
324
+ "epoch": 2.02,
325
+ "learning_rate": 2.4643755238893543e-05,
326
+ "loss": 0.2978,
327
+ "step": 250
328
+ },
329
+ {
330
+ "epoch": 2.06,
331
+ "learning_rate": 2.4518021793797152e-05,
332
+ "loss": 0.287,
333
+ "step": 255
334
+ },
335
+ {
336
+ "epoch": 2.1,
337
+ "learning_rate": 2.4392288348700754e-05,
338
+ "loss": 0.2841,
339
+ "step": 260
340
+ },
341
+ {
342
+ "epoch": 2.14,
343
+ "learning_rate": 2.426655490360436e-05,
344
+ "loss": 0.2755,
345
+ "step": 265
346
+ },
347
+ {
348
+ "epoch": 2.18,
349
+ "learning_rate": 2.4140821458507962e-05,
350
+ "loss": 0.2786,
351
+ "step": 270
352
+ },
353
+ {
354
+ "epoch": 2.22,
355
+ "learning_rate": 2.4015088013411568e-05,
356
+ "loss": 0.2837,
357
+ "step": 275
358
+ },
359
+ {
360
+ "epoch": 2.26,
361
+ "learning_rate": 2.3889354568315173e-05,
362
+ "loss": 0.2691,
363
+ "step": 280
364
+ },
365
+ {
366
+ "epoch": 2.3,
367
+ "learning_rate": 2.3763621123218775e-05,
368
+ "loss": 0.2785,
369
+ "step": 285
370
+ },
371
+ {
372
+ "epoch": 2.34,
373
+ "learning_rate": 2.363788767812238e-05,
374
+ "loss": 0.2863,
375
+ "step": 290
376
+ },
377
+ {
378
+ "epoch": 2.39,
379
+ "learning_rate": 2.3512154233025986e-05,
380
+ "loss": 0.277,
381
+ "step": 295
382
+ },
383
+ {
384
+ "epoch": 2.43,
385
+ "learning_rate": 2.338642078792959e-05,
386
+ "loss": 0.2755,
387
+ "step": 300
388
+ },
389
+ {
390
+ "epoch": 2.47,
391
+ "learning_rate": 2.3260687342833194e-05,
392
+ "loss": 0.2726,
393
+ "step": 305
394
+ },
395
+ {
396
+ "epoch": 2.51,
397
+ "learning_rate": 2.3134953897736796e-05,
398
+ "loss": 0.2732,
399
+ "step": 310
400
+ },
401
+ {
402
+ "epoch": 2.55,
403
+ "learning_rate": 2.3009220452640405e-05,
404
+ "loss": 0.2772,
405
+ "step": 315
406
+ },
407
+ {
408
+ "epoch": 2.59,
409
+ "learning_rate": 2.2883487007544007e-05,
410
+ "loss": 0.2853,
411
+ "step": 320
412
+ },
413
+ {
414
+ "epoch": 2.63,
415
+ "learning_rate": 2.275775356244761e-05,
416
+ "loss": 0.2803,
417
+ "step": 325
418
+ },
419
+ {
420
+ "epoch": 2.67,
421
+ "learning_rate": 2.263202011735122e-05,
422
+ "loss": 0.2777,
423
+ "step": 330
424
+ },
425
+ {
426
+ "epoch": 2.71,
427
+ "learning_rate": 2.250628667225482e-05,
428
+ "loss": 0.2728,
429
+ "step": 335
430
+ },
431
+ {
432
+ "epoch": 2.75,
433
+ "learning_rate": 2.2380553227158426e-05,
434
+ "loss": 0.2793,
435
+ "step": 340
436
+ },
437
+ {
438
+ "epoch": 2.79,
439
+ "learning_rate": 2.2254819782062028e-05,
440
+ "loss": 0.2574,
441
+ "step": 345
442
+ },
443
+ {
444
+ "epoch": 2.83,
445
+ "learning_rate": 2.2129086336965634e-05,
446
+ "loss": 0.2722,
447
+ "step": 350
448
+ },
449
+ {
450
+ "epoch": 2.87,
451
+ "learning_rate": 2.200335289186924e-05,
452
+ "loss": 0.2847,
453
+ "step": 355
454
+ },
455
+ {
456
+ "epoch": 2.91,
457
+ "learning_rate": 2.187761944677284e-05,
458
+ "loss": 0.2808,
459
+ "step": 360
460
+ },
461
+ {
462
+ "epoch": 2.95,
463
+ "learning_rate": 2.1751886001676447e-05,
464
+ "loss": 0.2837,
465
+ "step": 365
466
+ },
467
+ {
468
+ "epoch": 2.99,
469
+ "learning_rate": 2.1626152556580052e-05,
470
+ "loss": 0.2705,
471
+ "step": 370
472
+ },
473
+ {
474
+ "epoch": 3.0,
475
+ "eval_f1": 0.4702956617850236,
476
+ "eval_loss": 0.2660216987133026,
477
+ "eval_runtime": 30.702,
478
+ "eval_samples_per_second": 32.213,
479
+ "eval_steps_per_second": 8.078,
480
+ "step": 371
481
+ },
482
+ {
483
+ "epoch": 3.03,
484
+ "learning_rate": 2.1500419111483655e-05,
485
+ "loss": 0.2717,
486
+ "step": 375
487
+ },
488
+ {
489
+ "epoch": 3.07,
490
+ "learning_rate": 2.137468566638726e-05,
491
+ "loss": 0.2599,
492
+ "step": 380
493
+ },
494
+ {
495
+ "epoch": 3.11,
496
+ "learning_rate": 2.1248952221290862e-05,
497
+ "loss": 0.2668,
498
+ "step": 385
499
+ },
500
+ {
501
+ "epoch": 3.15,
502
+ "learning_rate": 2.112321877619447e-05,
503
+ "loss": 0.2689,
504
+ "step": 390
505
+ },
506
+ {
507
+ "epoch": 3.19,
508
+ "learning_rate": 2.0997485331098073e-05,
509
+ "loss": 0.2767,
510
+ "step": 395
511
+ },
512
+ {
513
+ "epoch": 3.23,
514
+ "learning_rate": 2.0871751886001675e-05,
515
+ "loss": 0.2755,
516
+ "step": 400
517
+ },
518
+ {
519
+ "epoch": 3.27,
520
+ "learning_rate": 2.074601844090528e-05,
521
+ "loss": 0.2645,
522
+ "step": 405
523
+ },
524
+ {
525
+ "epoch": 3.31,
526
+ "learning_rate": 2.0620284995808886e-05,
527
+ "loss": 0.2656,
528
+ "step": 410
529
+ },
530
+ {
531
+ "epoch": 3.36,
532
+ "learning_rate": 2.0494551550712492e-05,
533
+ "loss": 0.2691,
534
+ "step": 415
535
+ },
536
+ {
537
+ "epoch": 3.4,
538
+ "learning_rate": 2.0368818105616094e-05,
539
+ "loss": 0.2641,
540
+ "step": 420
541
+ },
542
+ {
543
+ "epoch": 3.44,
544
+ "learning_rate": 2.0243084660519696e-05,
545
+ "loss": 0.2576,
546
+ "step": 425
547
+ },
548
+ {
549
+ "epoch": 3.48,
550
+ "learning_rate": 2.0117351215423305e-05,
551
+ "loss": 0.2577,
552
+ "step": 430
553
+ },
554
+ {
555
+ "epoch": 3.52,
556
+ "learning_rate": 1.9991617770326907e-05,
557
+ "loss": 0.2609,
558
+ "step": 435
559
+ },
560
+ {
561
+ "epoch": 3.56,
562
+ "learning_rate": 1.986588432523051e-05,
563
+ "loss": 0.2558,
564
+ "step": 440
565
+ },
566
+ {
567
+ "epoch": 3.6,
568
+ "learning_rate": 1.9740150880134115e-05,
569
+ "loss": 0.2645,
570
+ "step": 445
571
+ },
572
+ {
573
+ "epoch": 3.64,
574
+ "learning_rate": 1.961441743503772e-05,
575
+ "loss": 0.2698,
576
+ "step": 450
577
+ },
578
+ {
579
+ "epoch": 3.68,
580
+ "learning_rate": 1.9488683989941326e-05,
581
+ "loss": 0.26,
582
+ "step": 455
583
+ },
584
+ {
585
+ "epoch": 3.72,
586
+ "learning_rate": 1.9362950544844928e-05,
587
+ "loss": 0.2641,
588
+ "step": 460
589
+ },
590
+ {
591
+ "epoch": 3.76,
592
+ "learning_rate": 1.923721709974853e-05,
593
+ "loss": 0.2672,
594
+ "step": 465
595
+ },
596
+ {
597
+ "epoch": 3.8,
598
+ "learning_rate": 1.911148365465214e-05,
599
+ "loss": 0.2592,
600
+ "step": 470
601
+ },
602
+ {
603
+ "epoch": 3.84,
604
+ "learning_rate": 1.898575020955574e-05,
605
+ "loss": 0.2662,
606
+ "step": 475
607
+ },
608
+ {
609
+ "epoch": 3.88,
610
+ "learning_rate": 1.8860016764459347e-05,
611
+ "loss": 0.26,
612
+ "step": 480
613
+ },
614
+ {
615
+ "epoch": 3.92,
616
+ "learning_rate": 1.8734283319362952e-05,
617
+ "loss": 0.264,
618
+ "step": 485
619
+ },
620
+ {
621
+ "epoch": 3.96,
622
+ "learning_rate": 1.8608549874266555e-05,
623
+ "loss": 0.2671,
624
+ "step": 490
625
+ },
626
+ {
627
+ "epoch": 3.99,
628
+ "eval_f1": 0.4754279403644394,
629
+ "eval_loss": 0.2679407000541687,
630
+ "eval_runtime": 34.3196,
631
+ "eval_samples_per_second": 28.817,
632
+ "eval_steps_per_second": 7.226,
633
+ "step": 494
634
+ },
635
+ {
636
+ "epoch": 4.0,
637
+ "learning_rate": 1.848281642917016e-05,
638
+ "loss": 0.2603,
639
+ "step": 495
640
+ },
641
+ {
642
+ "epoch": 4.04,
643
+ "learning_rate": 1.8357082984073762e-05,
644
+ "loss": 0.2476,
645
+ "step": 500
646
+ },
647
+ {
648
+ "epoch": 4.08,
649
+ "learning_rate": 1.823134953897737e-05,
650
+ "loss": 0.2623,
651
+ "step": 505
652
+ },
653
+ {
654
+ "epoch": 4.12,
655
+ "learning_rate": 1.8105616093880973e-05,
656
+ "loss": 0.2614,
657
+ "step": 510
658
+ },
659
+ {
660
+ "epoch": 4.16,
661
+ "learning_rate": 1.7979882648784575e-05,
662
+ "loss": 0.2405,
663
+ "step": 515
664
+ },
665
+ {
666
+ "epoch": 4.2,
667
+ "learning_rate": 1.785414920368818e-05,
668
+ "loss": 0.2519,
669
+ "step": 520
670
+ },
671
+ {
672
+ "epoch": 4.24,
673
+ "learning_rate": 1.7728415758591787e-05,
674
+ "loss": 0.2499,
675
+ "step": 525
676
+ },
677
+ {
678
+ "epoch": 4.28,
679
+ "learning_rate": 1.7602682313495392e-05,
680
+ "loss": 0.2678,
681
+ "step": 530
682
+ },
683
+ {
684
+ "epoch": 4.33,
685
+ "learning_rate": 1.7476948868398994e-05,
686
+ "loss": 0.2639,
687
+ "step": 535
688
+ },
689
+ {
690
+ "epoch": 4.37,
691
+ "learning_rate": 1.7351215423302596e-05,
692
+ "loss": 0.2554,
693
+ "step": 540
694
+ },
695
+ {
696
+ "epoch": 4.41,
697
+ "learning_rate": 1.7225481978206205e-05,
698
+ "loss": 0.2615,
699
+ "step": 545
700
+ },
701
+ {
702
+ "epoch": 4.45,
703
+ "learning_rate": 1.7099748533109807e-05,
704
+ "loss": 0.2518,
705
+ "step": 550
706
+ },
707
+ {
708
+ "epoch": 4.49,
709
+ "learning_rate": 1.6974015088013413e-05,
710
+ "loss": 0.2612,
711
+ "step": 555
712
+ },
713
+ {
714
+ "epoch": 4.53,
715
+ "learning_rate": 1.6848281642917015e-05,
716
+ "loss": 0.2608,
717
+ "step": 560
718
+ },
719
+ {
720
+ "epoch": 4.57,
721
+ "learning_rate": 1.672254819782062e-05,
722
+ "loss": 0.2446,
723
+ "step": 565
724
+ },
725
+ {
726
+ "epoch": 4.61,
727
+ "learning_rate": 1.6596814752724226e-05,
728
+ "loss": 0.2556,
729
+ "step": 570
730
+ },
731
+ {
732
+ "epoch": 4.65,
733
+ "learning_rate": 1.6471081307627828e-05,
734
+ "loss": 0.2477,
735
+ "step": 575
736
+ },
737
+ {
738
+ "epoch": 4.69,
739
+ "learning_rate": 1.6345347862531434e-05,
740
+ "loss": 0.2649,
741
+ "step": 580
742
+ },
743
+ {
744
+ "epoch": 4.73,
745
+ "learning_rate": 1.621961441743504e-05,
746
+ "loss": 0.2611,
747
+ "step": 585
748
+ },
749
+ {
750
+ "epoch": 4.77,
751
+ "learning_rate": 1.609388097233864e-05,
752
+ "loss": 0.259,
753
+ "step": 590
754
+ },
755
+ {
756
+ "epoch": 4.81,
757
+ "learning_rate": 1.5968147527242247e-05,
758
+ "loss": 0.2618,
759
+ "step": 595
760
+ },
761
+ {
762
+ "epoch": 4.85,
763
+ "learning_rate": 1.584241408214585e-05,
764
+ "loss": 0.2478,
765
+ "step": 600
766
+ },
767
+ {
768
+ "epoch": 4.89,
769
+ "learning_rate": 1.5716680637049458e-05,
770
+ "loss": 0.2542,
771
+ "step": 605
772
+ },
773
+ {
774
+ "epoch": 4.93,
775
+ "learning_rate": 1.559094719195306e-05,
776
+ "loss": 0.2488,
777
+ "step": 610
778
+ },
779
+ {
780
+ "epoch": 4.97,
781
+ "learning_rate": 1.5465213746856662e-05,
782
+ "loss": 0.2529,
783
+ "step": 615
784
+ },
785
+ {
786
+ "epoch": 5.0,
787
+ "eval_f1": 0.5162986330178759,
788
+ "eval_loss": 0.2654779851436615,
789
+ "eval_runtime": 32.5281,
790
+ "eval_samples_per_second": 30.405,
791
+ "eval_steps_per_second": 7.624,
792
+ "step": 618
793
+ },
794
+ {
795
+ "epoch": 5.01,
796
+ "learning_rate": 1.533948030176027e-05,
797
+ "loss": 0.2474,
798
+ "step": 620
799
+ },
800
+ {
801
+ "epoch": 5.05,
802
+ "learning_rate": 1.5213746856663873e-05,
803
+ "loss": 0.2442,
804
+ "step": 625
805
+ },
806
+ {
807
+ "epoch": 5.09,
808
+ "learning_rate": 1.5088013411567477e-05,
809
+ "loss": 0.2494,
810
+ "step": 630
811
+ },
812
+ {
813
+ "epoch": 5.13,
814
+ "learning_rate": 1.4962279966471083e-05,
815
+ "loss": 0.2505,
816
+ "step": 635
817
+ },
818
+ {
819
+ "epoch": 5.17,
820
+ "learning_rate": 1.4836546521374687e-05,
821
+ "loss": 0.2345,
822
+ "step": 640
823
+ },
824
+ {
825
+ "epoch": 5.21,
826
+ "learning_rate": 1.471081307627829e-05,
827
+ "loss": 0.2491,
828
+ "step": 645
829
+ },
830
+ {
831
+ "epoch": 5.26,
832
+ "learning_rate": 1.4585079631181894e-05,
833
+ "loss": 0.2455,
834
+ "step": 650
835
+ },
836
+ {
837
+ "epoch": 5.3,
838
+ "learning_rate": 1.44593461860855e-05,
839
+ "loss": 0.2483,
840
+ "step": 655
841
+ },
842
+ {
843
+ "epoch": 5.34,
844
+ "learning_rate": 1.4333612740989104e-05,
845
+ "loss": 0.2387,
846
+ "step": 660
847
+ },
848
+ {
849
+ "epoch": 5.38,
850
+ "learning_rate": 1.4207879295892709e-05,
851
+ "loss": 0.2568,
852
+ "step": 665
853
+ },
854
+ {
855
+ "epoch": 5.42,
856
+ "learning_rate": 1.4082145850796311e-05,
857
+ "loss": 0.2539,
858
+ "step": 670
859
+ },
860
+ {
861
+ "epoch": 5.46,
862
+ "learning_rate": 1.3956412405699917e-05,
863
+ "loss": 0.2306,
864
+ "step": 675
865
+ },
866
+ {
867
+ "epoch": 5.5,
868
+ "learning_rate": 1.383067896060352e-05,
869
+ "loss": 0.2453,
870
+ "step": 680
871
+ },
872
+ {
873
+ "epoch": 5.54,
874
+ "learning_rate": 1.3704945515507126e-05,
875
+ "loss": 0.2411,
876
+ "step": 685
877
+ },
878
+ {
879
+ "epoch": 5.58,
880
+ "learning_rate": 1.357921207041073e-05,
881
+ "loss": 0.2526,
882
+ "step": 690
883
+ },
884
+ {
885
+ "epoch": 5.62,
886
+ "learning_rate": 1.3453478625314334e-05,
887
+ "loss": 0.2384,
888
+ "step": 695
889
+ },
890
+ {
891
+ "epoch": 5.66,
892
+ "learning_rate": 1.3327745180217938e-05,
893
+ "loss": 0.2486,
894
+ "step": 700
895
+ },
896
+ {
897
+ "epoch": 5.7,
898
+ "learning_rate": 1.3202011735121543e-05,
899
+ "loss": 0.2357,
900
+ "step": 705
901
+ },
902
+ {
903
+ "epoch": 5.74,
904
+ "learning_rate": 1.3076278290025147e-05,
905
+ "loss": 0.2484,
906
+ "step": 710
907
+ },
908
+ {
909
+ "epoch": 5.78,
910
+ "learning_rate": 1.295054484492875e-05,
911
+ "loss": 0.2405,
912
+ "step": 715
913
+ },
914
+ {
915
+ "epoch": 5.82,
916
+ "learning_rate": 1.2824811399832355e-05,
917
+ "loss": 0.2521,
918
+ "step": 720
919
+ },
920
+ {
921
+ "epoch": 5.86,
922
+ "learning_rate": 1.269907795473596e-05,
923
+ "loss": 0.2575,
924
+ "step": 725
925
+ },
926
+ {
927
+ "epoch": 5.9,
928
+ "learning_rate": 1.2573344509639564e-05,
929
+ "loss": 0.2516,
930
+ "step": 730
931
+ },
932
+ {
933
+ "epoch": 5.94,
934
+ "learning_rate": 1.244761106454317e-05,
935
+ "loss": 0.2353,
936
+ "step": 735
937
+ },
938
+ {
939
+ "epoch": 5.98,
940
+ "learning_rate": 1.2321877619446772e-05,
941
+ "loss": 0.2395,
942
+ "step": 740
943
+ },
944
+ {
945
+ "epoch": 6.0,
946
+ "eval_f1": 0.5213319458896981,
947
+ "eval_loss": 0.26373252272605896,
948
+ "eval_runtime": 29.187,
949
+ "eval_samples_per_second": 33.885,
950
+ "eval_steps_per_second": 8.497,
951
+ "step": 742
952
+ },
953
+ {
954
+ "epoch": 6.02,
955
+ "learning_rate": 1.2196144174350377e-05,
956
+ "loss": 0.256,
957
+ "step": 745
958
+ },
959
+ {
960
+ "epoch": 6.06,
961
+ "learning_rate": 1.2070410729253981e-05,
962
+ "loss": 0.2359,
963
+ "step": 750
964
+ },
965
+ {
966
+ "epoch": 6.1,
967
+ "learning_rate": 1.1944677284157587e-05,
968
+ "loss": 0.239,
969
+ "step": 755
970
+ },
971
+ {
972
+ "epoch": 6.14,
973
+ "learning_rate": 1.181894383906119e-05,
974
+ "loss": 0.2485,
975
+ "step": 760
976
+ },
977
+ {
978
+ "epoch": 6.18,
979
+ "learning_rate": 1.1693210393964794e-05,
980
+ "loss": 0.254,
981
+ "step": 765
982
+ },
983
+ {
984
+ "epoch": 6.23,
985
+ "learning_rate": 1.1567476948868398e-05,
986
+ "loss": 0.2459,
987
+ "step": 770
988
+ },
989
+ {
990
+ "epoch": 6.27,
991
+ "learning_rate": 1.1441743503772004e-05,
992
+ "loss": 0.2328,
993
+ "step": 775
994
+ },
995
+ {
996
+ "epoch": 6.31,
997
+ "learning_rate": 1.131601005867561e-05,
998
+ "loss": 0.2422,
999
+ "step": 780
1000
+ },
1001
+ {
1002
+ "epoch": 6.35,
1003
+ "learning_rate": 1.1190276613579213e-05,
1004
+ "loss": 0.2296,
1005
+ "step": 785
1006
+ },
1007
+ {
1008
+ "epoch": 6.39,
1009
+ "learning_rate": 1.1064543168482817e-05,
1010
+ "loss": 0.2375,
1011
+ "step": 790
1012
+ },
1013
+ {
1014
+ "epoch": 6.43,
1015
+ "learning_rate": 1.093880972338642e-05,
1016
+ "loss": 0.2341,
1017
+ "step": 795
1018
+ },
1019
+ {
1020
+ "epoch": 6.47,
1021
+ "learning_rate": 1.0813076278290026e-05,
1022
+ "loss": 0.2424,
1023
+ "step": 800
1024
+ },
1025
+ {
1026
+ "epoch": 6.51,
1027
+ "learning_rate": 1.068734283319363e-05,
1028
+ "loss": 0.2253,
1029
+ "step": 805
1030
+ },
1031
+ {
1032
+ "epoch": 6.55,
1033
+ "learning_rate": 1.0561609388097236e-05,
1034
+ "loss": 0.2386,
1035
+ "step": 810
1036
+ },
1037
+ {
1038
+ "epoch": 6.59,
1039
+ "learning_rate": 1.0435875943000838e-05,
1040
+ "loss": 0.2546,
1041
+ "step": 815
1042
+ },
1043
+ {
1044
+ "epoch": 6.63,
1045
+ "learning_rate": 1.0310142497904443e-05,
1046
+ "loss": 0.224,
1047
+ "step": 820
1048
+ },
1049
+ {
1050
+ "epoch": 6.67,
1051
+ "learning_rate": 1.0184409052808047e-05,
1052
+ "loss": 0.2337,
1053
+ "step": 825
1054
+ },
1055
+ {
1056
+ "epoch": 6.71,
1057
+ "learning_rate": 1.0058675607711653e-05,
1058
+ "loss": 0.2415,
1059
+ "step": 830
1060
+ },
1061
+ {
1062
+ "epoch": 6.75,
1063
+ "learning_rate": 9.932942162615255e-06,
1064
+ "loss": 0.2309,
1065
+ "step": 835
1066
+ },
1067
+ {
1068
+ "epoch": 6.79,
1069
+ "learning_rate": 9.80720871751886e-06,
1070
+ "loss": 0.2257,
1071
+ "step": 840
1072
+ },
1073
+ {
1074
+ "epoch": 6.83,
1075
+ "learning_rate": 9.681475272422464e-06,
1076
+ "loss": 0.2373,
1077
+ "step": 845
1078
+ },
1079
+ {
1080
+ "epoch": 6.87,
1081
+ "learning_rate": 9.55574182732607e-06,
1082
+ "loss": 0.2483,
1083
+ "step": 850
1084
+ },
1085
+ {
1086
+ "epoch": 6.91,
1087
+ "learning_rate": 9.430008382229673e-06,
1088
+ "loss": 0.2437,
1089
+ "step": 855
1090
+ },
1091
+ {
1092
+ "epoch": 6.95,
1093
+ "learning_rate": 9.304274937133277e-06,
1094
+ "loss": 0.2344,
1095
+ "step": 860
1096
+ },
1097
+ {
1098
+ "epoch": 6.99,
1099
+ "learning_rate": 9.178541492036881e-06,
1100
+ "loss": 0.2469,
1101
+ "step": 865
1102
+ },
1103
+ {
1104
+ "epoch": 6.99,
1105
+ "eval_f1": 0.5234760051880674,
1106
+ "eval_loss": 0.26504993438720703,
1107
+ "eval_runtime": 33.4377,
1108
+ "eval_samples_per_second": 29.577,
1109
+ "eval_steps_per_second": 7.417,
1110
+ "step": 865
1111
+ },
1112
+ {
1113
+ "epoch": 7.03,
1114
+ "learning_rate": 9.052808046940487e-06,
1115
+ "loss": 0.2417,
1116
+ "step": 870
1117
+ },
1118
+ {
1119
+ "epoch": 7.07,
1120
+ "learning_rate": 8.92707460184409e-06,
1121
+ "loss": 0.2271,
1122
+ "step": 875
1123
+ },
1124
+ {
1125
+ "epoch": 7.11,
1126
+ "learning_rate": 8.801341156747696e-06,
1127
+ "loss": 0.2289,
1128
+ "step": 880
1129
+ },
1130
+ {
1131
+ "epoch": 7.16,
1132
+ "learning_rate": 8.675607711651298e-06,
1133
+ "loss": 0.2285,
1134
+ "step": 885
1135
+ },
1136
+ {
1137
+ "epoch": 7.2,
1138
+ "learning_rate": 8.549874266554904e-06,
1139
+ "loss": 0.2377,
1140
+ "step": 890
1141
+ },
1142
+ {
1143
+ "epoch": 7.24,
1144
+ "learning_rate": 8.424140821458508e-06,
1145
+ "loss": 0.2371,
1146
+ "step": 895
1147
+ },
1148
+ {
1149
+ "epoch": 7.28,
1150
+ "learning_rate": 8.298407376362113e-06,
1151
+ "loss": 0.2407,
1152
+ "step": 900
1153
+ },
1154
+ {
1155
+ "epoch": 7.32,
1156
+ "learning_rate": 8.172673931265717e-06,
1157
+ "loss": 0.2325,
1158
+ "step": 905
1159
+ },
1160
+ {
1161
+ "epoch": 7.36,
1162
+ "learning_rate": 8.04694048616932e-06,
1163
+ "loss": 0.2289,
1164
+ "step": 910
1165
+ },
1166
+ {
1167
+ "epoch": 7.4,
1168
+ "learning_rate": 7.921207041072925e-06,
1169
+ "loss": 0.2383,
1170
+ "step": 915
1171
+ },
1172
+ {
1173
+ "epoch": 7.44,
1174
+ "learning_rate": 7.79547359597653e-06,
1175
+ "loss": 0.2312,
1176
+ "step": 920
1177
+ },
1178
+ {
1179
+ "epoch": 7.48,
1180
+ "learning_rate": 7.669740150880136e-06,
1181
+ "loss": 0.2255,
1182
+ "step": 925
1183
+ },
1184
+ {
1185
+ "epoch": 7.52,
1186
+ "learning_rate": 7.544006705783739e-06,
1187
+ "loss": 0.2387,
1188
+ "step": 930
1189
+ },
1190
+ {
1191
+ "epoch": 7.56,
1192
+ "learning_rate": 7.418273260687343e-06,
1193
+ "loss": 0.2328,
1194
+ "step": 935
1195
+ },
1196
+ {
1197
+ "epoch": 7.6,
1198
+ "learning_rate": 7.292539815590947e-06,
1199
+ "loss": 0.2359,
1200
+ "step": 940
1201
+ },
1202
+ {
1203
+ "epoch": 7.64,
1204
+ "learning_rate": 7.166806370494552e-06,
1205
+ "loss": 0.2322,
1206
+ "step": 945
1207
+ },
1208
+ {
1209
+ "epoch": 7.68,
1210
+ "learning_rate": 7.041072925398156e-06,
1211
+ "loss": 0.2372,
1212
+ "step": 950
1213
+ },
1214
+ {
1215
+ "epoch": 7.72,
1216
+ "learning_rate": 6.91533948030176e-06,
1217
+ "loss": 0.2297,
1218
+ "step": 955
1219
+ },
1220
+ {
1221
+ "epoch": 7.76,
1222
+ "learning_rate": 6.789606035205365e-06,
1223
+ "loss": 0.2345,
1224
+ "step": 960
1225
+ },
1226
+ {
1227
+ "epoch": 7.8,
1228
+ "learning_rate": 6.663872590108969e-06,
1229
+ "loss": 0.2229,
1230
+ "step": 965
1231
+ },
1232
+ {
1233
+ "epoch": 7.84,
1234
+ "learning_rate": 6.5381391450125735e-06,
1235
+ "loss": 0.2395,
1236
+ "step": 970
1237
+ },
1238
+ {
1239
+ "epoch": 7.88,
1240
+ "learning_rate": 6.412405699916177e-06,
1241
+ "loss": 0.2395,
1242
+ "step": 975
1243
+ },
1244
+ {
1245
+ "epoch": 7.92,
1246
+ "learning_rate": 6.286672254819782e-06,
1247
+ "loss": 0.2226,
1248
+ "step": 980
1249
+ },
1250
+ {
1251
+ "epoch": 7.96,
1252
+ "learning_rate": 6.160938809723386e-06,
1253
+ "loss": 0.2301,
1254
+ "step": 985
1255
+ },
1256
+ {
1257
+ "epoch": 8.0,
1258
+ "eval_f1": 0.5250844814140889,
1259
+ "eval_loss": 0.2636159062385559,
1260
+ "eval_runtime": 39.4887,
1261
+ "eval_samples_per_second": 25.045,
1262
+ "eval_steps_per_second": 6.28,
1263
+ "step": 989
1264
+ },
1265
+ {
1266
+ "epoch": 8.0,
1267
+ "learning_rate": 6.0352053646269905e-06,
1268
+ "loss": 0.2295,
1269
+ "step": 990
1270
+ },
1271
+ {
1272
+ "epoch": 8.04,
1273
+ "learning_rate": 5.909471919530595e-06,
1274
+ "loss": 0.2384,
1275
+ "step": 995
1276
+ },
1277
+ {
1278
+ "epoch": 8.08,
1279
+ "learning_rate": 5.783738474434199e-06,
1280
+ "loss": 0.2277,
1281
+ "step": 1000
1282
+ },
1283
+ {
1284
+ "epoch": 8.13,
1285
+ "learning_rate": 5.658005029337805e-06,
1286
+ "loss": 0.2295,
1287
+ "step": 1005
1288
+ },
1289
+ {
1290
+ "epoch": 8.17,
1291
+ "learning_rate": 5.532271584241408e-06,
1292
+ "loss": 0.2323,
1293
+ "step": 1010
1294
+ },
1295
+ {
1296
+ "epoch": 8.21,
1297
+ "learning_rate": 5.406538139145013e-06,
1298
+ "loss": 0.2292,
1299
+ "step": 1015
1300
+ },
1301
+ {
1302
+ "epoch": 8.25,
1303
+ "learning_rate": 5.280804694048618e-06,
1304
+ "loss": 0.2265,
1305
+ "step": 1020
1306
+ },
1307
+ {
1308
+ "epoch": 8.29,
1309
+ "learning_rate": 5.155071248952222e-06,
1310
+ "loss": 0.2213,
1311
+ "step": 1025
1312
+ },
1313
+ {
1314
+ "epoch": 8.33,
1315
+ "learning_rate": 5.029337803855826e-06,
1316
+ "loss": 0.219,
1317
+ "step": 1030
1318
+ },
1319
+ {
1320
+ "epoch": 8.37,
1321
+ "learning_rate": 4.90360435875943e-06,
1322
+ "loss": 0.2201,
1323
+ "step": 1035
1324
+ },
1325
+ {
1326
+ "epoch": 8.41,
1327
+ "learning_rate": 4.777870913663035e-06,
1328
+ "loss": 0.2318,
1329
+ "step": 1040
1330
+ },
1331
+ {
1332
+ "epoch": 8.45,
1333
+ "learning_rate": 4.652137468566639e-06,
1334
+ "loss": 0.2316,
1335
+ "step": 1045
1336
+ },
1337
+ {
1338
+ "epoch": 8.49,
1339
+ "learning_rate": 4.526404023470243e-06,
1340
+ "loss": 0.2279,
1341
+ "step": 1050
1342
+ },
1343
+ {
1344
+ "epoch": 8.53,
1345
+ "learning_rate": 4.400670578373848e-06,
1346
+ "loss": 0.2237,
1347
+ "step": 1055
1348
+ },
1349
+ {
1350
+ "epoch": 8.57,
1351
+ "learning_rate": 4.274937133277452e-06,
1352
+ "loss": 0.2469,
1353
+ "step": 1060
1354
+ },
1355
+ {
1356
+ "epoch": 8.61,
1357
+ "learning_rate": 4.1492036881810565e-06,
1358
+ "loss": 0.224,
1359
+ "step": 1065
1360
+ },
1361
+ {
1362
+ "epoch": 8.65,
1363
+ "learning_rate": 4.02347024308466e-06,
1364
+ "loss": 0.2277,
1365
+ "step": 1070
1366
+ },
1367
+ {
1368
+ "epoch": 8.69,
1369
+ "learning_rate": 3.897736797988265e-06,
1370
+ "loss": 0.2334,
1371
+ "step": 1075
1372
+ },
1373
+ {
1374
+ "epoch": 8.73,
1375
+ "learning_rate": 3.7720033528918693e-06,
1376
+ "loss": 0.2291,
1377
+ "step": 1080
1378
+ },
1379
+ {
1380
+ "epoch": 8.77,
1381
+ "learning_rate": 3.6462699077954736e-06,
1382
+ "loss": 0.2333,
1383
+ "step": 1085
1384
+ },
1385
+ {
1386
+ "epoch": 8.81,
1387
+ "learning_rate": 3.520536462699078e-06,
1388
+ "loss": 0.2358,
1389
+ "step": 1090
1390
+ },
1391
+ {
1392
+ "epoch": 8.85,
1393
+ "learning_rate": 3.3948030176026825e-06,
1394
+ "loss": 0.2314,
1395
+ "step": 1095
1396
+ },
1397
+ {
1398
+ "epoch": 8.89,
1399
+ "learning_rate": 3.2690695725062868e-06,
1400
+ "loss": 0.2263,
1401
+ "step": 1100
1402
+ },
1403
+ {
1404
+ "epoch": 8.93,
1405
+ "learning_rate": 3.143336127409891e-06,
1406
+ "loss": 0.2274,
1407
+ "step": 1105
1408
+ },
1409
+ {
1410
+ "epoch": 8.97,
1411
+ "learning_rate": 3.0176026823134953e-06,
1412
+ "loss": 0.2273,
1413
+ "step": 1110
1414
+ },
1415
+ {
1416
+ "epoch": 9.0,
1417
+ "eval_f1": 0.5394001016776816,
1418
+ "eval_loss": 0.26480013132095337,
1419
+ "eval_runtime": 32.8402,
1420
+ "eval_samples_per_second": 30.116,
1421
+ "eval_steps_per_second": 7.552,
1422
+ "step": 1113
1423
+ },
1424
+ {
1425
+ "epoch": 9.01,
1426
+ "learning_rate": 2.8918692372170995e-06,
1427
+ "loss": 0.2256,
1428
+ "step": 1115
1429
+ },
1430
+ {
1431
+ "epoch": 9.06,
1432
+ "learning_rate": 2.766135792120704e-06,
1433
+ "loss": 0.2162,
1434
+ "step": 1120
1435
+ },
1436
+ {
1437
+ "epoch": 9.1,
1438
+ "learning_rate": 2.640402347024309e-06,
1439
+ "loss": 0.231,
1440
+ "step": 1125
1441
+ },
1442
+ {
1443
+ "epoch": 9.14,
1444
+ "learning_rate": 2.514668901927913e-06,
1445
+ "loss": 0.2298,
1446
+ "step": 1130
1447
+ },
1448
+ {
1449
+ "epoch": 9.18,
1450
+ "learning_rate": 2.3889354568315174e-06,
1451
+ "loss": 0.2291,
1452
+ "step": 1135
1453
+ },
1454
+ {
1455
+ "epoch": 9.22,
1456
+ "learning_rate": 2.2632020117351217e-06,
1457
+ "loss": 0.2263,
1458
+ "step": 1140
1459
+ },
1460
+ {
1461
+ "epoch": 9.26,
1462
+ "learning_rate": 2.137468566638726e-06,
1463
+ "loss": 0.2211,
1464
+ "step": 1145
1465
+ },
1466
+ {
1467
+ "epoch": 9.3,
1468
+ "learning_rate": 2.01173512154233e-06,
1469
+ "loss": 0.2302,
1470
+ "step": 1150
1471
+ },
1472
+ {
1473
+ "epoch": 9.34,
1474
+ "learning_rate": 1.8860016764459346e-06,
1475
+ "loss": 0.2218,
1476
+ "step": 1155
1477
+ },
1478
+ {
1479
+ "epoch": 9.38,
1480
+ "learning_rate": 1.760268231349539e-06,
1481
+ "loss": 0.2366,
1482
+ "step": 1160
1483
+ },
1484
+ {
1485
+ "epoch": 9.42,
1486
+ "learning_rate": 1.6345347862531434e-06,
1487
+ "loss": 0.2307,
1488
+ "step": 1165
1489
+ },
1490
+ {
1491
+ "epoch": 9.46,
1492
+ "learning_rate": 1.5088013411567476e-06,
1493
+ "loss": 0.234,
1494
+ "step": 1170
1495
+ },
1496
+ {
1497
+ "epoch": 9.5,
1498
+ "learning_rate": 1.383067896060352e-06,
1499
+ "loss": 0.2296,
1500
+ "step": 1175
1501
+ },
1502
+ {
1503
+ "epoch": 9.54,
1504
+ "learning_rate": 1.2573344509639566e-06,
1505
+ "loss": 0.2307,
1506
+ "step": 1180
1507
+ },
1508
+ {
1509
+ "epoch": 9.58,
1510
+ "learning_rate": 1.1316010058675608e-06,
1511
+ "loss": 0.2234,
1512
+ "step": 1185
1513
+ },
1514
+ {
1515
+ "epoch": 9.62,
1516
+ "learning_rate": 1.005867560771165e-06,
1517
+ "loss": 0.2299,
1518
+ "step": 1190
1519
+ },
1520
+ {
1521
+ "epoch": 9.66,
1522
+ "learning_rate": 8.801341156747695e-07,
1523
+ "loss": 0.2325,
1524
+ "step": 1195
1525
+ },
1526
+ {
1527
+ "epoch": 9.7,
1528
+ "learning_rate": 7.544006705783738e-07,
1529
+ "loss": 0.2238,
1530
+ "step": 1200
1531
+ },
1532
+ {
1533
+ "epoch": 9.74,
1534
+ "learning_rate": 6.286672254819783e-07,
1535
+ "loss": 0.2293,
1536
+ "step": 1205
1537
+ },
1538
+ {
1539
+ "epoch": 9.78,
1540
+ "learning_rate": 5.029337803855825e-07,
1541
+ "loss": 0.2265,
1542
+ "step": 1210
1543
+ },
1544
+ {
1545
+ "epoch": 9.82,
1546
+ "learning_rate": 3.772003352891869e-07,
1547
+ "loss": 0.2245,
1548
+ "step": 1215
1549
+ },
1550
+ {
1551
+ "epoch": 9.86,
1552
+ "learning_rate": 2.5146689019279127e-07,
1553
+ "loss": 0.2133,
1554
+ "step": 1220
1555
+ },
1556
+ {
1557
+ "epoch": 9.9,
1558
+ "learning_rate": 1.2573344509639564e-07,
1559
+ "loss": 0.2202,
1560
+ "step": 1225
1561
+ },
1562
+ {
1563
+ "epoch": 9.94,
1564
+ "learning_rate": 0.0,
1565
+ "loss": 0.2169,
1566
+ "step": 1230
1567
+ },
1568
+ {
1569
+ "epoch": 9.94,
1570
+ "eval_f1": 0.5395372489193999,
1571
+ "eval_loss": 0.2657696008682251,
1572
+ "eval_runtime": 31.9099,
1573
+ "eval_samples_per_second": 30.993,
1574
+ "eval_steps_per_second": 7.772,
1575
+ "step": 1230
1576
+ },
1577
+ {
1578
+ "epoch": 9.94,
1579
+ "step": 1230,
1580
+ "total_flos": 462869108028288.0,
1581
+ "train_loss": 47546.01565268776,
1582
+ "train_runtime": 9264.1071,
1583
+ "train_samples_per_second": 8.543,
1584
+ "train_steps_per_second": 0.133
1585
+ }
1586
+ ],
1587
+ "logging_steps": 5,
1588
+ "max_steps": 1230,
1589
+ "num_train_epochs": 10,
1590
+ "save_steps": 500,
1591
+ "total_flos": 462869108028288.0,
1592
+ "trial_name": null,
1593
+ "trial_params": null
1594
+ }