nbasatish commited on
Commit
5404319
1 Parent(s): 2effa34

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1166 -0
trainer_state.json ADDED
@@ -0,0 +1,1166 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 54.3016,
3
+ "best_model_checkpoint": "/uoa/home/s02sd1/Desktop/Project data/pytorch_project/facebook_28.03/checkpoint-43000",
4
+ "epoch": 64.88011283497885,
5
+ "global_step": 46000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.71,
12
+ "learning_rate": 2.978843441466855e-06,
13
+ "loss": 0.6224,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 1.41,
18
+ "learning_rate": 2.9576868829337095e-06,
19
+ "loss": 0.5049,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 1.41,
24
+ "eval_gen_len": 167.2707,
25
+ "eval_loss": 0.75015789270401,
26
+ "eval_rouge1": 61.2286,
27
+ "eval_rouge2": 43.2506,
28
+ "eval_rougeL": 42.2051,
29
+ "eval_rougeLsum": 46.3795,
30
+ "eval_runtime": 5804.0101,
31
+ "eval_samples_per_second": 0.06,
32
+ "eval_steps_per_second": 0.003,
33
+ "step": 1000
34
+ },
35
+ {
36
+ "epoch": 2.12,
37
+ "learning_rate": 2.9365303244005645e-06,
38
+ "loss": 0.4992,
39
+ "step": 1500
40
+ },
41
+ {
42
+ "epoch": 2.82,
43
+ "learning_rate": 2.915373765867419e-06,
44
+ "loss": 0.4523,
45
+ "step": 2000
46
+ },
47
+ {
48
+ "epoch": 2.82,
49
+ "eval_gen_len": 169.4872,
50
+ "eval_loss": 0.7313751578330994,
51
+ "eval_rouge1": 63.5466,
52
+ "eval_rouge2": 45.7831,
53
+ "eval_rougeL": 44.4885,
54
+ "eval_rougeLsum": 48.6313,
55
+ "eval_runtime": 4918.4156,
56
+ "eval_samples_per_second": 0.071,
57
+ "eval_steps_per_second": 0.004,
58
+ "step": 2000
59
+ },
60
+ {
61
+ "epoch": 3.53,
62
+ "learning_rate": 2.894217207334274e-06,
63
+ "loss": 0.4422,
64
+ "step": 2500
65
+ },
66
+ {
67
+ "epoch": 4.23,
68
+ "learning_rate": 2.8730606488011285e-06,
69
+ "loss": 0.4036,
70
+ "step": 3000
71
+ },
72
+ {
73
+ "epoch": 4.23,
74
+ "eval_gen_len": 160.4074,
75
+ "eval_loss": 0.7353909611701965,
76
+ "eval_rouge1": 64.7439,
77
+ "eval_rouge2": 47.1723,
78
+ "eval_rougeL": 45.9479,
79
+ "eval_rougeLsum": 50.2341,
80
+ "eval_runtime": 4241.5946,
81
+ "eval_samples_per_second": 0.083,
82
+ "eval_steps_per_second": 0.004,
83
+ "step": 3000
84
+ },
85
+ {
86
+ "epoch": 4.94,
87
+ "learning_rate": 2.851904090267983e-06,
88
+ "loss": 0.4028,
89
+ "step": 3500
90
+ },
91
+ {
92
+ "epoch": 5.64,
93
+ "learning_rate": 2.830747531734838e-06,
94
+ "loss": 0.3722,
95
+ "step": 4000
96
+ },
97
+ {
98
+ "epoch": 5.64,
99
+ "eval_gen_len": 162.0085,
100
+ "eval_loss": 0.72761070728302,
101
+ "eval_rouge1": 65.6958,
102
+ "eval_rouge2": 48.5007,
103
+ "eval_rougeL": 46.6737,
104
+ "eval_rougeLsum": 50.9835,
105
+ "eval_runtime": 4232.4341,
106
+ "eval_samples_per_second": 0.083,
107
+ "eval_steps_per_second": 0.004,
108
+ "step": 4000
109
+ },
110
+ {
111
+ "epoch": 6.35,
112
+ "learning_rate": 2.8095909732016925e-06,
113
+ "loss": 0.3749,
114
+ "step": 4500
115
+ },
116
+ {
117
+ "epoch": 7.05,
118
+ "learning_rate": 2.7884344146685474e-06,
119
+ "loss": 0.3482,
120
+ "step": 5000
121
+ },
122
+ {
123
+ "epoch": 7.05,
124
+ "eval_gen_len": 163.9231,
125
+ "eval_loss": 0.7417466640472412,
126
+ "eval_rouge1": 65.933,
127
+ "eval_rouge2": 48.921,
128
+ "eval_rougeL": 47.5877,
129
+ "eval_rougeLsum": 52.0114,
130
+ "eval_runtime": 4836.5569,
131
+ "eval_samples_per_second": 0.073,
132
+ "eval_steps_per_second": 0.004,
133
+ "step": 5000
134
+ },
135
+ {
136
+ "epoch": 7.76,
137
+ "learning_rate": 2.767277856135402e-06,
138
+ "loss": 0.3312,
139
+ "step": 5500
140
+ },
141
+ {
142
+ "epoch": 8.46,
143
+ "learning_rate": 2.746121297602257e-06,
144
+ "loss": 0.3174,
145
+ "step": 6000
146
+ },
147
+ {
148
+ "epoch": 8.46,
149
+ "eval_gen_len": 163.5527,
150
+ "eval_loss": 0.7546955943107605,
151
+ "eval_rouge1": 65.9445,
152
+ "eval_rouge2": 48.9117,
153
+ "eval_rougeL": 48.312,
154
+ "eval_rougeLsum": 52.4923,
155
+ "eval_runtime": 5417.8435,
156
+ "eval_samples_per_second": 0.065,
157
+ "eval_steps_per_second": 0.003,
158
+ "step": 6000
159
+ },
160
+ {
161
+ "epoch": 9.17,
162
+ "learning_rate": 2.7249647390691114e-06,
163
+ "loss": 0.3074,
164
+ "step": 6500
165
+ },
166
+ {
167
+ "epoch": 9.87,
168
+ "learning_rate": 2.7038081805359663e-06,
169
+ "loss": 0.2922,
170
+ "step": 7000
171
+ },
172
+ {
173
+ "epoch": 9.87,
174
+ "eval_gen_len": 163.1311,
175
+ "eval_loss": 0.754240095615387,
176
+ "eval_rouge1": 66.6619,
177
+ "eval_rouge2": 49.9256,
178
+ "eval_rougeL": 48.7757,
179
+ "eval_rougeLsum": 53.1916,
180
+ "eval_runtime": 5340.7776,
181
+ "eval_samples_per_second": 0.066,
182
+ "eval_steps_per_second": 0.003,
183
+ "step": 7000
184
+ },
185
+ {
186
+ "epoch": 10.58,
187
+ "learning_rate": 2.6826516220028213e-06,
188
+ "loss": 0.281,
189
+ "step": 7500
190
+ },
191
+ {
192
+ "epoch": 11.28,
193
+ "learning_rate": 2.6614950634696754e-06,
194
+ "loss": 0.276,
195
+ "step": 8000
196
+ },
197
+ {
198
+ "epoch": 11.28,
199
+ "eval_gen_len": 165.0826,
200
+ "eval_loss": 0.7793148756027222,
201
+ "eval_rouge1": 66.8911,
202
+ "eval_rouge2": 50.331,
203
+ "eval_rougeL": 49.8396,
204
+ "eval_rougeLsum": 54.0617,
205
+ "eval_runtime": 6037.4348,
206
+ "eval_samples_per_second": 0.058,
207
+ "eval_steps_per_second": 0.003,
208
+ "step": 8000
209
+ },
210
+ {
211
+ "epoch": 11.99,
212
+ "learning_rate": 2.6403385049365303e-06,
213
+ "loss": 0.2557,
214
+ "step": 8500
215
+ },
216
+ {
217
+ "epoch": 12.69,
218
+ "learning_rate": 2.6191819464033853e-06,
219
+ "loss": 0.2464,
220
+ "step": 9000
221
+ },
222
+ {
223
+ "epoch": 12.69,
224
+ "eval_gen_len": 164.9943,
225
+ "eval_loss": 0.7877324819564819,
226
+ "eval_rouge1": 67.1466,
227
+ "eval_rouge2": 50.939,
228
+ "eval_rougeL": 50.2255,
229
+ "eval_rougeLsum": 54.5145,
230
+ "eval_runtime": 5837.726,
231
+ "eval_samples_per_second": 0.06,
232
+ "eval_steps_per_second": 0.003,
233
+ "step": 9000
234
+ },
235
+ {
236
+ "epoch": 13.4,
237
+ "learning_rate": 2.5980253878702398e-06,
238
+ "loss": 0.2369,
239
+ "step": 9500
240
+ },
241
+ {
242
+ "epoch": 14.1,
243
+ "learning_rate": 2.5768688293370947e-06,
244
+ "loss": 0.2265,
245
+ "step": 10000
246
+ },
247
+ {
248
+ "epoch": 14.1,
249
+ "eval_gen_len": 162.1966,
250
+ "eval_loss": 0.806422233581543,
251
+ "eval_rouge1": 67.5137,
252
+ "eval_rouge2": 51.2801,
253
+ "eval_rougeL": 51.2323,
254
+ "eval_rougeLsum": 55.426,
255
+ "eval_runtime": 5856.0164,
256
+ "eval_samples_per_second": 0.06,
257
+ "eval_steps_per_second": 0.003,
258
+ "step": 10000
259
+ },
260
+ {
261
+ "epoch": 14.81,
262
+ "learning_rate": 2.5557122708039492e-06,
263
+ "loss": 0.2215,
264
+ "step": 10500
265
+ },
266
+ {
267
+ "epoch": 15.51,
268
+ "learning_rate": 2.534555712270804e-06,
269
+ "loss": 0.2004,
270
+ "step": 11000
271
+ },
272
+ {
273
+ "epoch": 15.51,
274
+ "eval_gen_len": 161.2365,
275
+ "eval_loss": 0.8361101746559143,
276
+ "eval_rouge1": 67.9515,
277
+ "eval_rouge2": 51.5857,
278
+ "eval_rougeL": 51.5796,
279
+ "eval_rougeLsum": 55.9738,
280
+ "eval_runtime": 5461.4431,
281
+ "eval_samples_per_second": 0.064,
282
+ "eval_steps_per_second": 0.003,
283
+ "step": 11000
284
+ },
285
+ {
286
+ "epoch": 16.22,
287
+ "learning_rate": 2.5133991537376587e-06,
288
+ "loss": 0.2039,
289
+ "step": 11500
290
+ },
291
+ {
292
+ "epoch": 16.93,
293
+ "learning_rate": 2.4922425952045136e-06,
294
+ "loss": 0.1903,
295
+ "step": 12000
296
+ },
297
+ {
298
+ "epoch": 16.93,
299
+ "eval_gen_len": 163.8234,
300
+ "eval_loss": 0.8543212413787842,
301
+ "eval_rouge1": 67.8403,
302
+ "eval_rouge2": 51.7839,
303
+ "eval_rougeL": 51.8869,
304
+ "eval_rougeLsum": 56.2681,
305
+ "eval_runtime": 5935.4542,
306
+ "eval_samples_per_second": 0.059,
307
+ "eval_steps_per_second": 0.003,
308
+ "step": 12000
309
+ },
310
+ {
311
+ "epoch": 17.63,
312
+ "learning_rate": 2.471086036671368e-06,
313
+ "loss": 0.1816,
314
+ "step": 12500
315
+ },
316
+ {
317
+ "epoch": 18.34,
318
+ "learning_rate": 2.4499294781382227e-06,
319
+ "loss": 0.1782,
320
+ "step": 13000
321
+ },
322
+ {
323
+ "epoch": 18.34,
324
+ "eval_gen_len": 161.2137,
325
+ "eval_loss": 0.8672023415565491,
326
+ "eval_rouge1": 68.2574,
327
+ "eval_rouge2": 52.2649,
328
+ "eval_rougeL": 52.3755,
329
+ "eval_rougeLsum": 56.4509,
330
+ "eval_runtime": 5698.6716,
331
+ "eval_samples_per_second": 0.062,
332
+ "eval_steps_per_second": 0.003,
333
+ "step": 13000
334
+ },
335
+ {
336
+ "epoch": 19.04,
337
+ "learning_rate": 2.4287729196050776e-06,
338
+ "loss": 0.1686,
339
+ "step": 13500
340
+ },
341
+ {
342
+ "epoch": 19.75,
343
+ "learning_rate": 2.407616361071932e-06,
344
+ "loss": 0.1581,
345
+ "step": 14000
346
+ },
347
+ {
348
+ "epoch": 19.75,
349
+ "eval_gen_len": 163.0057,
350
+ "eval_loss": 0.9022919535636902,
351
+ "eval_rouge1": 67.8924,
352
+ "eval_rouge2": 51.8802,
353
+ "eval_rougeL": 52.5074,
354
+ "eval_rougeLsum": 56.5946,
355
+ "eval_runtime": 5838.2781,
356
+ "eval_samples_per_second": 0.06,
357
+ "eval_steps_per_second": 0.003,
358
+ "step": 14000
359
+ },
360
+ {
361
+ "epoch": 20.45,
362
+ "learning_rate": 2.386459802538787e-06,
363
+ "loss": 0.1525,
364
+ "step": 14500
365
+ },
366
+ {
367
+ "epoch": 21.16,
368
+ "learning_rate": 2.3653032440056416e-06,
369
+ "loss": 0.1516,
370
+ "step": 15000
371
+ },
372
+ {
373
+ "epoch": 21.16,
374
+ "eval_gen_len": 163.0883,
375
+ "eval_loss": 0.9338411092758179,
376
+ "eval_rouge1": 68.0101,
377
+ "eval_rouge2": 51.7535,
378
+ "eval_rougeL": 52.2777,
379
+ "eval_rougeLsum": 56.692,
380
+ "eval_runtime": 6063.036,
381
+ "eval_samples_per_second": 0.058,
382
+ "eval_steps_per_second": 0.003,
383
+ "step": 15000
384
+ },
385
+ {
386
+ "epoch": 21.86,
387
+ "learning_rate": 2.3441466854724966e-06,
388
+ "loss": 0.143,
389
+ "step": 15500
390
+ },
391
+ {
392
+ "epoch": 22.57,
393
+ "learning_rate": 2.3229901269393515e-06,
394
+ "loss": 0.1358,
395
+ "step": 16000
396
+ },
397
+ {
398
+ "epoch": 22.57,
399
+ "eval_gen_len": 162.3504,
400
+ "eval_loss": 0.9403561353683472,
401
+ "eval_rouge1": 68.4997,
402
+ "eval_rouge2": 52.385,
403
+ "eval_rougeL": 52.7137,
404
+ "eval_rougeLsum": 57.2165,
405
+ "eval_runtime": 5678.9898,
406
+ "eval_samples_per_second": 0.062,
407
+ "eval_steps_per_second": 0.003,
408
+ "step": 16000
409
+ },
410
+ {
411
+ "epoch": 23.27,
412
+ "learning_rate": 2.301833568406206e-06,
413
+ "loss": 0.1268,
414
+ "step": 16500
415
+ },
416
+ {
417
+ "epoch": 23.98,
418
+ "learning_rate": 2.280677009873061e-06,
419
+ "loss": 0.1256,
420
+ "step": 17000
421
+ },
422
+ {
423
+ "epoch": 23.98,
424
+ "eval_gen_len": 163.6467,
425
+ "eval_loss": 0.9707176685333252,
426
+ "eval_rouge1": 68.3426,
427
+ "eval_rouge2": 52.1326,
428
+ "eval_rougeL": 52.6279,
429
+ "eval_rougeLsum": 57.0803,
430
+ "eval_runtime": 6169.909,
431
+ "eval_samples_per_second": 0.057,
432
+ "eval_steps_per_second": 0.003,
433
+ "step": 17000
434
+ },
435
+ {
436
+ "epoch": 24.68,
437
+ "learning_rate": 2.2595204513399155e-06,
438
+ "loss": 0.1158,
439
+ "step": 17500
440
+ },
441
+ {
442
+ "epoch": 25.39,
443
+ "learning_rate": 2.23836389280677e-06,
444
+ "loss": 0.1119,
445
+ "step": 18000
446
+ },
447
+ {
448
+ "epoch": 25.39,
449
+ "eval_gen_len": 160.9316,
450
+ "eval_loss": 1.0022860765457153,
451
+ "eval_rouge1": 69.1266,
452
+ "eval_rouge2": 52.9311,
453
+ "eval_rougeL": 53.9465,
454
+ "eval_rougeLsum": 58.0804,
455
+ "eval_runtime": 5772.1926,
456
+ "eval_samples_per_second": 0.061,
457
+ "eval_steps_per_second": 0.003,
458
+ "step": 18000
459
+ },
460
+ {
461
+ "epoch": 26.09,
462
+ "learning_rate": 2.217207334273625e-06,
463
+ "loss": 0.1125,
464
+ "step": 18500
465
+ },
466
+ {
467
+ "epoch": 26.8,
468
+ "learning_rate": 2.1960507757404795e-06,
469
+ "loss": 0.1038,
470
+ "step": 19000
471
+ },
472
+ {
473
+ "epoch": 26.8,
474
+ "eval_gen_len": 161.8746,
475
+ "eval_loss": 1.019147515296936,
476
+ "eval_rouge1": 68.6629,
477
+ "eval_rouge2": 52.6518,
478
+ "eval_rougeL": 53.1635,
479
+ "eval_rougeLsum": 57.4514,
480
+ "eval_runtime": 5939.4826,
481
+ "eval_samples_per_second": 0.059,
482
+ "eval_steps_per_second": 0.003,
483
+ "step": 19000
484
+ },
485
+ {
486
+ "epoch": 27.5,
487
+ "learning_rate": 2.1748942172073344e-06,
488
+ "loss": 0.098,
489
+ "step": 19500
490
+ },
491
+ {
492
+ "epoch": 28.21,
493
+ "learning_rate": 2.153737658674189e-06,
494
+ "loss": 0.0944,
495
+ "step": 20000
496
+ },
497
+ {
498
+ "epoch": 28.21,
499
+ "eval_gen_len": 161.4359,
500
+ "eval_loss": 1.0427296161651611,
501
+ "eval_rouge1": 69.1063,
502
+ "eval_rouge2": 53.0226,
503
+ "eval_rougeL": 53.5979,
504
+ "eval_rougeLsum": 57.8602,
505
+ "eval_runtime": 6011.7516,
506
+ "eval_samples_per_second": 0.058,
507
+ "eval_steps_per_second": 0.003,
508
+ "step": 20000
509
+ },
510
+ {
511
+ "epoch": 28.91,
512
+ "learning_rate": 2.132581100141044e-06,
513
+ "loss": 0.0929,
514
+ "step": 20500
515
+ },
516
+ {
517
+ "epoch": 29.62,
518
+ "learning_rate": 2.1114245416078984e-06,
519
+ "loss": 0.0843,
520
+ "step": 21000
521
+ },
522
+ {
523
+ "epoch": 29.62,
524
+ "eval_gen_len": 161.849,
525
+ "eval_loss": 1.0581331253051758,
526
+ "eval_rouge1": 68.0982,
527
+ "eval_rouge2": 52.0153,
528
+ "eval_rougeL": 53.1935,
529
+ "eval_rougeLsum": 57.443,
530
+ "eval_runtime": 5905.725,
531
+ "eval_samples_per_second": 0.059,
532
+ "eval_steps_per_second": 0.003,
533
+ "step": 21000
534
+ },
535
+ {
536
+ "epoch": 30.32,
537
+ "learning_rate": 2.0902679830747533e-06,
538
+ "loss": 0.0846,
539
+ "step": 21500
540
+ },
541
+ {
542
+ "epoch": 31.03,
543
+ "learning_rate": 2.069111424541608e-06,
544
+ "loss": 0.082,
545
+ "step": 22000
546
+ },
547
+ {
548
+ "epoch": 31.03,
549
+ "eval_gen_len": 161.0684,
550
+ "eval_loss": 1.0757286548614502,
551
+ "eval_rouge1": 68.6318,
552
+ "eval_rouge2": 52.8093,
553
+ "eval_rougeL": 53.7214,
554
+ "eval_rougeLsum": 57.775,
555
+ "eval_runtime": 6004.4709,
556
+ "eval_samples_per_second": 0.058,
557
+ "eval_steps_per_second": 0.003,
558
+ "step": 22000
559
+ },
560
+ {
561
+ "epoch": 31.73,
562
+ "learning_rate": 2.047954866008463e-06,
563
+ "loss": 0.0735,
564
+ "step": 22500
565
+ },
566
+ {
567
+ "epoch": 32.44,
568
+ "learning_rate": 2.0267983074753173e-06,
569
+ "loss": 0.0756,
570
+ "step": 23000
571
+ },
572
+ {
573
+ "epoch": 32.44,
574
+ "eval_gen_len": 162.4245,
575
+ "eval_loss": 1.0798823833465576,
576
+ "eval_rouge1": 68.557,
577
+ "eval_rouge2": 52.4644,
578
+ "eval_rougeL": 53.7606,
579
+ "eval_rougeLsum": 57.8967,
580
+ "eval_runtime": 6130.0883,
581
+ "eval_samples_per_second": 0.057,
582
+ "eval_steps_per_second": 0.003,
583
+ "step": 23000
584
+ },
585
+ {
586
+ "epoch": 33.15,
587
+ "learning_rate": 2.005641748942172e-06,
588
+ "loss": 0.0706,
589
+ "step": 23500
590
+ },
591
+ {
592
+ "epoch": 33.85,
593
+ "learning_rate": 1.9844851904090268e-06,
594
+ "loss": 0.0674,
595
+ "step": 24000
596
+ },
597
+ {
598
+ "epoch": 33.85,
599
+ "eval_gen_len": 162.963,
600
+ "eval_loss": 1.1047182083129883,
601
+ "eval_rouge1": 68.6765,
602
+ "eval_rouge2": 52.713,
603
+ "eval_rougeL": 53.5339,
604
+ "eval_rougeLsum": 57.7006,
605
+ "eval_runtime": 6102.6451,
606
+ "eval_samples_per_second": 0.058,
607
+ "eval_steps_per_second": 0.003,
608
+ "step": 24000
609
+ },
610
+ {
611
+ "epoch": 34.56,
612
+ "learning_rate": 1.9633286318758817e-06,
613
+ "loss": 0.0647,
614
+ "step": 24500
615
+ },
616
+ {
617
+ "epoch": 35.26,
618
+ "learning_rate": 1.9421720733427362e-06,
619
+ "loss": 0.063,
620
+ "step": 25000
621
+ },
622
+ {
623
+ "epoch": 35.26,
624
+ "eval_gen_len": 161.9003,
625
+ "eval_loss": 1.1172865629196167,
626
+ "eval_rouge1": 69.1911,
627
+ "eval_rouge2": 53.1382,
628
+ "eval_rougeL": 53.8971,
629
+ "eval_rougeLsum": 58.2079,
630
+ "eval_runtime": 5980.899,
631
+ "eval_samples_per_second": 0.059,
632
+ "eval_steps_per_second": 0.003,
633
+ "step": 25000
634
+ },
635
+ {
636
+ "epoch": 35.97,
637
+ "learning_rate": 1.921015514809591e-06,
638
+ "loss": 0.0599,
639
+ "step": 25500
640
+ },
641
+ {
642
+ "epoch": 36.67,
643
+ "learning_rate": 1.8998589562764457e-06,
644
+ "loss": 0.0566,
645
+ "step": 26000
646
+ },
647
+ {
648
+ "epoch": 36.67,
649
+ "eval_gen_len": 162.3362,
650
+ "eval_loss": 1.131187081336975,
651
+ "eval_rouge1": 69.0624,
652
+ "eval_rouge2": 53.101,
653
+ "eval_rougeL": 54.2012,
654
+ "eval_rougeLsum": 58.5665,
655
+ "eval_runtime": 5792.9256,
656
+ "eval_samples_per_second": 0.061,
657
+ "eval_steps_per_second": 0.003,
658
+ "step": 26000
659
+ },
660
+ {
661
+ "epoch": 37.38,
662
+ "learning_rate": 1.8787023977433004e-06,
663
+ "loss": 0.0557,
664
+ "step": 26500
665
+ },
666
+ {
667
+ "epoch": 38.08,
668
+ "learning_rate": 1.8575458392101554e-06,
669
+ "loss": 0.0532,
670
+ "step": 27000
671
+ },
672
+ {
673
+ "epoch": 38.08,
674
+ "eval_gen_len": 161.7009,
675
+ "eval_loss": 1.1492334604263306,
676
+ "eval_rouge1": 68.8232,
677
+ "eval_rouge2": 52.7547,
678
+ "eval_rougeL": 53.914,
679
+ "eval_rougeLsum": 58.1061,
680
+ "eval_runtime": 6017.4685,
681
+ "eval_samples_per_second": 0.058,
682
+ "eval_steps_per_second": 0.003,
683
+ "step": 27000
684
+ },
685
+ {
686
+ "epoch": 38.79,
687
+ "learning_rate": 1.8363892806770101e-06,
688
+ "loss": 0.0518,
689
+ "step": 27500
690
+ },
691
+ {
692
+ "epoch": 39.49,
693
+ "learning_rate": 1.8152327221438644e-06,
694
+ "loss": 0.0482,
695
+ "step": 28000
696
+ },
697
+ {
698
+ "epoch": 39.49,
699
+ "eval_gen_len": 161.4046,
700
+ "eval_loss": 1.1425426006317139,
701
+ "eval_rouge1": 68.8641,
702
+ "eval_rouge2": 52.7579,
703
+ "eval_rougeL": 54.0604,
704
+ "eval_rougeLsum": 58.3338,
705
+ "eval_runtime": 5870.7466,
706
+ "eval_samples_per_second": 0.06,
707
+ "eval_steps_per_second": 0.003,
708
+ "step": 28000
709
+ },
710
+ {
711
+ "epoch": 40.2,
712
+ "learning_rate": 1.7940761636107192e-06,
713
+ "loss": 0.0459,
714
+ "step": 28500
715
+ },
716
+ {
717
+ "epoch": 40.9,
718
+ "learning_rate": 1.772919605077574e-06,
719
+ "loss": 0.0458,
720
+ "step": 29000
721
+ },
722
+ {
723
+ "epoch": 40.9,
724
+ "eval_gen_len": 156.8718,
725
+ "eval_loss": 1.1709474325180054,
726
+ "eval_rouge1": 69.9022,
727
+ "eval_rouge2": 53.8817,
728
+ "eval_rougeL": 55.3029,
729
+ "eval_rougeLsum": 59.6658,
730
+ "eval_runtime": 5805.0163,
731
+ "eval_samples_per_second": 0.06,
732
+ "eval_steps_per_second": 0.003,
733
+ "step": 29000
734
+ },
735
+ {
736
+ "epoch": 41.61,
737
+ "learning_rate": 1.7517630465444288e-06,
738
+ "loss": 0.0417,
739
+ "step": 29500
740
+ },
741
+ {
742
+ "epoch": 42.31,
743
+ "learning_rate": 1.7306064880112836e-06,
744
+ "loss": 0.0421,
745
+ "step": 30000
746
+ },
747
+ {
748
+ "epoch": 42.31,
749
+ "eval_gen_len": 161.1595,
750
+ "eval_loss": 1.1962451934814453,
751
+ "eval_rouge1": 69.8385,
752
+ "eval_rouge2": 53.9521,
753
+ "eval_rougeL": 55.043,
754
+ "eval_rougeLsum": 59.4319,
755
+ "eval_runtime": 5621.1651,
756
+ "eval_samples_per_second": 0.062,
757
+ "eval_steps_per_second": 0.003,
758
+ "step": 30000
759
+ },
760
+ {
761
+ "epoch": 43.02,
762
+ "learning_rate": 1.7094499294781383e-06,
763
+ "loss": 0.0415,
764
+ "step": 30500
765
+ },
766
+ {
767
+ "epoch": 43.72,
768
+ "learning_rate": 1.688293370944993e-06,
769
+ "loss": 0.0378,
770
+ "step": 31000
771
+ },
772
+ {
773
+ "epoch": 43.72,
774
+ "eval_gen_len": 161.1368,
775
+ "eval_loss": 1.1894584894180298,
776
+ "eval_rouge1": 69.2784,
777
+ "eval_rouge2": 53.3896,
778
+ "eval_rougeL": 54.7309,
779
+ "eval_rougeLsum": 58.9909,
780
+ "eval_runtime": 5848.8931,
781
+ "eval_samples_per_second": 0.06,
782
+ "eval_steps_per_second": 0.003,
783
+ "step": 31000
784
+ },
785
+ {
786
+ "epoch": 44.43,
787
+ "learning_rate": 1.6671368124118478e-06,
788
+ "loss": 0.0382,
789
+ "step": 31500
790
+ },
791
+ {
792
+ "epoch": 45.13,
793
+ "learning_rate": 1.6459802538787025e-06,
794
+ "loss": 0.0351,
795
+ "step": 32000
796
+ },
797
+ {
798
+ "epoch": 45.13,
799
+ "eval_gen_len": 161.0627,
800
+ "eval_loss": 1.2170298099517822,
801
+ "eval_rouge1": 69.2182,
802
+ "eval_rouge2": 53.2021,
803
+ "eval_rougeL": 54.8192,
804
+ "eval_rougeLsum": 58.858,
805
+ "eval_runtime": 5747.8165,
806
+ "eval_samples_per_second": 0.061,
807
+ "eval_steps_per_second": 0.003,
808
+ "step": 32000
809
+ },
810
+ {
811
+ "epoch": 45.84,
812
+ "learning_rate": 1.6248236953455572e-06,
813
+ "loss": 0.0348,
814
+ "step": 32500
815
+ },
816
+ {
817
+ "epoch": 46.54,
818
+ "learning_rate": 1.6036671368124117e-06,
819
+ "loss": 0.0328,
820
+ "step": 33000
821
+ },
822
+ {
823
+ "epoch": 46.54,
824
+ "eval_gen_len": 162.3647,
825
+ "eval_loss": 1.2242525815963745,
826
+ "eval_rouge1": 69.2151,
827
+ "eval_rouge2": 53.5006,
828
+ "eval_rougeL": 54.6811,
829
+ "eval_rougeLsum": 58.899,
830
+ "eval_runtime": 5815.6916,
831
+ "eval_samples_per_second": 0.06,
832
+ "eval_steps_per_second": 0.003,
833
+ "step": 33000
834
+ },
835
+ {
836
+ "epoch": 47.25,
837
+ "learning_rate": 1.5825105782792665e-06,
838
+ "loss": 0.0334,
839
+ "step": 33500
840
+ },
841
+ {
842
+ "epoch": 47.95,
843
+ "learning_rate": 1.5613540197461212e-06,
844
+ "loss": 0.0316,
845
+ "step": 34000
846
+ },
847
+ {
848
+ "epoch": 47.95,
849
+ "eval_gen_len": 163.4872,
850
+ "eval_loss": 1.2345460653305054,
851
+ "eval_rouge1": 68.9079,
852
+ "eval_rouge2": 52.7923,
853
+ "eval_rougeL": 54.7024,
854
+ "eval_rougeLsum": 59.1333,
855
+ "eval_runtime": 5738.4019,
856
+ "eval_samples_per_second": 0.061,
857
+ "eval_steps_per_second": 0.003,
858
+ "step": 34000
859
+ },
860
+ {
861
+ "epoch": 48.66,
862
+ "learning_rate": 1.540197461212976e-06,
863
+ "loss": 0.0303,
864
+ "step": 34500
865
+ },
866
+ {
867
+ "epoch": 49.37,
868
+ "learning_rate": 1.5190409026798307e-06,
869
+ "loss": 0.0289,
870
+ "step": 35000
871
+ },
872
+ {
873
+ "epoch": 49.37,
874
+ "eval_gen_len": 161.2165,
875
+ "eval_loss": 1.2448240518569946,
876
+ "eval_rouge1": 69.3435,
877
+ "eval_rouge2": 53.5007,
878
+ "eval_rougeL": 54.6771,
879
+ "eval_rougeLsum": 59.067,
880
+ "eval_runtime": 5523.8322,
881
+ "eval_samples_per_second": 0.064,
882
+ "eval_steps_per_second": 0.003,
883
+ "step": 35000
884
+ },
885
+ {
886
+ "epoch": 50.07,
887
+ "learning_rate": 1.4978843441466856e-06,
888
+ "loss": 0.0287,
889
+ "step": 35500
890
+ },
891
+ {
892
+ "epoch": 50.78,
893
+ "learning_rate": 1.4767277856135403e-06,
894
+ "loss": 0.0279,
895
+ "step": 36000
896
+ },
897
+ {
898
+ "epoch": 50.78,
899
+ "eval_gen_len": 163.7664,
900
+ "eval_loss": 1.254445195198059,
901
+ "eval_rouge1": 68.8472,
902
+ "eval_rouge2": 53.1054,
903
+ "eval_rougeL": 54.3767,
904
+ "eval_rougeLsum": 58.4479,
905
+ "eval_runtime": 5626.7741,
906
+ "eval_samples_per_second": 0.062,
907
+ "eval_steps_per_second": 0.003,
908
+ "step": 36000
909
+ },
910
+ {
911
+ "epoch": 51.48,
912
+ "learning_rate": 1.455571227080395e-06,
913
+ "loss": 0.0266,
914
+ "step": 36500
915
+ },
916
+ {
917
+ "epoch": 52.19,
918
+ "learning_rate": 1.4344146685472496e-06,
919
+ "loss": 0.0266,
920
+ "step": 37000
921
+ },
922
+ {
923
+ "epoch": 52.19,
924
+ "eval_gen_len": 163.3675,
925
+ "eval_loss": 1.267912745475769,
926
+ "eval_rouge1": 69.1523,
927
+ "eval_rouge2": 53.3283,
928
+ "eval_rougeL": 54.8059,
929
+ "eval_rougeLsum": 58.9981,
930
+ "eval_runtime": 5635.0806,
931
+ "eval_samples_per_second": 0.062,
932
+ "eval_steps_per_second": 0.003,
933
+ "step": 37000
934
+ },
935
+ {
936
+ "epoch": 52.89,
937
+ "learning_rate": 1.4132581100141043e-06,
938
+ "loss": 0.025,
939
+ "step": 37500
940
+ },
941
+ {
942
+ "epoch": 53.6,
943
+ "learning_rate": 1.392101551480959e-06,
944
+ "loss": 0.0237,
945
+ "step": 38000
946
+ },
947
+ {
948
+ "epoch": 53.6,
949
+ "eval_gen_len": 161.943,
950
+ "eval_loss": 1.278545618057251,
951
+ "eval_rouge1": 69.7623,
952
+ "eval_rouge2": 53.9089,
953
+ "eval_rougeL": 55.4188,
954
+ "eval_rougeLsum": 59.7174,
955
+ "eval_runtime": 5517.6551,
956
+ "eval_samples_per_second": 0.064,
957
+ "eval_steps_per_second": 0.003,
958
+ "step": 38000
959
+ },
960
+ {
961
+ "epoch": 54.3,
962
+ "learning_rate": 1.3709449929478138e-06,
963
+ "loss": 0.0243,
964
+ "step": 38500
965
+ },
966
+ {
967
+ "epoch": 55.01,
968
+ "learning_rate": 1.3497884344146687e-06,
969
+ "loss": 0.0225,
970
+ "step": 39000
971
+ },
972
+ {
973
+ "epoch": 55.01,
974
+ "eval_gen_len": 162.8234,
975
+ "eval_loss": 1.2910034656524658,
976
+ "eval_rouge1": 69.4568,
977
+ "eval_rouge2": 53.5993,
978
+ "eval_rougeL": 55.3547,
979
+ "eval_rougeLsum": 59.7055,
980
+ "eval_runtime": 5493.2949,
981
+ "eval_samples_per_second": 0.064,
982
+ "eval_steps_per_second": 0.003,
983
+ "step": 39000
984
+ },
985
+ {
986
+ "epoch": 55.71,
987
+ "learning_rate": 1.3286318758815233e-06,
988
+ "loss": 0.0222,
989
+ "step": 39500
990
+ },
991
+ {
992
+ "epoch": 56.42,
993
+ "learning_rate": 1.307475317348378e-06,
994
+ "loss": 0.0207,
995
+ "step": 40000
996
+ },
997
+ {
998
+ "epoch": 56.42,
999
+ "eval_gen_len": 161.9145,
1000
+ "eval_loss": 1.302707314491272,
1001
+ "eval_rouge1": 69.3521,
1002
+ "eval_rouge2": 53.657,
1003
+ "eval_rougeL": 55.4158,
1004
+ "eval_rougeLsum": 59.7595,
1005
+ "eval_runtime": 5493.618,
1006
+ "eval_samples_per_second": 0.064,
1007
+ "eval_steps_per_second": 0.003,
1008
+ "step": 40000
1009
+ },
1010
+ {
1011
+ "epoch": 57.12,
1012
+ "learning_rate": 1.2863187588152327e-06,
1013
+ "loss": 0.0216,
1014
+ "step": 40500
1015
+ },
1016
+ {
1017
+ "epoch": 57.83,
1018
+ "learning_rate": 1.2651622002820875e-06,
1019
+ "loss": 0.0207,
1020
+ "step": 41000
1021
+ },
1022
+ {
1023
+ "epoch": 57.83,
1024
+ "eval_gen_len": 161.4017,
1025
+ "eval_loss": 1.2961533069610596,
1026
+ "eval_rouge1": 69.5378,
1027
+ "eval_rouge2": 53.6568,
1028
+ "eval_rougeL": 55.4833,
1029
+ "eval_rougeLsum": 59.8609,
1030
+ "eval_runtime": 5729.7323,
1031
+ "eval_samples_per_second": 0.061,
1032
+ "eval_steps_per_second": 0.003,
1033
+ "step": 41000
1034
+ },
1035
+ {
1036
+ "epoch": 58.53,
1037
+ "learning_rate": 1.2440056417489422e-06,
1038
+ "loss": 0.0202,
1039
+ "step": 41500
1040
+ },
1041
+ {
1042
+ "epoch": 59.24,
1043
+ "learning_rate": 1.222849083215797e-06,
1044
+ "loss": 0.019,
1045
+ "step": 42000
1046
+ },
1047
+ {
1048
+ "epoch": 59.24,
1049
+ "eval_gen_len": 161.5954,
1050
+ "eval_loss": 1.3159691095352173,
1051
+ "eval_rouge1": 69.6692,
1052
+ "eval_rouge2": 53.8824,
1053
+ "eval_rougeL": 56.153,
1054
+ "eval_rougeLsum": 60.5724,
1055
+ "eval_runtime": 5770.1856,
1056
+ "eval_samples_per_second": 0.061,
1057
+ "eval_steps_per_second": 0.003,
1058
+ "step": 42000
1059
+ },
1060
+ {
1061
+ "epoch": 59.94,
1062
+ "learning_rate": 1.2016925246826517e-06,
1063
+ "loss": 0.0193,
1064
+ "step": 42500
1065
+ },
1066
+ {
1067
+ "epoch": 60.65,
1068
+ "learning_rate": 1.1805359661495064e-06,
1069
+ "loss": 0.0183,
1070
+ "step": 43000
1071
+ },
1072
+ {
1073
+ "epoch": 60.65,
1074
+ "eval_gen_len": 159.6097,
1075
+ "eval_loss": 1.3009204864501953,
1076
+ "eval_rouge1": 69.8056,
1077
+ "eval_rouge2": 54.3016,
1078
+ "eval_rougeL": 56.2054,
1079
+ "eval_rougeLsum": 60.4544,
1080
+ "eval_runtime": 5689.5789,
1081
+ "eval_samples_per_second": 0.062,
1082
+ "eval_steps_per_second": 0.003,
1083
+ "step": 43000
1084
+ },
1085
+ {
1086
+ "epoch": 61.35,
1087
+ "learning_rate": 1.1593794076163611e-06,
1088
+ "loss": 0.0186,
1089
+ "step": 43500
1090
+ },
1091
+ {
1092
+ "epoch": 62.06,
1093
+ "learning_rate": 1.1382228490832158e-06,
1094
+ "loss": 0.0179,
1095
+ "step": 44000
1096
+ },
1097
+ {
1098
+ "epoch": 62.06,
1099
+ "eval_gen_len": 159.4046,
1100
+ "eval_loss": 1.3091576099395752,
1101
+ "eval_rouge1": 69.9596,
1102
+ "eval_rouge2": 54.1221,
1103
+ "eval_rougeL": 56.3196,
1104
+ "eval_rougeLsum": 60.6711,
1105
+ "eval_runtime": 5735.0694,
1106
+ "eval_samples_per_second": 0.061,
1107
+ "eval_steps_per_second": 0.003,
1108
+ "step": 44000
1109
+ },
1110
+ {
1111
+ "epoch": 62.76,
1112
+ "learning_rate": 1.1170662905500706e-06,
1113
+ "loss": 0.0179,
1114
+ "step": 44500
1115
+ },
1116
+ {
1117
+ "epoch": 63.47,
1118
+ "learning_rate": 1.0959097320169253e-06,
1119
+ "loss": 0.0162,
1120
+ "step": 45000
1121
+ },
1122
+ {
1123
+ "epoch": 63.47,
1124
+ "eval_gen_len": 161.7721,
1125
+ "eval_loss": 1.3298344612121582,
1126
+ "eval_rouge1": 69.7464,
1127
+ "eval_rouge2": 54.0211,
1128
+ "eval_rougeL": 56.0314,
1129
+ "eval_rougeLsum": 60.4478,
1130
+ "eval_runtime": 5746.6407,
1131
+ "eval_samples_per_second": 0.061,
1132
+ "eval_steps_per_second": 0.003,
1133
+ "step": 45000
1134
+ },
1135
+ {
1136
+ "epoch": 64.17,
1137
+ "learning_rate": 1.07475317348378e-06,
1138
+ "loss": 0.0177,
1139
+ "step": 45500
1140
+ },
1141
+ {
1142
+ "epoch": 64.88,
1143
+ "learning_rate": 1.0535966149506348e-06,
1144
+ "loss": 0.0161,
1145
+ "step": 46000
1146
+ },
1147
+ {
1148
+ "epoch": 64.88,
1149
+ "eval_gen_len": 161.3932,
1150
+ "eval_loss": 1.3238565921783447,
1151
+ "eval_rouge1": 69.6076,
1152
+ "eval_rouge2": 53.7339,
1153
+ "eval_rougeL": 55.9025,
1154
+ "eval_rougeLsum": 60.3293,
1155
+ "eval_runtime": 5658.4636,
1156
+ "eval_samples_per_second": 0.062,
1157
+ "eval_steps_per_second": 0.003,
1158
+ "step": 46000
1159
+ }
1160
+ ],
1161
+ "max_steps": 70900,
1162
+ "num_train_epochs": 100,
1163
+ "total_flos": 7.747033211322532e+17,
1164
+ "trial_name": null,
1165
+ "trial_params": null
1166
+ }