diallomama commited on
Commit
26a0219
1 Parent(s): a4ef5cb

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_gen_len": 13.938963345712901,
4
+ "eval_loss": 1.534104043798834e-08,
5
+ "eval_rouge1": 100.0,
6
+ "eval_rouge2": 100.0,
7
+ "eval_rougeL": 100.0,
8
+ "eval_rougeLsum": 100.0,
9
+ "eval_runtime": 411.2876,
10
+ "eval_samples": 6193,
11
+ "eval_samples_per_second": 15.058,
12
+ "eval_steps_per_second": 3.766,
13
+ "train_loss": 0.00022073277012470594,
14
+ "train_runtime": 10077.5092,
15
+ "train_samples": 43423,
16
+ "train_samples_per_second": 12.927,
17
+ "train_steps_per_second": 3.232
18
+ }
eval_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_gen_len": 13.938963345712901,
4
+ "eval_loss": 1.534104043798834e-08,
5
+ "eval_rouge1": 100.0,
6
+ "eval_rouge2": 100.0,
7
+ "eval_rougeL": 100.0,
8
+ "eval_rougeLsum": 100.0,
9
+ "eval_runtime": 411.2876,
10
+ "eval_samples": 6193,
11
+ "eval_samples_per_second": 15.058,
12
+ "eval_steps_per_second": 3.766
13
+ }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "decoder_start_token_id": 0,
4
  "eos_token_id": 1,
5
  "pad_token_id": 0,
 
1
  {
 
2
  "decoder_start_token_id": 0,
3
  "eos_token_id": 1,
4
  "pad_token_id": 0,
runs/Jun09_20-49-43_eddea62cdcba/events.out.tfevents.1686354446.eddea62cdcba.4553.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc654d865cbfce57d998b1fb6192b3bdf049bb911caaef2d59b8163bc8b8cacc
3
+ size 623
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "train_loss": 0.00022073277012470594,
4
+ "train_runtime": 10077.5092,
5
+ "train_samples": 43423,
6
+ "train_samples_per_second": 12.927,
7
+ "train_steps_per_second": 3.232
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,415 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 32568,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.05,
12
+ "learning_rate": 4.923237533775485e-05,
13
+ "loss": 0.008,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.09,
18
+ "learning_rate": 4.8464750675509704e-05,
19
+ "loss": 0.0007,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.14,
24
+ "learning_rate": 4.769712601326456e-05,
25
+ "loss": 0.0008,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.18,
30
+ "learning_rate": 4.6929501351019406e-05,
31
+ "loss": 0.0003,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.23,
36
+ "learning_rate": 4.616187668877426e-05,
37
+ "loss": 0.0004,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.28,
42
+ "learning_rate": 4.539425202652911e-05,
43
+ "loss": 0.0001,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.32,
48
+ "learning_rate": 4.4626627364283966e-05,
49
+ "loss": 0.0007,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.37,
54
+ "learning_rate": 4.3859002702038817e-05,
55
+ "loss": 0.0004,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.41,
60
+ "learning_rate": 4.309137803979367e-05,
61
+ "loss": 0.0006,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.46,
66
+ "learning_rate": 4.232375337754851e-05,
67
+ "loss": 0.0009,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.51,
72
+ "learning_rate": 4.155612871530337e-05,
73
+ "loss": 0.0,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.55,
78
+ "learning_rate": 4.078850405305822e-05,
79
+ "loss": 0.0,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.6,
84
+ "learning_rate": 4.002087939081307e-05,
85
+ "loss": 0.0,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.64,
90
+ "learning_rate": 3.925325472856792e-05,
91
+ "loss": 0.0,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.69,
96
+ "learning_rate": 3.848563006632277e-05,
97
+ "loss": 0.0,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.74,
102
+ "learning_rate": 3.7718005404077624e-05,
103
+ "loss": 0.0,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.78,
108
+ "learning_rate": 3.6950380741832475e-05,
109
+ "loss": 0.0001,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.83,
114
+ "learning_rate": 3.6182756079587326e-05,
115
+ "loss": 0.0008,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.88,
120
+ "learning_rate": 3.541513141734218e-05,
121
+ "loss": 0.0,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.92,
126
+ "learning_rate": 3.4647506755097034e-05,
127
+ "loss": 0.0001,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.97,
132
+ "learning_rate": 3.3879882092851885e-05,
133
+ "loss": 0.0,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 1.01,
138
+ "learning_rate": 3.311225743060673e-05,
139
+ "loss": 0.0,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 1.06,
144
+ "learning_rate": 3.234463276836158e-05,
145
+ "loss": 0.0,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 1.11,
150
+ "learning_rate": 3.157700810611644e-05,
151
+ "loss": 0.0,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 1.15,
156
+ "learning_rate": 3.080938344387129e-05,
157
+ "loss": 0.0,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 1.2,
162
+ "learning_rate": 3.004175878162614e-05,
163
+ "loss": 0.0,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 1.24,
168
+ "learning_rate": 2.927413411938099e-05,
169
+ "loss": 0.0,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 1.29,
174
+ "learning_rate": 2.8506509457135838e-05,
175
+ "loss": 0.0,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 1.34,
180
+ "learning_rate": 2.773888479489069e-05,
181
+ "loss": 0.0,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 1.38,
186
+ "learning_rate": 2.6971260132645544e-05,
187
+ "loss": 0.0,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 1.43,
192
+ "learning_rate": 2.6203635470400394e-05,
193
+ "loss": 0.0,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 1.47,
198
+ "learning_rate": 2.543601080815525e-05,
199
+ "loss": 0.0,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 1.52,
204
+ "learning_rate": 2.4668386145910096e-05,
205
+ "loss": 0.0,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 1.57,
210
+ "learning_rate": 2.390076148366495e-05,
211
+ "loss": 0.0,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 1.61,
216
+ "learning_rate": 2.3133136821419798e-05,
217
+ "loss": 0.0,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 1.66,
222
+ "learning_rate": 2.2365512159174652e-05,
223
+ "loss": 0.0,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 1.7,
228
+ "learning_rate": 2.1597887496929503e-05,
229
+ "loss": 0.0,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 1.75,
234
+ "learning_rate": 2.0830262834684354e-05,
235
+ "loss": 0.0,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 1.8,
240
+ "learning_rate": 2.0062638172439205e-05,
241
+ "loss": 0.0,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 1.84,
246
+ "learning_rate": 1.9295013510194056e-05,
247
+ "loss": 0.0,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 1.89,
252
+ "learning_rate": 1.8527388847948907e-05,
253
+ "loss": 0.0,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 1.93,
258
+ "learning_rate": 1.7759764185703758e-05,
259
+ "loss": 0.0,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 1.98,
264
+ "learning_rate": 1.6992139523458612e-05,
265
+ "loss": 0.0,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 2.03,
270
+ "learning_rate": 1.622451486121346e-05,
271
+ "loss": 0.0,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 2.07,
276
+ "learning_rate": 1.5456890198968314e-05,
277
+ "loss": 0.0,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 2.12,
282
+ "learning_rate": 1.4689265536723165e-05,
283
+ "loss": 0.0,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 2.16,
288
+ "learning_rate": 1.3921640874478014e-05,
289
+ "loss": 0.0,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 2.21,
294
+ "learning_rate": 1.3154016212232867e-05,
295
+ "loss": 0.0,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 2.26,
300
+ "learning_rate": 1.238639154998772e-05,
301
+ "loss": 0.0,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 2.3,
306
+ "learning_rate": 1.161876688774257e-05,
307
+ "loss": 0.0,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 2.35,
312
+ "learning_rate": 1.0851142225497421e-05,
313
+ "loss": 0.0,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 2.39,
318
+ "learning_rate": 1.0083517563252274e-05,
319
+ "loss": 0.0002,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 2.44,
324
+ "learning_rate": 9.315892901007125e-06,
325
+ "loss": 0.0,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 2.49,
330
+ "learning_rate": 8.548268238761976e-06,
331
+ "loss": 0.0,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 2.53,
336
+ "learning_rate": 7.780643576516827e-06,
337
+ "loss": 0.0,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 2.58,
342
+ "learning_rate": 7.0130189142716784e-06,
343
+ "loss": 0.0,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 2.63,
348
+ "learning_rate": 6.245394252026529e-06,
349
+ "loss": 0.0,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 2.67,
354
+ "learning_rate": 5.47776958978138e-06,
355
+ "loss": 0.0,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 2.72,
360
+ "learning_rate": 4.710144927536232e-06,
361
+ "loss": 0.0,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 2.76,
366
+ "learning_rate": 3.942520265291084e-06,
367
+ "loss": 0.0,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 2.81,
372
+ "learning_rate": 3.1748956030459347e-06,
373
+ "loss": 0.0,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 2.86,
378
+ "learning_rate": 2.407270940800786e-06,
379
+ "loss": 0.0,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 2.9,
384
+ "learning_rate": 1.6396462785556374e-06,
385
+ "loss": 0.0,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 2.95,
390
+ "learning_rate": 8.720216163104889e-07,
391
+ "loss": 0.0,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 2.99,
396
+ "learning_rate": 1.0439695406534021e-07,
397
+ "loss": 0.0,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 3.0,
402
+ "step": 32568,
403
+ "total_flos": 550963785498624.0,
404
+ "train_loss": 0.00022073277012470594,
405
+ "train_runtime": 10077.5092,
406
+ "train_samples_per_second": 12.927,
407
+ "train_steps_per_second": 3.232
408
+ }
409
+ ],
410
+ "max_steps": 32568,
411
+ "num_train_epochs": 3,
412
+ "total_flos": 550963785498624.0,
413
+ "trial_name": null,
414
+ "trial_params": null
415
+ }