Hugging Face's logo

wbz0505
/
GSPretrained-base

PyTorch
t5
Model card Files
xet
Community
1
wbz0505 commited on
Commit
3ad08ad
·
verified ·
1 Parent(s): 8f793d9

Upload 10 files

Browse files
Files changed (10) hide show
  1. added_tokens.json +518 -0
  2. config.json +60 -0
  3. generation_config.json +7 -0
  4. rng_state.pth +3 -0
  5. scheduler.pt +3 -0
  6. special_tokens_map.json +107 -0
  7. spiece.model +3 -0
  8. tokenizer_config.json +112 -0
  9. trainer_state.json +3826 -0
  10. training_args.bin +3 -0
added_tokens.json ADDED
@@ -0,0 +1,518 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</Motion Tokens>": 32613,
3
+ "<0>": 32100,
4
+ "<100>": 32200,
5
+ "<101>": 32201,
6
+ "<102>": 32202,
7
+ "<103>": 32203,
8
+ "<104>": 32204,
9
+ "<105>": 32205,
10
+ "<106>": 32206,
11
+ "<107>": 32207,
12
+ "<108>": 32208,
13
+ "<109>": 32209,
14
+ "<10>": 32110,
15
+ "<110>": 32210,
16
+ "<111>": 32211,
17
+ "<112>": 32212,
18
+ "<113>": 32213,
19
+ "<114>": 32214,
20
+ "<115>": 32215,
21
+ "<116>": 32216,
22
+ "<117>": 32217,
23
+ "<118>": 32218,
24
+ "<119>": 32219,
25
+ "<11>": 32111,
26
+ "<120>": 32220,
27
+ "<121>": 32221,
28
+ "<122>": 32222,
29
+ "<123>": 32223,
30
+ "<124>": 32224,
31
+ "<125>": 32225,
32
+ "<126>": 32226,
33
+ "<127>": 32227,
34
+ "<128>": 32228,
35
+ "<129>": 32229,
36
+ "<12>": 32112,
37
+ "<130>": 32230,
38
+ "<131>": 32231,
39
+ "<132>": 32232,
40
+ "<133>": 32233,
41
+ "<134>": 32234,
42
+ "<135>": 32235,
43
+ "<136>": 32236,
44
+ "<137>": 32237,
45
+ "<138>": 32238,
46
+ "<139>": 32239,
47
+ "<13>": 32113,
48
+ "<140>": 32240,
49
+ "<141>": 32241,
50
+ "<142>": 32242,
51
+ "<143>": 32243,
52
+ "<144>": 32244,
53
+ "<145>": 32245,
54
+ "<146>": 32246,
55
+ "<147>": 32247,
56
+ "<148>": 32248,
57
+ "<149>": 32249,
58
+ "<14>": 32114,
59
+ "<150>": 32250,
60
+ "<151>": 32251,
61
+ "<152>": 32252,
62
+ "<153>": 32253,
63
+ "<154>": 32254,
64
+ "<155>": 32255,
65
+ "<156>": 32256,
66
+ "<157>": 32257,
67
+ "<158>": 32258,
68
+ "<159>": 32259,
69
+ "<15>": 32115,
70
+ "<160>": 32260,
71
+ "<161>": 32261,
72
+ "<162>": 32262,
73
+ "<163>": 32263,
74
+ "<164>": 32264,
75
+ "<165>": 32265,
76
+ "<166>": 32266,
77
+ "<167>": 32267,
78
+ "<168>": 32268,
79
+ "<169>": 32269,
80
+ "<16>": 32116,
81
+ "<170>": 32270,
82
+ "<171>": 32271,
83
+ "<172>": 32272,
84
+ "<173>": 32273,
85
+ "<174>": 32274,
86
+ "<175>": 32275,
87
+ "<176>": 32276,
88
+ "<177>": 32277,
89
+ "<178>": 32278,
90
+ "<179>": 32279,
91
+ "<17>": 32117,
92
+ "<180>": 32280,
93
+ "<181>": 32281,
94
+ "<182>": 32282,
95
+ "<183>": 32283,
96
+ "<184>": 32284,
97
+ "<185>": 32285,
98
+ "<186>": 32286,
99
+ "<187>": 32287,
100
+ "<188>": 32288,
101
+ "<189>": 32289,
102
+ "<18>": 32118,
103
+ "<190>": 32290,
104
+ "<191>": 32291,
105
+ "<192>": 32292,
106
+ "<193>": 32293,
107
+ "<194>": 32294,
108
+ "<195>": 32295,
109
+ "<196>": 32296,
110
+ "<197>": 32297,
111
+ "<198>": 32298,
112
+ "<199>": 32299,
113
+ "<19>": 32119,
114
+ "<1>": 32101,
115
+ "<200>": 32300,
116
+ "<201>": 32301,
117
+ "<202>": 32302,
118
+ "<203>": 32303,
119
+ "<204>": 32304,
120
+ "<205>": 32305,
121
+ "<206>": 32306,
122
+ "<207>": 32307,
123
+ "<208>": 32308,
124
+ "<209>": 32309,
125
+ "<20>": 32120,
126
+ "<210>": 32310,
127
+ "<211>": 32311,
128
+ "<212>": 32312,
129
+ "<213>": 32313,
130
+ "<214>": 32314,
131
+ "<215>": 32315,
132
+ "<216>": 32316,
133
+ "<217>": 32317,
134
+ "<218>": 32318,
135
+ "<219>": 32319,
136
+ "<21>": 32121,
137
+ "<220>": 32320,
138
+ "<221>": 32321,
139
+ "<222>": 32322,
140
+ "<223>": 32323,
141
+ "<224>": 32324,
142
+ "<225>": 32325,
143
+ "<226>": 32326,
144
+ "<227>": 32327,
145
+ "<228>": 32328,
146
+ "<229>": 32329,
147
+ "<22>": 32122,
148
+ "<230>": 32330,
149
+ "<231>": 32331,
150
+ "<232>": 32332,
151
+ "<233>": 32333,
152
+ "<234>": 32334,
153
+ "<235>": 32335,
154
+ "<236>": 32336,
155
+ "<237>": 32337,
156
+ "<238>": 32338,
157
+ "<239>": 32339,
158
+ "<23>": 32123,
159
+ "<240>": 32340,
160
+ "<241>": 32341,
161
+ "<242>": 32342,
162
+ "<243>": 32343,
163
+ "<244>": 32344,
164
+ "<245>": 32345,
165
+ "<246>": 32346,
166
+ "<247>": 32347,
167
+ "<248>": 32348,
168
+ "<249>": 32349,
169
+ "<24>": 32124,
170
+ "<250>": 32350,
171
+ "<251>": 32351,
172
+ "<252>": 32352,
173
+ "<253>": 32353,
174
+ "<254>": 32354,
175
+ "<255>": 32355,
176
+ "<256>": 32356,
177
+ "<257>": 32357,
178
+ "<258>": 32358,
179
+ "<259>": 32359,
180
+ "<25>": 32125,
181
+ "<260>": 32360,
182
+ "<261>": 32361,
183
+ "<262>": 32362,
184
+ "<263>": 32363,
185
+ "<264>": 32364,
186
+ "<265>": 32365,
187
+ "<266>": 32366,
188
+ "<267>": 32367,
189
+ "<268>": 32368,
190
+ "<269>": 32369,
191
+ "<26>": 32126,
192
+ "<270>": 32370,
193
+ "<271>": 32371,
194
+ "<272>": 32372,
195
+ "<273>": 32373,
196
+ "<274>": 32374,
197
+ "<275>": 32375,
198
+ "<276>": 32376,
199
+ "<277>": 32377,
200
+ "<278>": 32378,
201
+ "<279>": 32379,
202
+ "<27>": 32127,
203
+ "<280>": 32380,
204
+ "<281>": 32381,
205
+ "<282>": 32382,
206
+ "<283>": 32383,
207
+ "<284>": 32384,
208
+ "<285>": 32385,
209
+ "<286>": 32386,
210
+ "<287>": 32387,
211
+ "<288>": 32388,
212
+ "<289>": 32389,
213
+ "<28>": 32128,
214
+ "<290>": 32390,
215
+ "<291>": 32391,
216
+ "<292>": 32392,
217
+ "<293>": 32393,
218
+ "<294>": 32394,
219
+ "<295>": 32395,
220
+ "<296>": 32396,
221
+ "<297>": 32397,
222
+ "<298>": 32398,
223
+ "<299>": 32399,
224
+ "<29>": 32129,
225
+ "<2>": 32102,
226
+ "<300>": 32400,
227
+ "<301>": 32401,
228
+ "<302>": 32402,
229
+ "<303>": 32403,
230
+ "<304>": 32404,
231
+ "<305>": 32405,
232
+ "<306>": 32406,
233
+ "<307>": 32407,
234
+ "<308>": 32408,
235
+ "<309>": 32409,
236
+ "<30>": 32130,
237
+ "<310>": 32410,
238
+ "<311>": 32411,
239
+ "<312>": 32412,
240
+ "<313>": 32413,
241
+ "<314>": 32414,
242
+ "<315>": 32415,
243
+ "<316>": 32416,
244
+ "<317>": 32417,
245
+ "<318>": 32418,
246
+ "<319>": 32419,
247
+ "<31>": 32131,
248
+ "<320>": 32420,
249
+ "<321>": 32421,
250
+ "<322>": 32422,
251
+ "<323>": 32423,
252
+ "<324>": 32424,
253
+ "<325>": 32425,
254
+ "<326>": 32426,
255
+ "<327>": 32427,
256
+ "<328>": 32428,
257
+ "<329>": 32429,
258
+ "<32>": 32132,
259
+ "<330>": 32430,
260
+ "<331>": 32431,
261
+ "<332>": 32432,
262
+ "<333>": 32433,
263
+ "<334>": 32434,
264
+ "<335>": 32435,
265
+ "<336>": 32436,
266
+ "<337>": 32437,
267
+ "<338>": 32438,
268
+ "<339>": 32439,
269
+ "<33>": 32133,
270
+ "<340>": 32440,
271
+ "<341>": 32441,
272
+ "<342>": 32442,
273
+ "<343>": 32443,
274
+ "<344>": 32444,
275
+ "<345>": 32445,
276
+ "<346>": 32446,
277
+ "<347>": 32447,
278
+ "<348>": 32448,
279
+ "<349>": 32449,
280
+ "<34>": 32134,
281
+ "<350>": 32450,
282
+ "<351>": 32451,
283
+ "<352>": 32452,
284
+ "<353>": 32453,
285
+ "<354>": 32454,
286
+ "<355>": 32455,
287
+ "<356>": 32456,
288
+ "<357>": 32457,
289
+ "<358>": 32458,
290
+ "<359>": 32459,
291
+ "<35>": 32135,
292
+ "<360>": 32460,
293
+ "<361>": 32461,
294
+ "<362>": 32462,
295
+ "<363>": 32463,
296
+ "<364>": 32464,
297
+ "<365>": 32465,
298
+ "<366>": 32466,
299
+ "<367>": 32467,
300
+ "<368>": 32468,
301
+ "<369>": 32469,
302
+ "<36>": 32136,
303
+ "<370>": 32470,
304
+ "<371>": 32471,
305
+ "<372>": 32472,
306
+ "<373>": 32473,
307
+ "<374>": 32474,
308
+ "<375>": 32475,
309
+ "<376>": 32476,
310
+ "<377>": 32477,
311
+ "<378>": 32478,
312
+ "<379>": 32479,
313
+ "<37>": 32137,
314
+ "<380>": 32480,
315
+ "<381>": 32481,
316
+ "<382>": 32482,
317
+ "<383>": 32483,
318
+ "<384>": 32484,
319
+ "<385>": 32485,
320
+ "<386>": 32486,
321
+ "<387>": 32487,
322
+ "<388>": 32488,
323
+ "<389>": 32489,
324
+ "<38>": 32138,
325
+ "<390>": 32490,
326
+ "<391>": 32491,
327
+ "<392>": 32492,
328
+ "<393>": 32493,
329
+ "<394>": 32494,
330
+ "<395>": 32495,
331
+ "<396>": 32496,
332
+ "<397>": 32497,
333
+ "<398>": 32498,
334
+ "<399>": 32499,
335
+ "<39>": 32139,
336
+ "<3>": 32103,
337
+ "<400>": 32500,
338
+ "<401>": 32501,
339
+ "<402>": 32502,
340
+ "<403>": 32503,
341
+ "<404>": 32504,
342
+ "<405>": 32505,
343
+ "<406>": 32506,
344
+ "<407>": 32507,
345
+ "<408>": 32508,
346
+ "<409>": 32509,
347
+ "<40>": 32140,
348
+ "<410>": 32510,
349
+ "<411>": 32511,
350
+ "<412>": 32512,
351
+ "<413>": 32513,
352
+ "<414>": 32514,
353
+ "<415>": 32515,
354
+ "<416>": 32516,
355
+ "<417>": 32517,
356
+ "<418>": 32518,
357
+ "<419>": 32519,
358
+ "<41>": 32141,
359
+ "<420>": 32520,
360
+ "<421>": 32521,
361
+ "<422>": 32522,
362
+ "<423>": 32523,
363
+ "<424>": 32524,
364
+ "<425>": 32525,
365
+ "<426>": 32526,
366
+ "<427>": 32527,
367
+ "<428>": 32528,
368
+ "<429>": 32529,
369
+ "<42>": 32142,
370
+ "<430>": 32530,
371
+ "<431>": 32531,
372
+ "<432>": 32532,
373
+ "<433>": 32533,
374
+ "<434>": 32534,
375
+ "<435>": 32535,
376
+ "<436>": 32536,
377
+ "<437>": 32537,
378
+ "<438>": 32538,
379
+ "<439>": 32539,
380
+ "<43>": 32143,
381
+ "<440>": 32540,
382
+ "<441>": 32541,
383
+ "<442>": 32542,
384
+ "<443>": 32543,
385
+ "<444>": 32544,
386
+ "<445>": 32545,
387
+ "<446>": 32546,
388
+ "<447>": 32547,
389
+ "<448>": 32548,
390
+ "<449>": 32549,
391
+ "<44>": 32144,
392
+ "<450>": 32550,
393
+ "<451>": 32551,
394
+ "<452>": 32552,
395
+ "<453>": 32553,
396
+ "<454>": 32554,
397
+ "<455>": 32555,
398
+ "<456>": 32556,
399
+ "<457>": 32557,
400
+ "<458>": 32558,
401
+ "<459>": 32559,
402
+ "<45>": 32145,
403
+ "<460>": 32560,
404
+ "<461>": 32561,
405
+ "<462>": 32562,
406
+ "<463>": 32563,
407
+ "<464>": 32564,
408
+ "<465>": 32565,
409
+ "<466>": 32566,
410
+ "<467>": 32567,
411
+ "<468>": 32568,
412
+ "<469>": 32569,
413
+ "<46>": 32146,
414
+ "<470>": 32570,
415
+ "<471>": 32571,
416
+ "<472>": 32572,
417
+ "<473>": 32573,
418
+ "<474>": 32574,
419
+ "<475>": 32575,
420
+ "<476>": 32576,
421
+ "<477>": 32577,
422
+ "<478>": 32578,
423
+ "<479>": 32579,
424
+ "<47>": 32147,
425
+ "<480>": 32580,
426
+ "<481>": 32581,
427
+ "<482>": 32582,
428
+ "<483>": 32583,
429
+ "<484>": 32584,
430
+ "<485>": 32585,
431
+ "<486>": 32586,
432
+ "<487>": 32587,
433
+ "<488>": 32588,
434
+ "<489>": 32589,
435
+ "<48>": 32148,
436
+ "<490>": 32590,
437
+ "<491>": 32591,
438
+ "<492>": 32592,
439
+ "<493>": 32593,
440
+ "<494>": 32594,
441
+ "<495>": 32595,
442
+ "<496>": 32596,
443
+ "<497>": 32597,
444
+ "<498>": 32598,
445
+ "<499>": 32599,
446
+ "<49>": 32149,
447
+ "<4>": 32104,
448
+ "<500>": 32600,
449
+ "<501>": 32601,
450
+ "<502>": 32602,
451
+ "<503>": 32603,
452
+ "<504>": 32604,
453
+ "<505>": 32605,
454
+ "<506>": 32606,
455
+ "<507>": 32607,
456
+ "<508>": 32608,
457
+ "<509>": 32609,
458
+ "<50>": 32150,
459
+ "<510>": 32610,
460
+ "<511>": 32611,
461
+ "<51>": 32151,
462
+ "<52>": 32152,
463
+ "<53>": 32153,
464
+ "<54>": 32154,
465
+ "<55>": 32155,
466
+ "<56>": 32156,
467
+ "<57>": 32157,
468
+ "<58>": 32158,
469
+ "<59>": 32159,
470
+ "<5>": 32105,
471
+ "<60>": 32160,
472
+ "<61>": 32161,
473
+ "<62>": 32162,
474
+ "<63>": 32163,
475
+ "<64>": 32164,
476
+ "<65>": 32165,
477
+ "<66>": 32166,
478
+ "<67>": 32167,
479
+ "<68>": 32168,
480
+ "<69>": 32169,
481
+ "<6>": 32106,
482
+ "<70>": 32170,
483
+ "<71>": 32171,
484
+ "<72>": 32172,
485
+ "<73>": 32173,
486
+ "<74>": 32174,
487
+ "<75>": 32175,
488
+ "<76>": 32176,
489
+ "<77>": 32177,
490
+ "<78>": 32178,
491
+ "<79>": 32179,
492
+ "<7>": 32107,
493
+ "<80>": 32180,
494
+ "<81>": 32181,
495
+ "<82>": 32182,
496
+ "<83>": 32183,
497
+ "<84>": 32184,
498
+ "<85>": 32185,
499
+ "<86>": 32186,
500
+ "<87>": 32187,
501
+ "<88>": 32188,
502
+ "<89>": 32189,
503
+ "<8>": 32108,
504
+ "<90>": 32190,
505
+ "<91>": 32191,
506
+ "<92>": 32192,
507
+ "<93>": 32193,
508
+ "<94>": 32194,
509
+ "<95>": 32195,
510
+ "<96>": 32196,
511
+ "<97>": 32197,
512
+ "<98>": 32198,
513
+ "<99>": 32199,
514
+ "<9>": 32109,
515
+ "<Motion Tokens>": 32612,
516
+ "<Motionless>": 32614,
517
+ "<SEP>": 32615
518
+ }
config.json ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google-t5/t5-base",
3
+ "architectures": [
4
+ "T5ForConditionalGeneration"
5
+ ],
6
+ "d_ff": 3072,
7
+ "d_kv": 64,
8
+ "d_model": 768,
9
+ "decoder_start_token_id": 0,
10
+ "dense_act_fn": "relu",
11
+ "dropout_rate": 0.1,
12
+ "eos_token_id": 1,
13
+ "feed_forward_proj": "relu",
14
+ "initializer_factor": 1.0,
15
+ "is_encoder_decoder": true,
16
+ "is_gated_act": false,
17
+ "layer_norm_epsilon": 1e-06,
18
+ "model_type": "t5",
19
+ "n_positions": 512,
20
+ "num_decoder_layers": 12,
21
+ "num_heads": 12,
22
+ "num_layers": 12,
23
+ "output_past": true,
24
+ "pad_token_id": 0,
25
+ "relative_attention_max_distance": 128,
26
+ "relative_attention_num_buckets": 32,
27
+ "task_specific_params": {
28
+ "summarization": {
29
+ "early_stopping": true,
30
+ "length_penalty": 2.0,
31
+ "max_length": 200,
32
+ "min_length": 30,
33
+ "no_repeat_ngram_size": 3,
34
+ "num_beams": 4,
35
+ "prefix": "summarize: "
36
+ },
37
+ "translation_en_to_de": {
38
+ "early_stopping": true,
39
+ "max_length": 300,
40
+ "num_beams": 4,
41
+ "prefix": "translate English to German: "
42
+ },
43
+ "translation_en_to_fr": {
44
+ "early_stopping": true,
45
+ "max_length": 300,
46
+ "num_beams": 4,
47
+ "prefix": "translate English to French: "
48
+ },
49
+ "translation_en_to_ro": {
50
+ "early_stopping": true,
51
+ "max_length": 300,
52
+ "num_beams": 4,
53
+ "prefix": "translate English to Romanian: "
54
+ }
55
+ },
56
+ "torch_dtype": "float32",
57
+ "transformers_version": "4.28.1",
58
+ "use_cache": true,
59
+ "vocab_size": 32616
60
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "decoder_start_token_id": 0,
4
+ "eos_token_id": 1,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.28.1"
7
+ }
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01fd4c21f6d88e39185be4e0e886113a38c2ce485d522ad12ad252dfd392f5d9
3
+ size 14308
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00000d7c209b1cdf066f0deb3f3cc14f68c294d852f98b115f2d0af44c907dc3
3
+ size 1064
special_tokens_map.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<extra_id_0>",
4
+ "<extra_id_1>",
5
+ "<extra_id_2>",
6
+ "<extra_id_3>",
7
+ "<extra_id_4>",
8
+ "<extra_id_5>",
9
+ "<extra_id_6>",
10
+ "<extra_id_7>",
11
+ "<extra_id_8>",
12
+ "<extra_id_9>",
13
+ "<extra_id_10>",
14
+ "<extra_id_11>",
15
+ "<extra_id_12>",
16
+ "<extra_id_13>",
17
+ "<extra_id_14>",
18
+ "<extra_id_15>",
19
+ "<extra_id_16>",
20
+ "<extra_id_17>",
21
+ "<extra_id_18>",
22
+ "<extra_id_19>",
23
+ "<extra_id_20>",
24
+ "<extra_id_21>",
25
+ "<extra_id_22>",
26
+ "<extra_id_23>",
27
+ "<extra_id_24>",
28
+ "<extra_id_25>",
29
+ "<extra_id_26>",
30
+ "<extra_id_27>",
31
+ "<extra_id_28>",
32
+ "<extra_id_29>",
33
+ "<extra_id_30>",
34
+ "<extra_id_31>",
35
+ "<extra_id_32>",
36
+ "<extra_id_33>",
37
+ "<extra_id_34>",
38
+ "<extra_id_35>",
39
+ "<extra_id_36>",
40
+ "<extra_id_37>",
41
+ "<extra_id_38>",
42
+ "<extra_id_39>",
43
+ "<extra_id_40>",
44
+ "<extra_id_41>",
45
+ "<extra_id_42>",
46
+ "<extra_id_43>",
47
+ "<extra_id_44>",
48
+ "<extra_id_45>",
49
+ "<extra_id_46>",
50
+ "<extra_id_47>",
51
+ "<extra_id_48>",
52
+ "<extra_id_49>",
53
+ "<extra_id_50>",
54
+ "<extra_id_51>",
55
+ "<extra_id_52>",
56
+ "<extra_id_53>",
57
+ "<extra_id_54>",
58
+ "<extra_id_55>",
59
+ "<extra_id_56>",
60
+ "<extra_id_57>",
61
+ "<extra_id_58>",
62
+ "<extra_id_59>",
63
+ "<extra_id_60>",
64
+ "<extra_id_61>",
65
+ "<extra_id_62>",
66
+ "<extra_id_63>",
67
+ "<extra_id_64>",
68
+ "<extra_id_65>",
69
+ "<extra_id_66>",
70
+ "<extra_id_67>",
71
+ "<extra_id_68>",
72
+ "<extra_id_69>",
73
+ "<extra_id_70>",
74
+ "<extra_id_71>",
75
+ "<extra_id_72>",
76
+ "<extra_id_73>",
77
+ "<extra_id_74>",
78
+ "<extra_id_75>",
79
+ "<extra_id_76>",
80
+ "<extra_id_77>",
81
+ "<extra_id_78>",
82
+ "<extra_id_79>",
83
+ "<extra_id_80>",
84
+ "<extra_id_81>",
85
+ "<extra_id_82>",
86
+ "<extra_id_83>",
87
+ "<extra_id_84>",
88
+ "<extra_id_85>",
89
+ "<extra_id_86>",
90
+ "<extra_id_87>",
91
+ "<extra_id_88>",
92
+ "<extra_id_89>",
93
+ "<extra_id_90>",
94
+ "<extra_id_91>",
95
+ "<extra_id_92>",
96
+ "<extra_id_93>",
97
+ "<extra_id_94>",
98
+ "<extra_id_95>",
99
+ "<extra_id_96>",
100
+ "<extra_id_97>",
101
+ "<extra_id_98>",
102
+ "<extra_id_99>"
103
+ ],
104
+ "eos_token": "</s>",
105
+ "pad_token": "<pad>",
106
+ "unk_token": "<unk>"
107
+ }
spiece.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
3
+ size 791656
tokenizer_config.json ADDED
@@ -0,0 +1,112 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<extra_id_0>",
4
+ "<extra_id_1>",
5
+ "<extra_id_2>",
6
+ "<extra_id_3>",
7
+ "<extra_id_4>",
8
+ "<extra_id_5>",
9
+ "<extra_id_6>",
10
+ "<extra_id_7>",
11
+ "<extra_id_8>",
12
+ "<extra_id_9>",
13
+ "<extra_id_10>",
14
+ "<extra_id_11>",
15
+ "<extra_id_12>",
16
+ "<extra_id_13>",
17
+ "<extra_id_14>",
18
+ "<extra_id_15>",
19
+ "<extra_id_16>",
20
+ "<extra_id_17>",
21
+ "<extra_id_18>",
22
+ "<extra_id_19>",
23
+ "<extra_id_20>",
24
+ "<extra_id_21>",
25
+ "<extra_id_22>",
26
+ "<extra_id_23>",
27
+ "<extra_id_24>",
28
+ "<extra_id_25>",
29
+ "<extra_id_26>",
30
+ "<extra_id_27>",
31
+ "<extra_id_28>",
32
+ "<extra_id_29>",
33
+ "<extra_id_30>",
34
+ "<extra_id_31>",
35
+ "<extra_id_32>",
36
+ "<extra_id_33>",
37
+ "<extra_id_34>",
38
+ "<extra_id_35>",
39
+ "<extra_id_36>",
40
+ "<extra_id_37>",
41
+ "<extra_id_38>",
42
+ "<extra_id_39>",
43
+ "<extra_id_40>",
44
+ "<extra_id_41>",
45
+ "<extra_id_42>",
46
+ "<extra_id_43>",
47
+ "<extra_id_44>",
48
+ "<extra_id_45>",
49
+ "<extra_id_46>",
50
+ "<extra_id_47>",
51
+ "<extra_id_48>",
52
+ "<extra_id_49>",
53
+ "<extra_id_50>",
54
+ "<extra_id_51>",
55
+ "<extra_id_52>",
56
+ "<extra_id_53>",
57
+ "<extra_id_54>",
58
+ "<extra_id_55>",
59
+ "<extra_id_56>",
60
+ "<extra_id_57>",
61
+ "<extra_id_58>",
62
+ "<extra_id_59>",
63
+ "<extra_id_60>",
64
+ "<extra_id_61>",
65
+ "<extra_id_62>",
66
+ "<extra_id_63>",
67
+ "<extra_id_64>",
68
+ "<extra_id_65>",
69
+ "<extra_id_66>",
70
+ "<extra_id_67>",
71
+ "<extra_id_68>",
72
+ "<extra_id_69>",
73
+ "<extra_id_70>",
74
+ "<extra_id_71>",
75
+ "<extra_id_72>",
76
+ "<extra_id_73>",
77
+ "<extra_id_74>",
78
+ "<extra_id_75>",
79
+ "<extra_id_76>",
80
+ "<extra_id_77>",
81
+ "<extra_id_78>",
82
+ "<extra_id_79>",
83
+ "<extra_id_80>",
84
+ "<extra_id_81>",
85
+ "<extra_id_82>",
86
+ "<extra_id_83>",
87
+ "<extra_id_84>",
88
+ "<extra_id_85>",
89
+ "<extra_id_86>",
90
+ "<extra_id_87>",
91
+ "<extra_id_88>",
92
+ "<extra_id_89>",
93
+ "<extra_id_90>",
94
+ "<extra_id_91>",
95
+ "<extra_id_92>",
96
+ "<extra_id_93>",
97
+ "<extra_id_94>",
98
+ "<extra_id_95>",
99
+ "<extra_id_96>",
100
+ "<extra_id_97>",
101
+ "<extra_id_98>",
102
+ "<extra_id_99>"
103
+ ],
104
+ "clean_up_tokenization_spaces": true,
105
+ "eos_token": "</s>",
106
+ "extra_ids": 100,
107
+ "model_max_length": 1000000000000000019884624838656,
108
+ "pad_token": "<pad>",
109
+ "sp_model_kwargs": {},
110
+ "tokenizer_class": "T5Tokenizer",
111
+ "unk_token": "<unk>"
112
+ }
trainer_state.json ADDED
@@ -0,0 +1,3826 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 195.09022923101935,
5
+ "global_step": 300000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.33,
12
+ "learning_rate": 0.00019999862922474268,
13
+ "loss": 3.8845,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.65,
18
+ "learning_rate": 0.00019999451693655123,
19
+ "loss": 3.3985,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.98,
24
+ "learning_rate": 0.00019998766324816607,
25
+ "loss": 3.2874,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 1.3,
30
+ "learning_rate": 0.00019997806834748456,
31
+ "loss": 3.0969,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 1.63,
36
+ "learning_rate": 0.00019996573249755572,
37
+ "loss": 3.1114,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 1.95,
42
+ "learning_rate": 0.00019995065603657316,
43
+ "loss": 3.0458,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 2.28,
48
+ "learning_rate": 0.00019993283937786563,
49
+ "loss": 2.9674,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 2.6,
54
+ "learning_rate": 0.00019991228300988585,
55
+ "loss": 3.0031,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 2.93,
60
+ "learning_rate": 0.00019988898749619702,
61
+ "loss": 2.8428,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 3.25,
66
+ "learning_rate": 0.0001998629534754574,
67
+ "loss": 2.7971,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 3.58,
72
+ "learning_rate": 0.00019983418166140285,
73
+ "loss": 2.8229,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 3.9,
78
+ "learning_rate": 0.00019980267284282717,
79
+ "loss": 2.7483,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 4.23,
84
+ "learning_rate": 0.00019976842788356055,
85
+ "loss": 2.7053,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 4.55,
90
+ "learning_rate": 0.00019973144772244582,
91
+ "loss": 2.6547,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 4.88,
96
+ "learning_rate": 0.0001996917333733128,
97
+ "loss": 2.6413,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 5.2,
102
+ "learning_rate": 0.00019964928592495045,
103
+ "loss": 2.6228,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 5.53,
108
+ "learning_rate": 0.00019960410654107697,
109
+ "loss": 2.5562,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 5.85,
114
+ "learning_rate": 0.00019955619646030802,
115
+ "loss": 2.5558,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 6.18,
120
+ "learning_rate": 0.00019950555699612267,
121
+ "loss": 2.5479,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 6.5,
126
+ "learning_rate": 0.00019945218953682734,
127
+ "loss": 2.5797,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 6.5,
132
+ "eval_runtime": 0.0015,
133
+ "eval_samples_per_second": 0.0,
134
+ "eval_steps_per_second": 0.0,
135
+ "step": 10000
136
+ },
137
+ {
138
+ "epoch": 6.83,
139
+ "learning_rate": 0.000199396095545518,
140
+ "loss": 2.5277,
141
+ "step": 10500
142
+ },
143
+ {
144
+ "epoch": 7.15,
145
+ "learning_rate": 0.00019933727656003963,
146
+ "loss": 2.4641,
147
+ "step": 11000
148
+ },
149
+ {
150
+ "epoch": 7.48,
151
+ "learning_rate": 0.00019927573419294456,
152
+ "loss": 2.4278,
153
+ "step": 11500
154
+ },
155
+ {
156
+ "epoch": 7.8,
157
+ "learning_rate": 0.0001992114701314478,
158
+ "loss": 2.3937,
159
+ "step": 12000
160
+ },
161
+ {
162
+ "epoch": 8.13,
163
+ "learning_rate": 0.00019914448613738106,
164
+ "loss": 2.3969,
165
+ "step": 12500
166
+ },
167
+ {
168
+ "epoch": 8.45,
169
+ "learning_rate": 0.00019907478404714436,
170
+ "loss": 2.38,
171
+ "step": 13000
172
+ },
173
+ {
174
+ "epoch": 8.78,
175
+ "learning_rate": 0.00019900236577165576,
176
+ "loss": 2.3317,
177
+ "step": 13500
178
+ },
179
+ {
180
+ "epoch": 9.1,
181
+ "learning_rate": 0.00019892723329629887,
182
+ "loss": 2.3562,
183
+ "step": 14000
184
+ },
185
+ {
186
+ "epoch": 9.43,
187
+ "learning_rate": 0.00019884938868086835,
188
+ "loss": 2.3601,
189
+ "step": 14500
190
+ },
191
+ {
192
+ "epoch": 9.75,
193
+ "learning_rate": 0.00019876883405951377,
194
+ "loss": 2.352,
195
+ "step": 15000
196
+ },
197
+ {
198
+ "epoch": 10.08,
199
+ "learning_rate": 0.00019868557164068074,
200
+ "loss": 2.3208,
201
+ "step": 15500
202
+ },
203
+ {
204
+ "epoch": 10.4,
205
+ "learning_rate": 0.0001985996037070505,
206
+ "loss": 2.2729,
207
+ "step": 16000
208
+ },
209
+ {
210
+ "epoch": 10.73,
211
+ "learning_rate": 0.0001985109326154774,
212
+ "loss": 2.2739,
213
+ "step": 16500
214
+ },
215
+ {
216
+ "epoch": 11.06,
217
+ "learning_rate": 0.0001984195607969242,
218
+ "loss": 2.2425,
219
+ "step": 17000
220
+ },
221
+ {
222
+ "epoch": 11.38,
223
+ "learning_rate": 0.0001983254907563955,
224
+ "loss": 2.2249,
225
+ "step": 17500
226
+ },
227
+ {
228
+ "epoch": 11.71,
229
+ "learning_rate": 0.0001982287250728689,
230
+ "loss": 2.211,
231
+ "step": 18000
232
+ },
233
+ {
234
+ "epoch": 12.03,
235
+ "learning_rate": 0.0001981292663992245,
236
+ "loss": 2.2012,
237
+ "step": 18500
238
+ },
239
+ {
240
+ "epoch": 12.36,
241
+ "learning_rate": 0.00019802711746217218,
242
+ "loss": 2.2002,
243
+ "step": 19000
244
+ },
245
+ {
246
+ "epoch": 12.68,
247
+ "learning_rate": 0.00019792228106217658,
248
+ "loss": 2.1889,
249
+ "step": 19500
250
+ },
251
+ {
252
+ "epoch": 13.01,
253
+ "learning_rate": 0.00019781476007338058,
254
+ "loss": 2.1837,
255
+ "step": 20000
256
+ },
257
+ {
258
+ "epoch": 13.01,
259
+ "eval_runtime": 0.0015,
260
+ "eval_samples_per_second": 0.0,
261
+ "eval_steps_per_second": 0.0,
262
+ "step": 20000
263
+ },
264
+ {
265
+ "epoch": 13.33,
266
+ "learning_rate": 0.0001977045574435264,
267
+ "loss": 2.1589,
268
+ "step": 20500
269
+ },
270
+ {
271
+ "epoch": 13.66,
272
+ "learning_rate": 0.00019759167619387476,
273
+ "loss": 2.1809,
274
+ "step": 21000
275
+ },
276
+ {
277
+ "epoch": 13.98,
278
+ "learning_rate": 0.0001974761194191222,
279
+ "loss": 2.1251,
280
+ "step": 21500
281
+ },
282
+ {
283
+ "epoch": 14.31,
284
+ "learning_rate": 0.00019735789028731604,
285
+ "loss": 2.0866,
286
+ "step": 22000
287
+ },
288
+ {
289
+ "epoch": 14.63,
290
+ "learning_rate": 0.00019723699203976766,
291
+ "loss": 2.1219,
292
+ "step": 22500
293
+ },
294
+ {
295
+ "epoch": 14.96,
296
+ "learning_rate": 0.00019711342799096361,
297
+ "loss": 2.121,
298
+ "step": 23000
299
+ },
300
+ {
301
+ "epoch": 15.28,
302
+ "learning_rate": 0.00019698720152847468,
303
+ "loss": 2.0853,
304
+ "step": 23500
305
+ },
306
+ {
307
+ "epoch": 15.61,
308
+ "learning_rate": 0.0001968583161128631,
309
+ "loss": 2.0996,
310
+ "step": 24000
311
+ },
312
+ {
313
+ "epoch": 15.93,
314
+ "learning_rate": 0.0001967267752775877,
315
+ "loss": 2.0758,
316
+ "step": 24500
317
+ },
318
+ {
319
+ "epoch": 16.26,
320
+ "learning_rate": 0.00019659258262890683,
321
+ "loss": 2.0532,
322
+ "step": 25000
323
+ },
324
+ {
325
+ "epoch": 16.58,
326
+ "learning_rate": 0.00019645574184577982,
327
+ "loss": 2.0133,
328
+ "step": 25500
329
+ },
330
+ {
331
+ "epoch": 16.91,
332
+ "learning_rate": 0.00019631625667976583,
333
+ "loss": 1.996,
334
+ "step": 26000
335
+ },
336
+ {
337
+ "epoch": 17.23,
338
+ "learning_rate": 0.00019617413095492114,
339
+ "loss": 2.0053,
340
+ "step": 26500
341
+ },
342
+ {
343
+ "epoch": 17.56,
344
+ "learning_rate": 0.0001960293685676943,
345
+ "loss": 2.0098,
346
+ "step": 27000
347
+ },
348
+ {
349
+ "epoch": 17.88,
350
+ "learning_rate": 0.0001958819734868193,
351
+ "loss": 2.0401,
352
+ "step": 27500
353
+ },
354
+ {
355
+ "epoch": 18.21,
356
+ "learning_rate": 0.00019573194975320673,
357
+ "loss": 1.9947,
358
+ "step": 28000
359
+ },
360
+ {
361
+ "epoch": 18.53,
362
+ "learning_rate": 0.00019557930147983302,
363
+ "loss": 1.9812,
364
+ "step": 28500
365
+ },
366
+ {
367
+ "epoch": 18.86,
368
+ "learning_rate": 0.0001954240328516277,
369
+ "loss": 2.0141,
370
+ "step": 29000
371
+ },
372
+ {
373
+ "epoch": 19.18,
374
+ "learning_rate": 0.00019526614812535864,
375
+ "loss": 1.9574,
376
+ "step": 29500
377
+ },
378
+ {
379
+ "epoch": 19.51,
380
+ "learning_rate": 0.00019510565162951537,
381
+ "loss": 1.9243,
382
+ "step": 30000
383
+ },
384
+ {
385
+ "epoch": 19.51,
386
+ "eval_runtime": 0.0019,
387
+ "eval_samples_per_second": 0.0,
388
+ "eval_steps_per_second": 0.0,
389
+ "step": 30000
390
+ },
391
+ {
392
+ "epoch": 19.83,
393
+ "learning_rate": 0.0001949425477641904,
394
+ "loss": 1.923,
395
+ "step": 30500
396
+ },
397
+ {
398
+ "epoch": 20.16,
399
+ "learning_rate": 0.0001947768410009586,
400
+ "loss": 1.9097,
401
+ "step": 31000
402
+ },
403
+ {
404
+ "epoch": 20.48,
405
+ "learning_rate": 0.00019460853588275454,
406
+ "loss": 1.8941,
407
+ "step": 31500
408
+ },
409
+ {
410
+ "epoch": 20.81,
411
+ "learning_rate": 0.00019443763702374812,
412
+ "loss": 1.9159,
413
+ "step": 32000
414
+ },
415
+ {
416
+ "epoch": 21.13,
417
+ "learning_rate": 0.00019426414910921787,
418
+ "loss": 1.8793,
419
+ "step": 32500
420
+ },
421
+ {
422
+ "epoch": 21.46,
423
+ "learning_rate": 0.00019408807689542257,
424
+ "loss": 1.9007,
425
+ "step": 33000
426
+ },
427
+ {
428
+ "epoch": 21.79,
429
+ "learning_rate": 0.0001939094252094709,
430
+ "loss": 1.8847,
431
+ "step": 33500
432
+ },
433
+ {
434
+ "epoch": 22.11,
435
+ "learning_rate": 0.00019372819894918915,
436
+ "loss": 1.9223,
437
+ "step": 34000
438
+ },
439
+ {
440
+ "epoch": 22.44,
441
+ "learning_rate": 0.00019354440308298675,
442
+ "loss": 1.8933,
443
+ "step": 34500
444
+ },
445
+ {
446
+ "epoch": 22.76,
447
+ "learning_rate": 0.00019335804264972018,
448
+ "loss": 1.8753,
449
+ "step": 35000
450
+ },
451
+ {
452
+ "epoch": 23.09,
453
+ "learning_rate": 0.0001931691227585549,
454
+ "loss": 1.8752,
455
+ "step": 35500
456
+ },
457
+ {
458
+ "epoch": 23.41,
459
+ "learning_rate": 0.00019297764858882514,
460
+ "loss": 1.8201,
461
+ "step": 36000
462
+ },
463
+ {
464
+ "epoch": 23.74,
465
+ "learning_rate": 0.000192783625389892,
466
+ "loss": 1.8452,
467
+ "step": 36500
468
+ },
469
+ {
470
+ "epoch": 24.06,
471
+ "learning_rate": 0.0001925870584809995,
472
+ "loss": 1.8877,
473
+ "step": 37000
474
+ },
475
+ {
476
+ "epoch": 24.39,
477
+ "learning_rate": 0.0001923879532511287,
478
+ "loss": 1.8377,
479
+ "step": 37500
480
+ },
481
+ {
482
+ "epoch": 24.71,
483
+ "learning_rate": 0.00019218631515885006,
484
+ "loss": 1.8613,
485
+ "step": 38000
486
+ },
487
+ {
488
+ "epoch": 25.04,
489
+ "learning_rate": 0.00019198214973217378,
490
+ "loss": 1.8599,
491
+ "step": 38500
492
+ },
493
+ {
494
+ "epoch": 25.36,
495
+ "learning_rate": 0.00019177546256839812,
496
+ "loss": 1.8242,
497
+ "step": 39000
498
+ },
499
+ {
500
+ "epoch": 25.69,
501
+ "learning_rate": 0.00019156625933395614,
502
+ "loss": 1.8917,
503
+ "step": 39500
504
+ },
505
+ {
506
+ "epoch": 26.01,
507
+ "learning_rate": 0.0001913545457642601,
508
+ "loss": 1.8431,
509
+ "step": 40000
510
+ },
511
+ {
512
+ "epoch": 26.01,
513
+ "eval_runtime": 0.0016,
514
+ "eval_samples_per_second": 0.0,
515
+ "eval_steps_per_second": 0.0,
516
+ "step": 40000
517
+ },
518
+ {
519
+ "epoch": 26.34,
520
+ "learning_rate": 0.00019114032766354453,
521
+ "loss": 1.8362,
522
+ "step": 40500
523
+ },
524
+ {
525
+ "epoch": 26.66,
526
+ "learning_rate": 0.00019092361090470688,
527
+ "loss": 1.8107,
528
+ "step": 41000
529
+ },
530
+ {
531
+ "epoch": 26.99,
532
+ "learning_rate": 0.0001907044014291465,
533
+ "loss": 1.846,
534
+ "step": 41500
535
+ },
536
+ {
537
+ "epoch": 27.31,
538
+ "learning_rate": 0.00019048270524660196,
539
+ "loss": 1.8259,
540
+ "step": 42000
541
+ },
542
+ {
543
+ "epoch": 27.64,
544
+ "learning_rate": 0.00019025852843498607,
545
+ "loss": 1.7882,
546
+ "step": 42500
547
+ },
548
+ {
549
+ "epoch": 27.96,
550
+ "learning_rate": 0.00019003187714021938,
551
+ "loss": 1.8067,
552
+ "step": 43000
553
+ },
554
+ {
555
+ "epoch": 28.29,
556
+ "learning_rate": 0.00018980275757606157,
557
+ "loss": 1.837,
558
+ "step": 43500
559
+ },
560
+ {
561
+ "epoch": 28.61,
562
+ "learning_rate": 0.0001895711760239413,
563
+ "loss": 1.7629,
564
+ "step": 44000
565
+ },
566
+ {
567
+ "epoch": 28.94,
568
+ "learning_rate": 0.00018933713883278376,
569
+ "loss": 1.7724,
570
+ "step": 44500
571
+ },
572
+ {
573
+ "epoch": 29.26,
574
+ "learning_rate": 0.0001891006524188368,
575
+ "loss": 1.7355,
576
+ "step": 45000
577
+ },
578
+ {
579
+ "epoch": 29.59,
580
+ "learning_rate": 0.0001888617232654949,
581
+ "loss": 1.7706,
582
+ "step": 45500
583
+ },
584
+ {
585
+ "epoch": 29.91,
586
+ "learning_rate": 0.00018862035792312147,
587
+ "loss": 1.7667,
588
+ "step": 46000
589
+ },
590
+ {
591
+ "epoch": 30.24,
592
+ "learning_rate": 0.00018837656300886937,
593
+ "loss": 1.7589,
594
+ "step": 46500
595
+ },
596
+ {
597
+ "epoch": 30.56,
598
+ "learning_rate": 0.0001881303452064992,
599
+ "loss": 1.7815,
600
+ "step": 47000
601
+ },
602
+ {
603
+ "epoch": 30.89,
604
+ "learning_rate": 0.00018788171126619653,
605
+ "loss": 1.7756,
606
+ "step": 47500
607
+ },
608
+ {
609
+ "epoch": 31.21,
610
+ "learning_rate": 0.00018763066800438636,
611
+ "loss": 1.7395,
612
+ "step": 48000
613
+ },
614
+ {
615
+ "epoch": 31.54,
616
+ "learning_rate": 0.00018737722230354655,
617
+ "loss": 1.7551,
618
+ "step": 48500
619
+ },
620
+ {
621
+ "epoch": 31.86,
622
+ "learning_rate": 0.00018712138111201895,
623
+ "loss": 1.7464,
624
+ "step": 49000
625
+ },
626
+ {
627
+ "epoch": 32.19,
628
+ "learning_rate": 0.00018686315144381913,
629
+ "loss": 1.7418,
630
+ "step": 49500
631
+ },
632
+ {
633
+ "epoch": 32.52,
634
+ "learning_rate": 0.00018660254037844388,
635
+ "loss": 1.7134,
636
+ "step": 50000
637
+ },
638
+ {
639
+ "epoch": 32.52,
640
+ "eval_runtime": 0.0015,
641
+ "eval_samples_per_second": 0.0,
642
+ "eval_steps_per_second": 0.0,
643
+ "step": 50000
644
+ },
645
+ {
646
+ "epoch": 32.84,
647
+ "learning_rate": 0.00018633955506067718,
648
+ "loss": 1.7335,
649
+ "step": 50500
650
+ },
651
+ {
652
+ "epoch": 33.17,
653
+ "learning_rate": 0.0001860742027003944,
654
+ "loss": 1.698,
655
+ "step": 51000
656
+ },
657
+ {
658
+ "epoch": 33.49,
659
+ "learning_rate": 0.00018580649057236447,
660
+ "loss": 1.7395,
661
+ "step": 51500
662
+ },
663
+ {
664
+ "epoch": 33.82,
665
+ "learning_rate": 0.00018553642601605068,
666
+ "loss": 1.7032,
667
+ "step": 52000
668
+ },
669
+ {
670
+ "epoch": 34.14,
671
+ "learning_rate": 0.00018526401643540922,
672
+ "loss": 1.7098,
673
+ "step": 52500
674
+ },
675
+ {
676
+ "epoch": 34.47,
677
+ "learning_rate": 0.00018498926929868642,
678
+ "loss": 1.7279,
679
+ "step": 53000
680
+ },
681
+ {
682
+ "epoch": 34.79,
683
+ "learning_rate": 0.00018471219213821375,
684
+ "loss": 1.735,
685
+ "step": 53500
686
+ },
687
+ {
688
+ "epoch": 35.12,
689
+ "learning_rate": 0.00018443279255020152,
690
+ "loss": 1.6936,
691
+ "step": 54000
692
+ },
693
+ {
694
+ "epoch": 35.44,
695
+ "learning_rate": 0.00018415107819453062,
696
+ "loss": 1.6992,
697
+ "step": 54500
698
+ },
699
+ {
700
+ "epoch": 35.77,
701
+ "learning_rate": 0.00018386705679454242,
702
+ "loss": 1.706,
703
+ "step": 55000
704
+ },
705
+ {
706
+ "epoch": 36.09,
707
+ "learning_rate": 0.00018358073613682706,
708
+ "loss": 1.7532,
709
+ "step": 55500
710
+ },
711
+ {
712
+ "epoch": 36.42,
713
+ "learning_rate": 0.00018329212407100994,
714
+ "loss": 1.6917,
715
+ "step": 56000
716
+ },
717
+ {
718
+ "epoch": 36.74,
719
+ "learning_rate": 0.00018300122850953675,
720
+ "loss": 1.6996,
721
+ "step": 56500
722
+ },
723
+ {
724
+ "epoch": 37.07,
725
+ "learning_rate": 0.00018270805742745617,
726
+ "loss": 1.7492,
727
+ "step": 57000
728
+ },
729
+ {
730
+ "epoch": 37.39,
731
+ "learning_rate": 0.00018241261886220154,
732
+ "loss": 1.6722,
733
+ "step": 57500
734
+ },
735
+ {
736
+ "epoch": 37.72,
737
+ "learning_rate": 0.00018211492091337042,
738
+ "loss": 1.6911,
739
+ "step": 58000
740
+ },
741
+ {
742
+ "epoch": 38.04,
743
+ "learning_rate": 0.00018181497174250236,
744
+ "loss": 1.657,
745
+ "step": 58500
746
+ },
747
+ {
748
+ "epoch": 38.37,
749
+ "learning_rate": 0.00018151277957285543,
750
+ "loss": 1.6558,
751
+ "step": 59000
752
+ },
753
+ {
754
+ "epoch": 38.69,
755
+ "learning_rate": 0.00018120835268918063,
756
+ "loss": 1.6926,
757
+ "step": 59500
758
+ },
759
+ {
760
+ "epoch": 39.02,
761
+ "learning_rate": 0.00018090169943749476,
762
+ "loss": 1.6785,
763
+ "step": 60000
764
+ },
765
+ {
766
+ "epoch": 39.02,
767
+ "eval_runtime": 0.0012,
768
+ "eval_samples_per_second": 0.0,
769
+ "eval_steps_per_second": 0.0,
770
+ "step": 60000
771
+ },
772
+ {
773
+ "epoch": 39.34,
774
+ "learning_rate": 0.00018059282822485158,
775
+ "loss": 1.6601,
776
+ "step": 60500
777
+ },
778
+ {
779
+ "epoch": 39.67,
780
+ "learning_rate": 0.00018028174751911146,
781
+ "loss": 1.7,
782
+ "step": 61000
783
+ },
784
+ {
785
+ "epoch": 39.99,
786
+ "learning_rate": 0.00017996846584870908,
787
+ "loss": 1.6339,
788
+ "step": 61500
789
+ },
790
+ {
791
+ "epoch": 40.32,
792
+ "learning_rate": 0.00017965299180241963,
793
+ "loss": 1.6275,
794
+ "step": 62000
795
+ },
796
+ {
797
+ "epoch": 40.64,
798
+ "learning_rate": 0.00017933533402912354,
799
+ "loss": 1.6315,
800
+ "step": 62500
801
+ },
802
+ {
803
+ "epoch": 40.97,
804
+ "learning_rate": 0.00017901550123756906,
805
+ "loss": 1.6423,
806
+ "step": 63000
807
+ },
808
+ {
809
+ "epoch": 41.29,
810
+ "learning_rate": 0.00017869350219613375,
811
+ "loss": 1.6913,
812
+ "step": 63500
813
+ },
814
+ {
815
+ "epoch": 41.62,
816
+ "learning_rate": 0.000178369345732584,
817
+ "loss": 1.623,
818
+ "step": 64000
819
+ },
820
+ {
821
+ "epoch": 41.94,
822
+ "learning_rate": 0.000178043040733833,
823
+ "loss": 1.6645,
824
+ "step": 64500
825
+ },
826
+ {
827
+ "epoch": 42.27,
828
+ "learning_rate": 0.0001777145961456971,
829
+ "loss": 1.6585,
830
+ "step": 65000
831
+ },
832
+ {
833
+ "epoch": 42.59,
834
+ "learning_rate": 0.00017738402097265064,
835
+ "loss": 1.5515,
836
+ "step": 65500
837
+ },
838
+ {
839
+ "epoch": 42.92,
840
+ "learning_rate": 0.00017705132427757895,
841
+ "loss": 1.6737,
842
+ "step": 66000
843
+ },
844
+ {
845
+ "epoch": 43.25,
846
+ "learning_rate": 0.00017671651518153,
847
+ "loss": 1.6143,
848
+ "step": 66500
849
+ },
850
+ {
851
+ "epoch": 43.57,
852
+ "learning_rate": 0.00017637960286346425,
853
+ "loss": 1.6923,
854
+ "step": 67000
855
+ },
856
+ {
857
+ "epoch": 43.9,
858
+ "learning_rate": 0.0001760405965600031,
859
+ "loss": 1.623,
860
+ "step": 67500
861
+ },
862
+ {
863
+ "epoch": 44.22,
864
+ "learning_rate": 0.00017569950556517566,
865
+ "loss": 1.6416,
866
+ "step": 68000
867
+ },
868
+ {
869
+ "epoch": 44.55,
870
+ "learning_rate": 0.0001753563392301638,
871
+ "loss": 1.6161,
872
+ "step": 68500
873
+ },
874
+ {
875
+ "epoch": 44.87,
876
+ "learning_rate": 0.00017501110696304596,
877
+ "loss": 1.6079,
878
+ "step": 69000
879
+ },
880
+ {
881
+ "epoch": 45.2,
882
+ "learning_rate": 0.00017466381822853915,
883
+ "loss": 1.5893,
884
+ "step": 69500
885
+ },
886
+ {
887
+ "epoch": 45.52,
888
+ "learning_rate": 0.00017431448254773944,
889
+ "loss": 1.6074,
890
+ "step": 70000
891
+ },
892
+ {
893
+ "epoch": 45.52,
894
+ "eval_runtime": 0.0016,
895
+ "eval_samples_per_second": 0.0,
896
+ "eval_steps_per_second": 0.0,
897
+ "step": 70000
898
+ },
899
+ {
900
+ "epoch": 45.85,
901
+ "learning_rate": 0.000173963109497861,
902
+ "loss": 1.6074,
903
+ "step": 70500
904
+ },
905
+ {
906
+ "epoch": 46.17,
907
+ "learning_rate": 0.00017360970871197346,
908
+ "loss": 1.6223,
909
+ "step": 71000
910
+ },
911
+ {
912
+ "epoch": 46.5,
913
+ "learning_rate": 0.0001732542898787379,
914
+ "loss": 1.6068,
915
+ "step": 71500
916
+ },
917
+ {
918
+ "epoch": 46.82,
919
+ "learning_rate": 0.00017289686274214118,
920
+ "loss": 1.615,
921
+ "step": 72000
922
+ },
923
+ {
924
+ "epoch": 47.15,
925
+ "learning_rate": 0.00017253743710122875,
926
+ "loss": 1.5937,
927
+ "step": 72500
928
+ },
929
+ {
930
+ "epoch": 47.47,
931
+ "learning_rate": 0.00017217602280983623,
932
+ "loss": 1.6432,
933
+ "step": 73000
934
+ },
935
+ {
936
+ "epoch": 47.8,
937
+ "learning_rate": 0.00017181262977631888,
938
+ "loss": 1.6094,
939
+ "step": 73500
940
+ },
941
+ {
942
+ "epoch": 48.12,
943
+ "learning_rate": 0.00017144726796328034,
944
+ "loss": 1.6062,
945
+ "step": 74000
946
+ },
947
+ {
948
+ "epoch": 48.45,
949
+ "learning_rate": 0.00017107994738729926,
950
+ "loss": 1.5807,
951
+ "step": 74500
952
+ },
953
+ {
954
+ "epoch": 48.77,
955
+ "learning_rate": 0.00017071067811865476,
956
+ "loss": 1.6068,
957
+ "step": 75000
958
+ },
959
+ {
960
+ "epoch": 49.1,
961
+ "learning_rate": 0.00017033947028105039,
962
+ "loss": 1.5722,
963
+ "step": 75500
964
+ },
965
+ {
966
+ "epoch": 49.42,
967
+ "learning_rate": 0.00016996633405133655,
968
+ "loss": 1.5687,
969
+ "step": 76000
970
+ },
971
+ {
972
+ "epoch": 49.75,
973
+ "learning_rate": 0.00016959127965923142,
974
+ "loss": 1.6048,
975
+ "step": 76500
976
+ },
977
+ {
978
+ "epoch": 50.07,
979
+ "learning_rate": 0.0001692143173870407,
980
+ "loss": 1.5781,
981
+ "step": 77000
982
+ },
983
+ {
984
+ "epoch": 50.4,
985
+ "learning_rate": 0.0001688354575693754,
986
+ "loss": 1.5699,
987
+ "step": 77500
988
+ },
989
+ {
990
+ "epoch": 50.72,
991
+ "learning_rate": 0.00016845471059286887,
992
+ "loss": 1.5764,
993
+ "step": 78000
994
+ },
995
+ {
996
+ "epoch": 51.05,
997
+ "learning_rate": 0.0001680720868958918,
998
+ "loss": 1.577,
999
+ "step": 78500
1000
+ },
1001
+ {
1002
+ "epoch": 51.37,
1003
+ "learning_rate": 0.00016768759696826608,
1004
+ "loss": 1.5438,
1005
+ "step": 79000
1006
+ },
1007
+ {
1008
+ "epoch": 51.7,
1009
+ "learning_rate": 0.00016730125135097735,
1010
+ "loss": 1.5751,
1011
+ "step": 79500
1012
+ },
1013
+ {
1014
+ "epoch": 52.02,
1015
+ "learning_rate": 0.00016691306063588583,
1016
+ "loss": 1.6465,
1017
+ "step": 80000
1018
+ },
1019
+ {
1020
+ "epoch": 52.02,
1021
+ "eval_runtime": 0.0012,
1022
+ "eval_samples_per_second": 0.0,
1023
+ "eval_steps_per_second": 0.0,
1024
+ "step": 80000
1025
+ },
1026
+ {
1027
+ "epoch": 52.35,
1028
+ "learning_rate": 0.00016652303546543608,
1029
+ "loss": 1.6022,
1030
+ "step": 80500
1031
+ },
1032
+ {
1033
+ "epoch": 52.67,
1034
+ "learning_rate": 0.00016613118653236518,
1035
+ "loss": 1.5599,
1036
+ "step": 81000
1037
+ },
1038
+ {
1039
+ "epoch": 53.0,
1040
+ "learning_rate": 0.0001657375245794096,
1041
+ "loss": 1.5909,
1042
+ "step": 81500
1043
+ },
1044
+ {
1045
+ "epoch": 53.32,
1046
+ "learning_rate": 0.00016534206039901057,
1047
+ "loss": 1.5483,
1048
+ "step": 82000
1049
+ },
1050
+ {
1051
+ "epoch": 53.65,
1052
+ "learning_rate": 0.00016494480483301836,
1053
+ "loss": 1.5531,
1054
+ "step": 82500
1055
+ },
1056
+ {
1057
+ "epoch": 53.97,
1058
+ "learning_rate": 0.00016454576877239507,
1059
+ "loss": 1.579,
1060
+ "step": 83000
1061
+ },
1062
+ {
1063
+ "epoch": 54.3,
1064
+ "learning_rate": 0.00016414496315691581,
1065
+ "loss": 1.5757,
1066
+ "step": 83500
1067
+ },
1068
+ {
1069
+ "epoch": 54.63,
1070
+ "learning_rate": 0.000163742398974869,
1071
+ "loss": 1.5794,
1072
+ "step": 84000
1073
+ },
1074
+ {
1075
+ "epoch": 54.95,
1076
+ "learning_rate": 0.000163338087262755,
1077
+ "loss": 1.6013,
1078
+ "step": 84500
1079
+ },
1080
+ {
1081
+ "epoch": 55.28,
1082
+ "learning_rate": 0.00016293203910498376,
1083
+ "loss": 1.5567,
1084
+ "step": 85000
1085
+ },
1086
+ {
1087
+ "epoch": 55.6,
1088
+ "learning_rate": 0.00016252426563357055,
1089
+ "loss": 1.5536,
1090
+ "step": 85500
1091
+ },
1092
+ {
1093
+ "epoch": 55.93,
1094
+ "learning_rate": 0.00016211477802783103,
1095
+ "loss": 1.5524,
1096
+ "step": 86000
1097
+ },
1098
+ {
1099
+ "epoch": 56.25,
1100
+ "learning_rate": 0.00016170358751407487,
1101
+ "loss": 1.5279,
1102
+ "step": 86500
1103
+ },
1104
+ {
1105
+ "epoch": 56.58,
1106
+ "learning_rate": 0.00016129070536529766,
1107
+ "loss": 1.5523,
1108
+ "step": 87000
1109
+ },
1110
+ {
1111
+ "epoch": 56.9,
1112
+ "learning_rate": 0.00016087614290087208,
1113
+ "loss": 1.5422,
1114
+ "step": 87500
1115
+ },
1116
+ {
1117
+ "epoch": 57.23,
1118
+ "learning_rate": 0.0001604599114862375,
1119
+ "loss": 1.561,
1120
+ "step": 88000
1121
+ },
1122
+ {
1123
+ "epoch": 57.55,
1124
+ "learning_rate": 0.00016004202253258842,
1125
+ "loss": 1.5354,
1126
+ "step": 88500
1127
+ },
1128
+ {
1129
+ "epoch": 57.88,
1130
+ "learning_rate": 0.0001596224874965616,
1131
+ "loss": 1.5443,
1132
+ "step": 89000
1133
+ },
1134
+ {
1135
+ "epoch": 58.2,
1136
+ "learning_rate": 0.00015920131787992197,
1137
+ "loss": 1.509,
1138
+ "step": 89500
1139
+ },
1140
+ {
1141
+ "epoch": 58.53,
1142
+ "learning_rate": 0.00015877852522924732,
1143
+ "loss": 1.5282,
1144
+ "step": 90000
1145
+ },
1146
+ {
1147
+ "epoch": 58.53,
1148
+ "eval_runtime": 0.0013,
1149
+ "eval_samples_per_second": 0.0,
1150
+ "eval_steps_per_second": 0.0,
1151
+ "step": 90000
1152
+ },
1153
+ {
1154
+ "epoch": 58.85,
1155
+ "learning_rate": 0.00015835412113561175,
1156
+ "loss": 1.5356,
1157
+ "step": 90500
1158
+ },
1159
+ {
1160
+ "epoch": 59.18,
1161
+ "learning_rate": 0.0001579281172342679,
1162
+ "loss": 1.5452,
1163
+ "step": 91000
1164
+ },
1165
+ {
1166
+ "epoch": 59.5,
1167
+ "learning_rate": 0.00015750052520432787,
1168
+ "loss": 1.5231,
1169
+ "step": 91500
1170
+ },
1171
+ {
1172
+ "epoch": 59.83,
1173
+ "learning_rate": 0.0001570713567684432,
1174
+ "loss": 1.5276,
1175
+ "step": 92000
1176
+ },
1177
+ {
1178
+ "epoch": 60.15,
1179
+ "learning_rate": 0.00015664062369248328,
1180
+ "loss": 1.525,
1181
+ "step": 92500
1182
+ },
1183
+ {
1184
+ "epoch": 60.48,
1185
+ "learning_rate": 0.00015620833778521307,
1186
+ "loss": 1.5317,
1187
+ "step": 93000
1188
+ },
1189
+ {
1190
+ "epoch": 60.8,
1191
+ "learning_rate": 0.00015577451089796905,
1192
+ "loss": 1.5324,
1193
+ "step": 93500
1194
+ },
1195
+ {
1196
+ "epoch": 61.13,
1197
+ "learning_rate": 0.00015533915492433443,
1198
+ "loss": 1.5415,
1199
+ "step": 94000
1200
+ },
1201
+ {
1202
+ "epoch": 61.45,
1203
+ "learning_rate": 0.0001549022817998132,
1204
+ "loss": 1.5377,
1205
+ "step": 94500
1206
+ },
1207
+ {
1208
+ "epoch": 61.78,
1209
+ "learning_rate": 0.00015446390350150273,
1210
+ "loss": 1.4902,
1211
+ "step": 95000
1212
+ },
1213
+ {
1214
+ "epoch": 62.1,
1215
+ "learning_rate": 0.0001540240320477655,
1216
+ "loss": 1.4984,
1217
+ "step": 95500
1218
+ },
1219
+ {
1220
+ "epoch": 62.43,
1221
+ "learning_rate": 0.00015358267949789966,
1222
+ "loss": 1.5049,
1223
+ "step": 96000
1224
+ },
1225
+ {
1226
+ "epoch": 62.75,
1227
+ "learning_rate": 0.00015313985795180828,
1228
+ "loss": 1.527,
1229
+ "step": 96500
1230
+ },
1231
+ {
1232
+ "epoch": 63.08,
1233
+ "learning_rate": 0.00015269557954966778,
1234
+ "loss": 1.4838,
1235
+ "step": 97000
1236
+ },
1237
+ {
1238
+ "epoch": 63.4,
1239
+ "learning_rate": 0.0001522498564715949,
1240
+ "loss": 1.5343,
1241
+ "step": 97500
1242
+ },
1243
+ {
1244
+ "epoch": 63.73,
1245
+ "learning_rate": 0.00015180270093731303,
1246
+ "loss": 1.5185,
1247
+ "step": 98000
1248
+ },
1249
+ {
1250
+ "epoch": 64.05,
1251
+ "learning_rate": 0.00015135412520581702,
1252
+ "loss": 1.5109,
1253
+ "step": 98500
1254
+ },
1255
+ {
1256
+ "epoch": 64.38,
1257
+ "learning_rate": 0.00015090414157503714,
1258
+ "loss": 1.5228,
1259
+ "step": 99000
1260
+ },
1261
+ {
1262
+ "epoch": 64.7,
1263
+ "learning_rate": 0.00015045276238150192,
1264
+ "loss": 1.4978,
1265
+ "step": 99500
1266
+ },
1267
+ {
1268
+ "epoch": 65.03,
1269
+ "learning_rate": 0.00015000000000000001,
1270
+ "loss": 1.4764,
1271
+ "step": 100000
1272
+ },
1273
+ {
1274
+ "epoch": 65.03,
1275
+ "eval_runtime": 0.0016,
1276
+ "eval_samples_per_second": 0.0,
1277
+ "eval_steps_per_second": 0.0,
1278
+ "step": 100000
1279
+ },
1280
+ {
1281
+ "epoch": 65.36,
1282
+ "learning_rate": 0.00014954586684324078,
1283
+ "loss": 1.5,
1284
+ "step": 100500
1285
+ },
1286
+ {
1287
+ "epoch": 65.68,
1288
+ "learning_rate": 0.00014909037536151409,
1289
+ "loss": 1.5275,
1290
+ "step": 101000
1291
+ },
1292
+ {
1293
+ "epoch": 66.01,
1294
+ "learning_rate": 0.00014863353804234905,
1295
+ "loss": 1.4656,
1296
+ "step": 101500
1297
+ },
1298
+ {
1299
+ "epoch": 66.33,
1300
+ "learning_rate": 0.00014817536741017152,
1301
+ "loss": 1.5234,
1302
+ "step": 102000
1303
+ },
1304
+ {
1305
+ "epoch": 66.66,
1306
+ "learning_rate": 0.00014771587602596084,
1307
+ "loss": 1.5018,
1308
+ "step": 102500
1309
+ },
1310
+ {
1311
+ "epoch": 66.98,
1312
+ "learning_rate": 0.00014725507648690543,
1313
+ "loss": 1.4856,
1314
+ "step": 103000
1315
+ },
1316
+ {
1317
+ "epoch": 67.31,
1318
+ "learning_rate": 0.00014679298142605734,
1319
+ "loss": 1.5146,
1320
+ "step": 103500
1321
+ },
1322
+ {
1323
+ "epoch": 67.63,
1324
+ "learning_rate": 0.00014632960351198618,
1325
+ "loss": 1.477,
1326
+ "step": 104000
1327
+ },
1328
+ {
1329
+ "epoch": 67.96,
1330
+ "learning_rate": 0.00014586495544843152,
1331
+ "loss": 1.5255,
1332
+ "step": 104500
1333
+ },
1334
+ {
1335
+ "epoch": 68.28,
1336
+ "learning_rate": 0.00014539904997395468,
1337
+ "loss": 1.4731,
1338
+ "step": 105000
1339
+ },
1340
+ {
1341
+ "epoch": 68.61,
1342
+ "learning_rate": 0.00014493189986158965,
1343
+ "loss": 1.5142,
1344
+ "step": 105500
1345
+ },
1346
+ {
1347
+ "epoch": 68.93,
1348
+ "learning_rate": 0.00014446351791849276,
1349
+ "loss": 1.4807,
1350
+ "step": 106000
1351
+ },
1352
+ {
1353
+ "epoch": 69.26,
1354
+ "learning_rate": 0.00014399391698559152,
1355
+ "loss": 1.4664,
1356
+ "step": 106500
1357
+ },
1358
+ {
1359
+ "epoch": 69.58,
1360
+ "learning_rate": 0.00014352310993723277,
1361
+ "loss": 1.4428,
1362
+ "step": 107000
1363
+ },
1364
+ {
1365
+ "epoch": 69.91,
1366
+ "learning_rate": 0.00014305110968082952,
1367
+ "loss": 1.4612,
1368
+ "step": 107500
1369
+ },
1370
+ {
1371
+ "epoch": 70.23,
1372
+ "learning_rate": 0.00014257792915650728,
1373
+ "loss": 1.4578,
1374
+ "step": 108000
1375
+ },
1376
+ {
1377
+ "epoch": 70.56,
1378
+ "learning_rate": 0.00014210358133674912,
1379
+ "loss": 1.4706,
1380
+ "step": 108500
1381
+ },
1382
+ {
1383
+ "epoch": 70.88,
1384
+ "learning_rate": 0.00014162807922604012,
1385
+ "loss": 1.4724,
1386
+ "step": 109000
1387
+ },
1388
+ {
1389
+ "epoch": 71.21,
1390
+ "learning_rate": 0.00014115143586051088,
1391
+ "loss": 1.4551,
1392
+ "step": 109500
1393
+ },
1394
+ {
1395
+ "epoch": 71.53,
1396
+ "learning_rate": 0.00014067366430758004,
1397
+ "loss": 1.4767,
1398
+ "step": 110000
1399
+ },
1400
+ {
1401
+ "epoch": 71.53,
1402
+ "eval_runtime": 0.0013,
1403
+ "eval_samples_per_second": 0.0,
1404
+ "eval_steps_per_second": 0.0,
1405
+ "step": 110000
1406
+ },
1407
+ {
1408
+ "epoch": 71.86,
1409
+ "learning_rate": 0.00014019477766559604,
1410
+ "loss": 1.4857,
1411
+ "step": 110500
1412
+ },
1413
+ {
1414
+ "epoch": 72.18,
1415
+ "learning_rate": 0.00013971478906347806,
1416
+ "loss": 1.4728,
1417
+ "step": 111000
1418
+ },
1419
+ {
1420
+ "epoch": 72.51,
1421
+ "learning_rate": 0.00013923371166035616,
1422
+ "loss": 1.4558,
1423
+ "step": 111500
1424
+ },
1425
+ {
1426
+ "epoch": 72.83,
1427
+ "learning_rate": 0.0001387515586452103,
1428
+ "loss": 1.3975,
1429
+ "step": 112000
1430
+ },
1431
+ {
1432
+ "epoch": 73.16,
1433
+ "learning_rate": 0.000138268343236509,
1434
+ "loss": 1.507,
1435
+ "step": 112500
1436
+ },
1437
+ {
1438
+ "epoch": 73.48,
1439
+ "learning_rate": 0.00013778407868184672,
1440
+ "loss": 1.4588,
1441
+ "step": 113000
1442
+ },
1443
+ {
1444
+ "epoch": 73.81,
1445
+ "learning_rate": 0.0001372987782575809,
1446
+ "loss": 1.4787,
1447
+ "step": 113500
1448
+ },
1449
+ {
1450
+ "epoch": 74.13,
1451
+ "learning_rate": 0.00013681245526846783,
1452
+ "loss": 1.4403,
1453
+ "step": 114000
1454
+ },
1455
+ {
1456
+ "epoch": 74.46,
1457
+ "learning_rate": 0.00013632512304729785,
1458
+ "loss": 1.4818,
1459
+ "step": 114500
1460
+ },
1461
+ {
1462
+ "epoch": 74.78,
1463
+ "learning_rate": 0.00013583679495453,
1464
+ "loss": 1.4553,
1465
+ "step": 115000
1466
+ },
1467
+ {
1468
+ "epoch": 75.11,
1469
+ "learning_rate": 0.00013534748437792573,
1470
+ "loss": 1.4651,
1471
+ "step": 115500
1472
+ },
1473
+ {
1474
+ "epoch": 75.43,
1475
+ "learning_rate": 0.00013485720473218154,
1476
+ "loss": 1.462,
1477
+ "step": 116000
1478
+ },
1479
+ {
1480
+ "epoch": 75.76,
1481
+ "learning_rate": 0.00013436596945856164,
1482
+ "loss": 1.4843,
1483
+ "step": 116500
1484
+ },
1485
+ {
1486
+ "epoch": 76.09,
1487
+ "learning_rate": 0.00013387379202452917,
1488
+ "loss": 1.4452,
1489
+ "step": 117000
1490
+ },
1491
+ {
1492
+ "epoch": 76.41,
1493
+ "learning_rate": 0.0001333806859233771,
1494
+ "loss": 1.4007,
1495
+ "step": 117500
1496
+ },
1497
+ {
1498
+ "epoch": 76.74,
1499
+ "learning_rate": 0.00013288666467385833,
1500
+ "loss": 1.4758,
1501
+ "step": 118000
1502
+ },
1503
+ {
1504
+ "epoch": 77.06,
1505
+ "learning_rate": 0.00013239174181981495,
1506
+ "loss": 1.505,
1507
+ "step": 118500
1508
+ },
1509
+ {
1510
+ "epoch": 77.39,
1511
+ "learning_rate": 0.00013189593092980702,
1512
+ "loss": 1.4362,
1513
+ "step": 119000
1514
+ },
1515
+ {
1516
+ "epoch": 77.71,
1517
+ "learning_rate": 0.00013139924559674052,
1518
+ "loss": 1.4407,
1519
+ "step": 119500
1520
+ },
1521
+ {
1522
+ "epoch": 78.04,
1523
+ "learning_rate": 0.00013090169943749476,
1524
+ "loss": 1.4661,
1525
+ "step": 120000
1526
+ },
1527
+ {
1528
+ "epoch": 78.04,
1529
+ "eval_runtime": 0.0013,
1530
+ "eval_samples_per_second": 0.0,
1531
+ "eval_steps_per_second": 0.0,
1532
+ "step": 120000
1533
+ },
1534
+ {
1535
+ "epoch": 78.36,
1536
+ "learning_rate": 0.00013040330609254903,
1537
+ "loss": 1.4859,
1538
+ "step": 120500
1539
+ },
1540
+ {
1541
+ "epoch": 78.69,
1542
+ "learning_rate": 0.00012990407922560868,
1543
+ "loss": 1.4639,
1544
+ "step": 121000
1545
+ },
1546
+ {
1547
+ "epoch": 79.01,
1548
+ "learning_rate": 0.0001294040325232304,
1549
+ "loss": 1.4493,
1550
+ "step": 121500
1551
+ },
1552
+ {
1553
+ "epoch": 79.34,
1554
+ "learning_rate": 0.00012890317969444716,
1555
+ "loss": 1.4161,
1556
+ "step": 122000
1557
+ },
1558
+ {
1559
+ "epoch": 79.66,
1560
+ "learning_rate": 0.00012840153447039228,
1561
+ "loss": 1.4757,
1562
+ "step": 122500
1563
+ },
1564
+ {
1565
+ "epoch": 79.99,
1566
+ "learning_rate": 0.00012789911060392294,
1567
+ "loss": 1.4217,
1568
+ "step": 123000
1569
+ },
1570
+ {
1571
+ "epoch": 80.31,
1572
+ "learning_rate": 0.00012739592186924328,
1573
+ "loss": 1.4079,
1574
+ "step": 123500
1575
+ },
1576
+ {
1577
+ "epoch": 80.64,
1578
+ "learning_rate": 0.00012689198206152657,
1579
+ "loss": 1.4328,
1580
+ "step": 124000
1581
+ },
1582
+ {
1583
+ "epoch": 80.96,
1584
+ "learning_rate": 0.0001263873049965373,
1585
+ "loss": 1.4509,
1586
+ "step": 124500
1587
+ },
1588
+ {
1589
+ "epoch": 81.29,
1590
+ "learning_rate": 0.00012588190451025207,
1591
+ "loss": 1.3947,
1592
+ "step": 125000
1593
+ },
1594
+ {
1595
+ "epoch": 81.61,
1596
+ "learning_rate": 0.00012537579445848058,
1597
+ "loss": 1.4561,
1598
+ "step": 125500
1599
+ },
1600
+ {
1601
+ "epoch": 81.94,
1602
+ "learning_rate": 0.0001248689887164855,
1603
+ "loss": 1.4229,
1604
+ "step": 126000
1605
+ },
1606
+ {
1607
+ "epoch": 82.26,
1608
+ "learning_rate": 0.00012436150117860225,
1609
+ "loss": 1.4349,
1610
+ "step": 126500
1611
+ },
1612
+ {
1613
+ "epoch": 82.59,
1614
+ "learning_rate": 0.0001238533457578581,
1615
+ "loss": 1.422,
1616
+ "step": 127000
1617
+ },
1618
+ {
1619
+ "epoch": 82.91,
1620
+ "learning_rate": 0.00012334453638559057,
1621
+ "loss": 1.4223,
1622
+ "step": 127500
1623
+ },
1624
+ {
1625
+ "epoch": 83.24,
1626
+ "learning_rate": 0.00012283508701106557,
1627
+ "loss": 1.404,
1628
+ "step": 128000
1629
+ },
1630
+ {
1631
+ "epoch": 83.56,
1632
+ "learning_rate": 0.00012232501160109514,
1633
+ "loss": 1.4299,
1634
+ "step": 128500
1635
+ },
1636
+ {
1637
+ "epoch": 83.89,
1638
+ "learning_rate": 0.00012181432413965428,
1639
+ "loss": 1.4128,
1640
+ "step": 129000
1641
+ },
1642
+ {
1643
+ "epoch": 84.21,
1644
+ "learning_rate": 0.00012130303862749767,
1645
+ "loss": 1.3965,
1646
+ "step": 129500
1647
+ },
1648
+ {
1649
+ "epoch": 84.54,
1650
+ "learning_rate": 0.00012079116908177593,
1651
+ "loss": 1.4154,
1652
+ "step": 130000
1653
+ },
1654
+ {
1655
+ "epoch": 84.54,
1656
+ "eval_runtime": 0.0013,
1657
+ "eval_samples_per_second": 0.0,
1658
+ "eval_steps_per_second": 0.0,
1659
+ "step": 130000
1660
+ },
1661
+ {
1662
+ "epoch": 84.86,
1663
+ "learning_rate": 0.00012027872953565125,
1664
+ "loss": 1.4085,
1665
+ "step": 130500
1666
+ },
1667
+ {
1668
+ "epoch": 85.19,
1669
+ "learning_rate": 0.00011976573403791262,
1670
+ "loss": 1.4114,
1671
+ "step": 131000
1672
+ },
1673
+ {
1674
+ "epoch": 85.51,
1675
+ "learning_rate": 0.00011925219665259075,
1676
+ "loss": 1.4066,
1677
+ "step": 131500
1678
+ },
1679
+ {
1680
+ "epoch": 85.84,
1681
+ "learning_rate": 0.00011873813145857249,
1682
+ "loss": 1.4308,
1683
+ "step": 132000
1684
+ },
1685
+ {
1686
+ "epoch": 86.16,
1687
+ "learning_rate": 0.00011822355254921478,
1688
+ "loss": 1.4277,
1689
+ "step": 132500
1690
+ },
1691
+ {
1692
+ "epoch": 86.49,
1693
+ "learning_rate": 0.00011770847403195834,
1694
+ "loss": 1.3809,
1695
+ "step": 133000
1696
+ },
1697
+ {
1698
+ "epoch": 86.82,
1699
+ "learning_rate": 0.00011719291002794096,
1700
+ "loss": 1.4093,
1701
+ "step": 133500
1702
+ },
1703
+ {
1704
+ "epoch": 87.14,
1705
+ "learning_rate": 0.00011667687467161024,
1706
+ "loss": 1.4209,
1707
+ "step": 134000
1708
+ },
1709
+ {
1710
+ "epoch": 87.47,
1711
+ "learning_rate": 0.00011616038211033613,
1712
+ "loss": 1.4072,
1713
+ "step": 134500
1714
+ },
1715
+ {
1716
+ "epoch": 87.79,
1717
+ "learning_rate": 0.0001156434465040231,
1718
+ "loss": 1.4197,
1719
+ "step": 135000
1720
+ },
1721
+ {
1722
+ "epoch": 88.12,
1723
+ "learning_rate": 0.00011512608202472194,
1724
+ "loss": 1.3941,
1725
+ "step": 135500
1726
+ },
1727
+ {
1728
+ "epoch": 88.44,
1729
+ "learning_rate": 0.00011460830285624118,
1730
+ "loss": 1.3687,
1731
+ "step": 136000
1732
+ },
1733
+ {
1734
+ "epoch": 88.77,
1735
+ "learning_rate": 0.00011409012319375827,
1736
+ "loss": 1.3804,
1737
+ "step": 136500
1738
+ },
1739
+ {
1740
+ "epoch": 89.09,
1741
+ "learning_rate": 0.00011357155724343045,
1742
+ "loss": 1.3745,
1743
+ "step": 137000
1744
+ },
1745
+ {
1746
+ "epoch": 89.42,
1747
+ "learning_rate": 0.00011305261922200519,
1748
+ "loss": 1.4224,
1749
+ "step": 137500
1750
+ },
1751
+ {
1752
+ "epoch": 89.74,
1753
+ "learning_rate": 0.00011253332335643043,
1754
+ "loss": 1.3903,
1755
+ "step": 138000
1756
+ },
1757
+ {
1758
+ "epoch": 90.07,
1759
+ "learning_rate": 0.00011201368388346471,
1760
+ "loss": 1.4107,
1761
+ "step": 138500
1762
+ },
1763
+ {
1764
+ "epoch": 90.39,
1765
+ "learning_rate": 0.00011149371504928668,
1766
+ "loss": 1.4075,
1767
+ "step": 139000
1768
+ },
1769
+ {
1770
+ "epoch": 90.72,
1771
+ "learning_rate": 0.00011097343110910452,
1772
+ "loss": 1.3999,
1773
+ "step": 139500
1774
+ },
1775
+ {
1776
+ "epoch": 91.04,
1777
+ "learning_rate": 0.00011045284632676536,
1778
+ "loss": 1.4005,
1779
+ "step": 140000
1780
+ },
1781
+ {
1782
+ "epoch": 91.04,
1783
+ "eval_runtime": 0.0012,
1784
+ "eval_samples_per_second": 0.0,
1785
+ "eval_steps_per_second": 0.0,
1786
+ "step": 140000
1787
+ },
1788
+ {
1789
+ "epoch": 91.37,
1790
+ "learning_rate": 0.00010993197497436391,
1791
+ "loss": 1.4144,
1792
+ "step": 140500
1793
+ },
1794
+ {
1795
+ "epoch": 91.69,
1796
+ "learning_rate": 0.00010941083133185146,
1797
+ "loss": 1.4213,
1798
+ "step": 141000
1799
+ },
1800
+ {
1801
+ "epoch": 92.02,
1802
+ "learning_rate": 0.00010888942968664417,
1803
+ "loss": 1.3982,
1804
+ "step": 141500
1805
+ },
1806
+ {
1807
+ "epoch": 92.34,
1808
+ "learning_rate": 0.00010836778433323158,
1809
+ "loss": 1.4067,
1810
+ "step": 142000
1811
+ },
1812
+ {
1813
+ "epoch": 92.67,
1814
+ "learning_rate": 0.0001078459095727845,
1815
+ "loss": 1.4014,
1816
+ "step": 142500
1817
+ },
1818
+ {
1819
+ "epoch": 92.99,
1820
+ "learning_rate": 0.00010732381971276318,
1821
+ "loss": 1.408,
1822
+ "step": 143000
1823
+ },
1824
+ {
1825
+ "epoch": 93.32,
1826
+ "learning_rate": 0.00010680152906652483,
1827
+ "loss": 1.3757,
1828
+ "step": 143500
1829
+ },
1830
+ {
1831
+ "epoch": 93.64,
1832
+ "learning_rate": 0.00010627905195293135,
1833
+ "loss": 1.3949,
1834
+ "step": 144000
1835
+ },
1836
+ {
1837
+ "epoch": 93.97,
1838
+ "learning_rate": 0.00010575640269595675,
1839
+ "loss": 1.3828,
1840
+ "step": 144500
1841
+ },
1842
+ {
1843
+ "epoch": 94.29,
1844
+ "learning_rate": 0.0001052335956242944,
1845
+ "loss": 1.3768,
1846
+ "step": 145000
1847
+ },
1848
+ {
1849
+ "epoch": 94.62,
1850
+ "learning_rate": 0.00010471064507096426,
1851
+ "loss": 1.3962,
1852
+ "step": 145500
1853
+ },
1854
+ {
1855
+ "epoch": 94.94,
1856
+ "learning_rate": 0.00010418756537291996,
1857
+ "loss": 1.3791,
1858
+ "step": 146000
1859
+ },
1860
+ {
1861
+ "epoch": 95.27,
1862
+ "learning_rate": 0.00010366437087065564,
1863
+ "loss": 1.3751,
1864
+ "step": 146500
1865
+ },
1866
+ {
1867
+ "epoch": 95.59,
1868
+ "learning_rate": 0.00010314107590781284,
1869
+ "loss": 1.3748,
1870
+ "step": 147000
1871
+ },
1872
+ {
1873
+ "epoch": 95.92,
1874
+ "learning_rate": 0.00010261769483078733,
1875
+ "loss": 1.3773,
1876
+ "step": 147500
1877
+ },
1878
+ {
1879
+ "epoch": 96.24,
1880
+ "learning_rate": 0.0001020942419883357,
1881
+ "loss": 1.3847,
1882
+ "step": 148000
1883
+ },
1884
+ {
1885
+ "epoch": 96.57,
1886
+ "learning_rate": 0.00010157073173118208,
1887
+ "loss": 1.3949,
1888
+ "step": 148500
1889
+ },
1890
+ {
1891
+ "epoch": 96.89,
1892
+ "learning_rate": 0.00010104717841162458,
1893
+ "loss": 1.3732,
1894
+ "step": 149000
1895
+ },
1896
+ {
1897
+ "epoch": 97.22,
1898
+ "learning_rate": 0.00010052359638314195,
1899
+ "loss": 1.3609,
1900
+ "step": 149500
1901
+ },
1902
+ {
1903
+ "epoch": 97.55,
1904
+ "learning_rate": 0.0001,
1905
+ "loss": 1.3965,
1906
+ "step": 150000
1907
+ },
1908
+ {
1909
+ "epoch": 97.55,
1910
+ "eval_runtime": 0.0012,
1911
+ "eval_samples_per_second": 0.0,
1912
+ "eval_steps_per_second": 0.0,
1913
+ "step": 150000
1914
+ },
1915
+ {
1916
+ "epoch": 97.87,
1917
+ "learning_rate": 9.947640361685804e-05,
1918
+ "loss": 1.3683,
1919
+ "step": 150500
1920
+ },
1921
+ {
1922
+ "epoch": 98.2,
1923
+ "learning_rate": 9.895282158837545e-05,
1924
+ "loss": 1.3975,
1925
+ "step": 151000
1926
+ },
1927
+ {
1928
+ "epoch": 98.52,
1929
+ "learning_rate": 9.842926826881796e-05,
1930
+ "loss": 1.3746,
1931
+ "step": 151500
1932
+ },
1933
+ {
1934
+ "epoch": 98.85,
1935
+ "learning_rate": 9.790575801166432e-05,
1936
+ "loss": 1.3471,
1937
+ "step": 152000
1938
+ },
1939
+ {
1940
+ "epoch": 99.17,
1941
+ "learning_rate": 9.73823051692127e-05,
1942
+ "loss": 1.3849,
1943
+ "step": 152500
1944
+ },
1945
+ {
1946
+ "epoch": 99.5,
1947
+ "learning_rate": 9.685892409218717e-05,
1948
+ "loss": 1.3635,
1949
+ "step": 153000
1950
+ },
1951
+ {
1952
+ "epoch": 99.82,
1953
+ "learning_rate": 9.633562912934436e-05,
1954
+ "loss": 1.3589,
1955
+ "step": 153500
1956
+ },
1957
+ {
1958
+ "epoch": 100.15,
1959
+ "learning_rate": 9.581243462708006e-05,
1960
+ "loss": 1.3696,
1961
+ "step": 154000
1962
+ },
1963
+ {
1964
+ "epoch": 100.47,
1965
+ "learning_rate": 9.528935492903575e-05,
1966
+ "loss": 1.3769,
1967
+ "step": 154500
1968
+ },
1969
+ {
1970
+ "epoch": 100.8,
1971
+ "learning_rate": 9.476640437570562e-05,
1972
+ "loss": 1.3547,
1973
+ "step": 155000
1974
+ },
1975
+ {
1976
+ "epoch": 101.12,
1977
+ "learning_rate": 9.424359730404329e-05,
1978
+ "loss": 1.3267,
1979
+ "step": 155500
1980
+ },
1981
+ {
1982
+ "epoch": 101.45,
1983
+ "learning_rate": 9.372094804706867e-05,
1984
+ "loss": 1.3589,
1985
+ "step": 156000
1986
+ },
1987
+ {
1988
+ "epoch": 101.77,
1989
+ "learning_rate": 9.319847093347522e-05,
1990
+ "loss": 1.3543,
1991
+ "step": 156500
1992
+ },
1993
+ {
1994
+ "epoch": 102.1,
1995
+ "learning_rate": 9.267618028723686e-05,
1996
+ "loss": 1.3551,
1997
+ "step": 157000
1998
+ },
1999
+ {
2000
+ "epoch": 102.42,
2001
+ "learning_rate": 9.215409042721552e-05,
2002
+ "loss": 1.3655,
2003
+ "step": 157500
2004
+ },
2005
+ {
2006
+ "epoch": 102.75,
2007
+ "learning_rate": 9.163221566676847e-05,
2008
+ "loss": 1.3621,
2009
+ "step": 158000
2010
+ },
2011
+ {
2012
+ "epoch": 103.07,
2013
+ "learning_rate": 9.111057031335585e-05,
2014
+ "loss": 1.3626,
2015
+ "step": 158500
2016
+ },
2017
+ {
2018
+ "epoch": 103.4,
2019
+ "learning_rate": 9.058916866814858e-05,
2020
+ "loss": 1.3766,
2021
+ "step": 159000
2022
+ },
2023
+ {
2024
+ "epoch": 103.72,
2025
+ "learning_rate": 9.006802502563612e-05,
2026
+ "loss": 1.3416,
2027
+ "step": 159500
2028
+ },
2029
+ {
2030
+ "epoch": 104.05,
2031
+ "learning_rate": 8.954715367323468e-05,
2032
+ "loss": 1.3788,
2033
+ "step": 160000
2034
+ },
2035
+ {
2036
+ "epoch": 104.05,
2037
+ "eval_runtime": 0.0016,
2038
+ "eval_samples_per_second": 0.0,
2039
+ "eval_steps_per_second": 0.0,
2040
+ "step": 160000
2041
+ },
2042
+ {
2043
+ "epoch": 104.37,
2044
+ "learning_rate": 8.902656889089548e-05,
2045
+ "loss": 1.321,
2046
+ "step": 160500
2047
+ },
2048
+ {
2049
+ "epoch": 104.7,
2050
+ "learning_rate": 8.850628495071336e-05,
2051
+ "loss": 1.3473,
2052
+ "step": 161000
2053
+ },
2054
+ {
2055
+ "epoch": 105.02,
2056
+ "learning_rate": 8.79863161165353e-05,
2057
+ "loss": 1.318,
2058
+ "step": 161500
2059
+ },
2060
+ {
2061
+ "epoch": 105.35,
2062
+ "learning_rate": 8.746667664356956e-05,
2063
+ "loss": 1.341,
2064
+ "step": 162000
2065
+ },
2066
+ {
2067
+ "epoch": 105.67,
2068
+ "learning_rate": 8.694738077799488e-05,
2069
+ "loss": 1.3762,
2070
+ "step": 162500
2071
+ },
2072
+ {
2073
+ "epoch": 106.0,
2074
+ "learning_rate": 8.642844275656957e-05,
2075
+ "loss": 1.3681,
2076
+ "step": 163000
2077
+ },
2078
+ {
2079
+ "epoch": 106.32,
2080
+ "learning_rate": 8.590987680624174e-05,
2081
+ "loss": 1.3809,
2082
+ "step": 163500
2083
+ },
2084
+ {
2085
+ "epoch": 106.65,
2086
+ "learning_rate": 8.539169714375885e-05,
2087
+ "loss": 1.3593,
2088
+ "step": 164000
2089
+ },
2090
+ {
2091
+ "epoch": 106.97,
2092
+ "learning_rate": 8.487391797527808e-05,
2093
+ "loss": 1.3202,
2094
+ "step": 164500
2095
+ },
2096
+ {
2097
+ "epoch": 107.3,
2098
+ "learning_rate": 8.435655349597689e-05,
2099
+ "loss": 1.3989,
2100
+ "step": 165000
2101
+ },
2102
+ {
2103
+ "epoch": 107.62,
2104
+ "learning_rate": 8.383961788966391e-05,
2105
+ "loss": 1.3576,
2106
+ "step": 165500
2107
+ },
2108
+ {
2109
+ "epoch": 107.95,
2110
+ "learning_rate": 8.332312532838978e-05,
2111
+ "loss": 1.3826,
2112
+ "step": 166000
2113
+ },
2114
+ {
2115
+ "epoch": 108.28,
2116
+ "learning_rate": 8.280708997205904e-05,
2117
+ "loss": 1.3641,
2118
+ "step": 166500
2119
+ },
2120
+ {
2121
+ "epoch": 108.6,
2122
+ "learning_rate": 8.229152596804168e-05,
2123
+ "loss": 1.3701,
2124
+ "step": 167000
2125
+ },
2126
+ {
2127
+ "epoch": 108.93,
2128
+ "learning_rate": 8.177644745078526e-05,
2129
+ "loss": 1.3053,
2130
+ "step": 167500
2131
+ },
2132
+ {
2133
+ "epoch": 109.25,
2134
+ "learning_rate": 8.126186854142752e-05,
2135
+ "loss": 1.3344,
2136
+ "step": 168000
2137
+ },
2138
+ {
2139
+ "epoch": 109.58,
2140
+ "learning_rate": 8.074780334740928e-05,
2141
+ "loss": 1.329,
2142
+ "step": 168500
2143
+ },
2144
+ {
2145
+ "epoch": 109.9,
2146
+ "learning_rate": 8.023426596208739e-05,
2147
+ "loss": 1.3179,
2148
+ "step": 169000
2149
+ },
2150
+ {
2151
+ "epoch": 110.23,
2152
+ "learning_rate": 7.972127046434878e-05,
2153
+ "loss": 1.3291,
2154
+ "step": 169500
2155
+ },
2156
+ {
2157
+ "epoch": 110.55,
2158
+ "learning_rate": 7.920883091822408e-05,
2159
+ "loss": 1.3191,
2160
+ "step": 170000
2161
+ },
2162
+ {
2163
+ "epoch": 110.55,
2164
+ "eval_runtime": 0.0014,
2165
+ "eval_samples_per_second": 0.0,
2166
+ "eval_steps_per_second": 0.0,
2167
+ "step": 170000
2168
+ },
2169
+ {
2170
+ "epoch": 110.88,
2171
+ "learning_rate": 7.869696137250235e-05,
2172
+ "loss": 1.3441,
2173
+ "step": 170500
2174
+ },
2175
+ {
2176
+ "epoch": 111.2,
2177
+ "learning_rate": 7.818567586034577e-05,
2178
+ "loss": 1.354,
2179
+ "step": 171000
2180
+ },
2181
+ {
2182
+ "epoch": 111.53,
2183
+ "learning_rate": 7.767498839890488e-05,
2184
+ "loss": 1.3219,
2185
+ "step": 171500
2186
+ },
2187
+ {
2188
+ "epoch": 111.85,
2189
+ "learning_rate": 7.716491298893442e-05,
2190
+ "loss": 1.3389,
2191
+ "step": 172000
2192
+ },
2193
+ {
2194
+ "epoch": 112.18,
2195
+ "learning_rate": 7.66554636144095e-05,
2196
+ "loss": 1.3039,
2197
+ "step": 172500
2198
+ },
2199
+ {
2200
+ "epoch": 112.5,
2201
+ "learning_rate": 7.614665424214193e-05,
2202
+ "loss": 1.3437,
2203
+ "step": 173000
2204
+ },
2205
+ {
2206
+ "epoch": 112.83,
2207
+ "learning_rate": 7.563849882139776e-05,
2208
+ "loss": 1.3324,
2209
+ "step": 173500
2210
+ },
2211
+ {
2212
+ "epoch": 113.15,
2213
+ "learning_rate": 7.513101128351454e-05,
2214
+ "loss": 1.3324,
2215
+ "step": 174000
2216
+ },
2217
+ {
2218
+ "epoch": 113.48,
2219
+ "learning_rate": 7.462420554151944e-05,
2220
+ "loss": 1.3431,
2221
+ "step": 174500
2222
+ },
2223
+ {
2224
+ "epoch": 113.8,
2225
+ "learning_rate": 7.411809548974792e-05,
2226
+ "loss": 1.3103,
2227
+ "step": 175000
2228
+ },
2229
+ {
2230
+ "epoch": 114.13,
2231
+ "learning_rate": 7.361269500346274e-05,
2232
+ "loss": 1.3188,
2233
+ "step": 175500
2234
+ },
2235
+ {
2236
+ "epoch": 114.45,
2237
+ "learning_rate": 7.310801793847344e-05,
2238
+ "loss": 1.3246,
2239
+ "step": 176000
2240
+ },
2241
+ {
2242
+ "epoch": 114.78,
2243
+ "learning_rate": 7.260407813075676e-05,
2244
+ "loss": 1.3444,
2245
+ "step": 176500
2246
+ },
2247
+ {
2248
+ "epoch": 115.1,
2249
+ "learning_rate": 7.210088939607708e-05,
2250
+ "loss": 1.3134,
2251
+ "step": 177000
2252
+ },
2253
+ {
2254
+ "epoch": 115.43,
2255
+ "learning_rate": 7.159846552960774e-05,
2256
+ "loss": 1.3163,
2257
+ "step": 177500
2258
+ },
2259
+ {
2260
+ "epoch": 115.75,
2261
+ "learning_rate": 7.109682030555283e-05,
2262
+ "loss": 1.3096,
2263
+ "step": 178000
2264
+ },
2265
+ {
2266
+ "epoch": 116.08,
2267
+ "learning_rate": 7.059596747676962e-05,
2268
+ "loss": 1.2948,
2269
+ "step": 178500
2270
+ },
2271
+ {
2272
+ "epoch": 116.4,
2273
+ "learning_rate": 7.009592077439134e-05,
2274
+ "loss": 1.3084,
2275
+ "step": 179000
2276
+ },
2277
+ {
2278
+ "epoch": 116.73,
2279
+ "learning_rate": 6.959669390745097e-05,
2280
+ "loss": 1.3069,
2281
+ "step": 179500
2282
+ },
2283
+ {
2284
+ "epoch": 117.05,
2285
+ "learning_rate": 6.909830056250527e-05,
2286
+ "loss": 1.3097,
2287
+ "step": 180000
2288
+ },
2289
+ {
2290
+ "epoch": 117.05,
2291
+ "eval_runtime": 0.0013,
2292
+ "eval_samples_per_second": 0.0,
2293
+ "eval_steps_per_second": 0.0,
2294
+ "step": 180000
2295
+ },
2296
+ {
2297
+ "epoch": 117.38,
2298
+ "learning_rate": 6.860075440325951e-05,
2299
+ "loss": 1.3253,
2300
+ "step": 180500
2301
+ },
2302
+ {
2303
+ "epoch": 117.7,
2304
+ "learning_rate": 6.8104069070193e-05,
2305
+ "loss": 1.326,
2306
+ "step": 181000
2307
+ },
2308
+ {
2309
+ "epoch": 118.03,
2310
+ "learning_rate": 6.760825818018508e-05,
2311
+ "loss": 1.2957,
2312
+ "step": 181500
2313
+ },
2314
+ {
2315
+ "epoch": 118.35,
2316
+ "learning_rate": 6.711333532614168e-05,
2317
+ "loss": 1.316,
2318
+ "step": 182000
2319
+ },
2320
+ {
2321
+ "epoch": 118.68,
2322
+ "learning_rate": 6.661931407662292e-05,
2323
+ "loss": 1.3295,
2324
+ "step": 182500
2325
+ },
2326
+ {
2327
+ "epoch": 119.01,
2328
+ "learning_rate": 6.612620797547087e-05,
2329
+ "loss": 1.2891,
2330
+ "step": 183000
2331
+ },
2332
+ {
2333
+ "epoch": 119.33,
2334
+ "learning_rate": 6.56340305414384e-05,
2335
+ "loss": 1.3008,
2336
+ "step": 183500
2337
+ },
2338
+ {
2339
+ "epoch": 119.66,
2340
+ "learning_rate": 6.51427952678185e-05,
2341
+ "loss": 1.3158,
2342
+ "step": 184000
2343
+ },
2344
+ {
2345
+ "epoch": 119.98,
2346
+ "learning_rate": 6.465251562207431e-05,
2347
+ "loss": 1.3154,
2348
+ "step": 184500
2349
+ },
2350
+ {
2351
+ "epoch": 120.31,
2352
+ "learning_rate": 6.416320504546997e-05,
2353
+ "loss": 1.3166,
2354
+ "step": 185000
2355
+ },
2356
+ {
2357
+ "epoch": 120.63,
2358
+ "learning_rate": 6.367487695270217e-05,
2359
+ "loss": 1.3134,
2360
+ "step": 185500
2361
+ },
2362
+ {
2363
+ "epoch": 120.96,
2364
+ "learning_rate": 6.318754473153221e-05,
2365
+ "loss": 1.308,
2366
+ "step": 186000
2367
+ },
2368
+ {
2369
+ "epoch": 121.28,
2370
+ "learning_rate": 6.27012217424191e-05,
2371
+ "loss": 1.3229,
2372
+ "step": 186500
2373
+ },
2374
+ {
2375
+ "epoch": 121.61,
2376
+ "learning_rate": 6.22159213181533e-05,
2377
+ "loss": 1.2987,
2378
+ "step": 187000
2379
+ },
2380
+ {
2381
+ "epoch": 121.93,
2382
+ "learning_rate": 6.173165676349103e-05,
2383
+ "loss": 1.3054,
2384
+ "step": 187500
2385
+ },
2386
+ {
2387
+ "epoch": 122.26,
2388
+ "learning_rate": 6.12484413547897e-05,
2389
+ "loss": 1.2986,
2390
+ "step": 188000
2391
+ },
2392
+ {
2393
+ "epoch": 122.58,
2394
+ "learning_rate": 6.076628833964388e-05,
2395
+ "loss": 1.2911,
2396
+ "step": 188500
2397
+ },
2398
+ {
2399
+ "epoch": 122.91,
2400
+ "learning_rate": 6.0285210936521955e-05,
2401
+ "loss": 1.2844,
2402
+ "step": 189000
2403
+ },
2404
+ {
2405
+ "epoch": 123.23,
2406
+ "learning_rate": 5.9805222334404e-05,
2407
+ "loss": 1.2987,
2408
+ "step": 189500
2409
+ },
2410
+ {
2411
+ "epoch": 123.56,
2412
+ "learning_rate": 5.9326335692419995e-05,
2413
+ "loss": 1.325,
2414
+ "step": 190000
2415
+ },
2416
+ {
2417
+ "epoch": 123.56,
2418
+ "eval_runtime": 0.0012,
2419
+ "eval_samples_per_second": 0.0,
2420
+ "eval_steps_per_second": 0.0,
2421
+ "step": 190000
2422
+ },
2423
+ {
2424
+ "epoch": 123.88,
2425
+ "learning_rate": 5.884856413948913e-05,
2426
+ "loss": 1.3178,
2427
+ "step": 190500
2428
+ },
2429
+ {
2430
+ "epoch": 124.21,
2431
+ "learning_rate": 5.83719207739599e-05,
2432
+ "loss": 1.3022,
2433
+ "step": 191000
2434
+ },
2435
+ {
2436
+ "epoch": 124.53,
2437
+ "learning_rate": 5.789641866325091e-05,
2438
+ "loss": 1.3017,
2439
+ "step": 191500
2440
+ },
2441
+ {
2442
+ "epoch": 124.86,
2443
+ "learning_rate": 5.7422070843492734e-05,
2444
+ "loss": 1.2894,
2445
+ "step": 192000
2446
+ },
2447
+ {
2448
+ "epoch": 125.18,
2449
+ "learning_rate": 5.694889031917047e-05,
2450
+ "loss": 1.2886,
2451
+ "step": 192500
2452
+ },
2453
+ {
2454
+ "epoch": 125.51,
2455
+ "learning_rate": 5.647689006276726e-05,
2456
+ "loss": 1.3058,
2457
+ "step": 193000
2458
+ },
2459
+ {
2460
+ "epoch": 125.83,
2461
+ "learning_rate": 5.6006083014408484e-05,
2462
+ "loss": 1.3201,
2463
+ "step": 193500
2464
+ },
2465
+ {
2466
+ "epoch": 126.16,
2467
+ "learning_rate": 5.553648208150728e-05,
2468
+ "loss": 1.2908,
2469
+ "step": 194000
2470
+ },
2471
+ {
2472
+ "epoch": 126.48,
2473
+ "learning_rate": 5.506810013841036e-05,
2474
+ "loss": 1.2965,
2475
+ "step": 194500
2476
+ },
2477
+ {
2478
+ "epoch": 126.81,
2479
+ "learning_rate": 5.4600950026045326e-05,
2480
+ "loss": 1.296,
2481
+ "step": 195000
2482
+ },
2483
+ {
2484
+ "epoch": 127.13,
2485
+ "learning_rate": 5.4135044551568546e-05,
2486
+ "loss": 1.2668,
2487
+ "step": 195500
2488
+ },
2489
+ {
2490
+ "epoch": 127.46,
2491
+ "learning_rate": 5.3670396488013854e-05,
2492
+ "loss": 1.3238,
2493
+ "step": 196000
2494
+ },
2495
+ {
2496
+ "epoch": 127.78,
2497
+ "learning_rate": 5.320701857394268e-05,
2498
+ "loss": 1.3058,
2499
+ "step": 196500
2500
+ },
2501
+ {
2502
+ "epoch": 128.11,
2503
+ "learning_rate": 5.274492351309461e-05,
2504
+ "loss": 1.2915,
2505
+ "step": 197000
2506
+ },
2507
+ {
2508
+ "epoch": 128.43,
2509
+ "learning_rate": 5.2284123974039154e-05,
2510
+ "loss": 1.2827,
2511
+ "step": 197500
2512
+ },
2513
+ {
2514
+ "epoch": 128.76,
2515
+ "learning_rate": 5.182463258982846e-05,
2516
+ "loss": 1.2545,
2517
+ "step": 198000
2518
+ },
2519
+ {
2520
+ "epoch": 129.08,
2521
+ "learning_rate": 5.1366461957650954e-05,
2522
+ "loss": 1.3182,
2523
+ "step": 198500
2524
+ },
2525
+ {
2526
+ "epoch": 129.41,
2527
+ "learning_rate": 5.090962463848592e-05,
2528
+ "loss": 1.2939,
2529
+ "step": 199000
2530
+ },
2531
+ {
2532
+ "epoch": 129.74,
2533
+ "learning_rate": 5.045413315675924e-05,
2534
+ "loss": 1.3114,
2535
+ "step": 199500
2536
+ },
2537
+ {
2538
+ "epoch": 130.06,
2539
+ "learning_rate": 5.000000000000002e-05,
2540
+ "loss": 1.285,
2541
+ "step": 200000
2542
+ },
2543
+ {
2544
+ "epoch": 130.06,
2545
+ "eval_runtime": 0.0014,
2546
+ "eval_samples_per_second": 0.0,
2547
+ "eval_steps_per_second": 0.0,
2548
+ "step": 200000
2549
+ },
2550
+ {
2551
+ "epoch": 130.39,
2552
+ "learning_rate": 4.9547237618498085e-05,
2553
+ "loss": 1.2967,
2554
+ "step": 200500
2555
+ },
2556
+ {
2557
+ "epoch": 130.71,
2558
+ "learning_rate": 4.909585842496287e-05,
2559
+ "loss": 1.2901,
2560
+ "step": 201000
2561
+ },
2562
+ {
2563
+ "epoch": 131.04,
2564
+ "learning_rate": 4.864587479418302e-05,
2565
+ "loss": 1.2764,
2566
+ "step": 201500
2567
+ },
2568
+ {
2569
+ "epoch": 131.36,
2570
+ "learning_rate": 4.8197299062686995e-05,
2571
+ "loss": 1.2933,
2572
+ "step": 202000
2573
+ },
2574
+ {
2575
+ "epoch": 131.69,
2576
+ "learning_rate": 4.7750143528405126e-05,
2577
+ "loss": 1.3121,
2578
+ "step": 202500
2579
+ },
2580
+ {
2581
+ "epoch": 132.01,
2582
+ "learning_rate": 4.7304420450332244e-05,
2583
+ "loss": 1.2743,
2584
+ "step": 203000
2585
+ },
2586
+ {
2587
+ "epoch": 132.34,
2588
+ "learning_rate": 4.68601420481917e-05,
2589
+ "loss": 1.2615,
2590
+ "step": 203500
2591
+ },
2592
+ {
2593
+ "epoch": 132.66,
2594
+ "learning_rate": 4.6417320502100316e-05,
2595
+ "loss": 1.2778,
2596
+ "step": 204000
2597
+ },
2598
+ {
2599
+ "epoch": 132.99,
2600
+ "learning_rate": 4.59759679522345e-05,
2601
+ "loss": 1.2898,
2602
+ "step": 204500
2603
+ },
2604
+ {
2605
+ "epoch": 133.31,
2606
+ "learning_rate": 4.5536096498497295e-05,
2607
+ "loss": 1.2782,
2608
+ "step": 205000
2609
+ },
2610
+ {
2611
+ "epoch": 133.64,
2612
+ "learning_rate": 4.5097718200186814e-05,
2613
+ "loss": 1.2726,
2614
+ "step": 205500
2615
+ },
2616
+ {
2617
+ "epoch": 133.96,
2618
+ "learning_rate": 4.46608450756656e-05,
2619
+ "loss": 1.2616,
2620
+ "step": 206000
2621
+ },
2622
+ {
2623
+ "epoch": 134.29,
2624
+ "learning_rate": 4.4225489102030995e-05,
2625
+ "loss": 1.2849,
2626
+ "step": 206500
2627
+ },
2628
+ {
2629
+ "epoch": 134.61,
2630
+ "learning_rate": 4.379166221478697e-05,
2631
+ "loss": 1.253,
2632
+ "step": 207000
2633
+ },
2634
+ {
2635
+ "epoch": 134.94,
2636
+ "learning_rate": 4.335937630751674e-05,
2637
+ "loss": 1.2782,
2638
+ "step": 207500
2639
+ },
2640
+ {
2641
+ "epoch": 135.26,
2642
+ "learning_rate": 4.2928643231556844e-05,
2643
+ "loss": 1.2926,
2644
+ "step": 208000
2645
+ },
2646
+ {
2647
+ "epoch": 135.59,
2648
+ "learning_rate": 4.249947479567218e-05,
2649
+ "loss": 1.3252,
2650
+ "step": 208500
2651
+ },
2652
+ {
2653
+ "epoch": 135.91,
2654
+ "learning_rate": 4.207188276573214e-05,
2655
+ "loss": 1.3277,
2656
+ "step": 209000
2657
+ },
2658
+ {
2659
+ "epoch": 136.24,
2660
+ "learning_rate": 4.1645878864388266e-05,
2661
+ "loss": 1.269,
2662
+ "step": 209500
2663
+ },
2664
+ {
2665
+ "epoch": 136.56,
2666
+ "learning_rate": 4.12214747707527e-05,
2667
+ "loss": 1.322,
2668
+ "step": 210000
2669
+ },
2670
+ {
2671
+ "epoch": 136.56,
2672
+ "eval_runtime": 0.0013,
2673
+ "eval_samples_per_second": 0.0,
2674
+ "eval_steps_per_second": 0.0,
2675
+ "step": 210000
2676
+ },
2677
+ {
2678
+ "epoch": 136.89,
2679
+ "learning_rate": 4.0798682120078044e-05,
2680
+ "loss": 1.3141,
2681
+ "step": 210500
2682
+ },
2683
+ {
2684
+ "epoch": 137.21,
2685
+ "learning_rate": 4.037751250343841e-05,
2686
+ "loss": 1.2815,
2687
+ "step": 211000
2688
+ },
2689
+ {
2690
+ "epoch": 137.54,
2691
+ "learning_rate": 3.9957977467411615e-05,
2692
+ "loss": 1.2729,
2693
+ "step": 211500
2694
+ },
2695
+ {
2696
+ "epoch": 137.86,
2697
+ "learning_rate": 3.954008851376252e-05,
2698
+ "loss": 1.2726,
2699
+ "step": 212000
2700
+ },
2701
+ {
2702
+ "epoch": 138.19,
2703
+ "learning_rate": 3.9123857099127936e-05,
2704
+ "loss": 1.2877,
2705
+ "step": 212500
2706
+ },
2707
+ {
2708
+ "epoch": 138.51,
2709
+ "learning_rate": 3.8709294634702376e-05,
2710
+ "loss": 1.2954,
2711
+ "step": 213000
2712
+ },
2713
+ {
2714
+ "epoch": 138.84,
2715
+ "learning_rate": 3.829641248592515e-05,
2716
+ "loss": 1.2716,
2717
+ "step": 213500
2718
+ },
2719
+ {
2720
+ "epoch": 139.16,
2721
+ "learning_rate": 3.788522197216897e-05,
2722
+ "loss": 1.2603,
2723
+ "step": 214000
2724
+ },
2725
+ {
2726
+ "epoch": 139.49,
2727
+ "learning_rate": 3.747573436642951e-05,
2728
+ "loss": 1.2564,
2729
+ "step": 214500
2730
+ },
2731
+ {
2732
+ "epoch": 139.81,
2733
+ "learning_rate": 3.7067960895016275e-05,
2734
+ "loss": 1.2978,
2735
+ "step": 215000
2736
+ },
2737
+ {
2738
+ "epoch": 140.14,
2739
+ "learning_rate": 3.6661912737245e-05,
2740
+ "loss": 1.2693,
2741
+ "step": 215500
2742
+ },
2743
+ {
2744
+ "epoch": 140.46,
2745
+ "learning_rate": 3.6257601025131026e-05,
2746
+ "loss": 1.2827,
2747
+ "step": 216000
2748
+ },
2749
+ {
2750
+ "epoch": 140.79,
2751
+ "learning_rate": 3.585503684308421e-05,
2752
+ "loss": 1.291,
2753
+ "step": 216500
2754
+ },
2755
+ {
2756
+ "epoch": 141.12,
2757
+ "learning_rate": 3.545423122760493e-05,
2758
+ "loss": 1.2743,
2759
+ "step": 217000
2760
+ },
2761
+ {
2762
+ "epoch": 141.44,
2763
+ "learning_rate": 3.5055195166981645e-05,
2764
+ "loss": 1.2899,
2765
+ "step": 217500
2766
+ },
2767
+ {
2768
+ "epoch": 141.77,
2769
+ "learning_rate": 3.465793960098945e-05,
2770
+ "loss": 1.2654,
2771
+ "step": 218000
2772
+ },
2773
+ {
2774
+ "epoch": 142.09,
2775
+ "learning_rate": 3.426247542059041e-05,
2776
+ "loss": 1.2695,
2777
+ "step": 218500
2778
+ },
2779
+ {
2780
+ "epoch": 142.42,
2781
+ "learning_rate": 3.386881346763483e-05,
2782
+ "loss": 1.2615,
2783
+ "step": 219000
2784
+ },
2785
+ {
2786
+ "epoch": 142.74,
2787
+ "learning_rate": 3.347696453456393e-05,
2788
+ "loss": 1.2506,
2789
+ "step": 219500
2790
+ },
2791
+ {
2792
+ "epoch": 143.07,
2793
+ "learning_rate": 3.308693936411421e-05,
2794
+ "loss": 1.2812,
2795
+ "step": 220000
2796
+ },
2797
+ {
2798
+ "epoch": 143.07,
2799
+ "eval_runtime": 0.0012,
2800
+ "eval_samples_per_second": 0.0,
2801
+ "eval_steps_per_second": 0.0,
2802
+ "step": 220000
2803
+ },
2804
+ {
2805
+ "epoch": 143.39,
2806
+ "learning_rate": 3.269874864902269e-05,
2807
+ "loss": 1.2625,
2808
+ "step": 220500
2809
+ },
2810
+ {
2811
+ "epoch": 143.72,
2812
+ "learning_rate": 3.231240303173394e-05,
2813
+ "loss": 1.2722,
2814
+ "step": 221000
2815
+ },
2816
+ {
2817
+ "epoch": 144.04,
2818
+ "learning_rate": 3.192791310410822e-05,
2819
+ "loss": 1.278,
2820
+ "step": 221500
2821
+ },
2822
+ {
2823
+ "epoch": 144.37,
2824
+ "learning_rate": 3.154528940713113e-05,
2825
+ "loss": 1.3162,
2826
+ "step": 222000
2827
+ },
2828
+ {
2829
+ "epoch": 144.69,
2830
+ "learning_rate": 3.116454243062459e-05,
2831
+ "loss": 1.2748,
2832
+ "step": 222500
2833
+ },
2834
+ {
2835
+ "epoch": 145.02,
2836
+ "learning_rate": 3.078568261295933e-05,
2837
+ "loss": 1.2756,
2838
+ "step": 223000
2839
+ },
2840
+ {
2841
+ "epoch": 145.34,
2842
+ "learning_rate": 3.0408720340768572e-05,
2843
+ "loss": 1.2389,
2844
+ "step": 223500
2845
+ },
2846
+ {
2847
+ "epoch": 145.67,
2848
+ "learning_rate": 3.0033665948663448e-05,
2849
+ "loss": 1.259,
2850
+ "step": 224000
2851
+ },
2852
+ {
2853
+ "epoch": 145.99,
2854
+ "learning_rate": 2.9660529718949627e-05,
2855
+ "loss": 1.2801,
2856
+ "step": 224500
2857
+ },
2858
+ {
2859
+ "epoch": 146.32,
2860
+ "learning_rate": 2.9289321881345254e-05,
2861
+ "loss": 1.2844,
2862
+ "step": 225000
2863
+ },
2864
+ {
2865
+ "epoch": 146.64,
2866
+ "learning_rate": 2.8920052612700754e-05,
2867
+ "loss": 1.3067,
2868
+ "step": 225500
2869
+ },
2870
+ {
2871
+ "epoch": 146.97,
2872
+ "learning_rate": 2.8552732036719687e-05,
2873
+ "loss": 1.2596,
2874
+ "step": 226000
2875
+ },
2876
+ {
2877
+ "epoch": 147.29,
2878
+ "learning_rate": 2.8187370223681132e-05,
2879
+ "loss": 1.2737,
2880
+ "step": 226500
2881
+ },
2882
+ {
2883
+ "epoch": 147.62,
2884
+ "learning_rate": 2.7823977190163786e-05,
2885
+ "loss": 1.2689,
2886
+ "step": 227000
2887
+ },
2888
+ {
2889
+ "epoch": 147.94,
2890
+ "learning_rate": 2.746256289877126e-05,
2891
+ "loss": 1.2418,
2892
+ "step": 227500
2893
+ },
2894
+ {
2895
+ "epoch": 148.27,
2896
+ "learning_rate": 2.7103137257858868e-05,
2897
+ "loss": 1.2921,
2898
+ "step": 228000
2899
+ },
2900
+ {
2901
+ "epoch": 148.59,
2902
+ "learning_rate": 2.6745710121262136e-05,
2903
+ "loss": 1.2561,
2904
+ "step": 228500
2905
+ },
2906
+ {
2907
+ "epoch": 148.92,
2908
+ "learning_rate": 2.639029128802657e-05,
2909
+ "loss": 1.2658,
2910
+ "step": 229000
2911
+ },
2912
+ {
2913
+ "epoch": 149.24,
2914
+ "learning_rate": 2.603689050213902e-05,
2915
+ "loss": 1.2722,
2916
+ "step": 229500
2917
+ },
2918
+ {
2919
+ "epoch": 149.57,
2920
+ "learning_rate": 2.5685517452260567e-05,
2921
+ "loss": 1.2175,
2922
+ "step": 230000
2923
+ },
2924
+ {
2925
+ "epoch": 149.57,
2926
+ "eval_runtime": 0.0012,
2927
+ "eval_samples_per_second": 0.0,
2928
+ "eval_steps_per_second": 0.0,
2929
+ "step": 230000
2930
+ },
2931
+ {
2932
+ "epoch": 149.89,
2933
+ "learning_rate": 2.5336181771460876e-05,
2934
+ "loss": 1.2331,
2935
+ "step": 230500
2936
+ },
2937
+ {
2938
+ "epoch": 150.22,
2939
+ "learning_rate": 2.4988893036954043e-05,
2940
+ "loss": 1.2488,
2941
+ "step": 231000
2942
+ },
2943
+ {
2944
+ "epoch": 150.54,
2945
+ "learning_rate": 2.464366076983623e-05,
2946
+ "loss": 1.2598,
2947
+ "step": 231500
2948
+ },
2949
+ {
2950
+ "epoch": 150.87,
2951
+ "learning_rate": 2.4300494434824373e-05,
2952
+ "loss": 1.2805,
2953
+ "step": 232000
2954
+ },
2955
+ {
2956
+ "epoch": 151.19,
2957
+ "learning_rate": 2.3959403439996907e-05,
2958
+ "loss": 1.2435,
2959
+ "step": 232500
2960
+ },
2961
+ {
2962
+ "epoch": 151.52,
2963
+ "learning_rate": 2.362039713653581e-05,
2964
+ "loss": 1.2641,
2965
+ "step": 233000
2966
+ },
2967
+ {
2968
+ "epoch": 151.85,
2969
+ "learning_rate": 2.328348481847006e-05,
2970
+ "loss": 1.2625,
2971
+ "step": 233500
2972
+ },
2973
+ {
2974
+ "epoch": 152.17,
2975
+ "learning_rate": 2.2948675722421086e-05,
2976
+ "loss": 1.2149,
2977
+ "step": 234000
2978
+ },
2979
+ {
2980
+ "epoch": 152.5,
2981
+ "learning_rate": 2.2615979027349387e-05,
2982
+ "loss": 1.2519,
2983
+ "step": 234500
2984
+ },
2985
+ {
2986
+ "epoch": 152.82,
2987
+ "learning_rate": 2.2285403854302912e-05,
2988
+ "loss": 1.2497,
2989
+ "step": 235000
2990
+ },
2991
+ {
2992
+ "epoch": 153.15,
2993
+ "learning_rate": 2.195695926616702e-05,
2994
+ "loss": 1.2638,
2995
+ "step": 235500
2996
+ },
2997
+ {
2998
+ "epoch": 153.47,
2999
+ "learning_rate": 2.163065426741603e-05,
3000
+ "loss": 1.2347,
3001
+ "step": 236000
3002
+ },
3003
+ {
3004
+ "epoch": 153.8,
3005
+ "learning_rate": 2.1306497803866277e-05,
3006
+ "loss": 1.2572,
3007
+ "step": 236500
3008
+ },
3009
+ {
3010
+ "epoch": 154.12,
3011
+ "learning_rate": 2.098449876243096e-05,
3012
+ "loss": 1.2383,
3013
+ "step": 237000
3014
+ },
3015
+ {
3016
+ "epoch": 154.45,
3017
+ "learning_rate": 2.0664665970876496e-05,
3018
+ "loss": 1.2517,
3019
+ "step": 237500
3020
+ },
3021
+ {
3022
+ "epoch": 154.77,
3023
+ "learning_rate": 2.0347008197580374e-05,
3024
+ "loss": 1.2381,
3025
+ "step": 238000
3026
+ },
3027
+ {
3028
+ "epoch": 155.1,
3029
+ "learning_rate": 2.0031534151290943e-05,
3030
+ "loss": 1.2762,
3031
+ "step": 238500
3032
+ },
3033
+ {
3034
+ "epoch": 155.42,
3035
+ "learning_rate": 1.9718252480888566e-05,
3036
+ "loss": 1.2562,
3037
+ "step": 239000
3038
+ },
3039
+ {
3040
+ "epoch": 155.75,
3041
+ "learning_rate": 1.9407171775148436e-05,
3042
+ "loss": 1.2569,
3043
+ "step": 239500
3044
+ },
3045
+ {
3046
+ "epoch": 156.07,
3047
+ "learning_rate": 1.9098300562505266e-05,
3048
+ "loss": 1.2424,
3049
+ "step": 240000
3050
+ },
3051
+ {
3052
+ "epoch": 156.07,
3053
+ "eval_runtime": 0.0012,
3054
+ "eval_samples_per_second": 0.0,
3055
+ "eval_steps_per_second": 0.0,
3056
+ "step": 240000
3057
+ },
3058
+ {
3059
+ "epoch": 156.4,
3060
+ "learning_rate": 1.879164731081937e-05,
3061
+ "loss": 1.2454,
3062
+ "step": 240500
3063
+ },
3064
+ {
3065
+ "epoch": 156.72,
3066
+ "learning_rate": 1.848722042714457e-05,
3067
+ "loss": 1.2628,
3068
+ "step": 241000
3069
+ },
3070
+ {
3071
+ "epoch": 157.05,
3072
+ "learning_rate": 1.818502825749764e-05,
3073
+ "loss": 1.2726,
3074
+ "step": 241500
3075
+ },
3076
+ {
3077
+ "epoch": 157.37,
3078
+ "learning_rate": 1.78850790866296e-05,
3079
+ "loss": 1.2418,
3080
+ "step": 242000
3081
+ },
3082
+ {
3083
+ "epoch": 157.7,
3084
+ "learning_rate": 1.7587381137798432e-05,
3085
+ "loss": 1.27,
3086
+ "step": 242500
3087
+ },
3088
+ {
3089
+ "epoch": 158.02,
3090
+ "learning_rate": 1.7291942572543807e-05,
3091
+ "loss": 1.2433,
3092
+ "step": 243000
3093
+ },
3094
+ {
3095
+ "epoch": 158.35,
3096
+ "learning_rate": 1.6998771490463262e-05,
3097
+ "loss": 1.2645,
3098
+ "step": 243500
3099
+ },
3100
+ {
3101
+ "epoch": 158.67,
3102
+ "learning_rate": 1.6707875928990058e-05,
3103
+ "loss": 1.2575,
3104
+ "step": 244000
3105
+ },
3106
+ {
3107
+ "epoch": 159.0,
3108
+ "learning_rate": 1.6419263863172997e-05,
3109
+ "loss": 1.2702,
3110
+ "step": 244500
3111
+ },
3112
+ {
3113
+ "epoch": 159.32,
3114
+ "learning_rate": 1.6132943205457606e-05,
3115
+ "loss": 1.254,
3116
+ "step": 245000
3117
+ },
3118
+ {
3119
+ "epoch": 159.65,
3120
+ "learning_rate": 1.5848921805469397e-05,
3121
+ "loss": 1.2707,
3122
+ "step": 245500
3123
+ },
3124
+ {
3125
+ "epoch": 159.97,
3126
+ "learning_rate": 1.5567207449798515e-05,
3127
+ "loss": 1.2871,
3128
+ "step": 246000
3129
+ },
3130
+ {
3131
+ "epoch": 160.3,
3132
+ "learning_rate": 1.528780786178631e-05,
3133
+ "loss": 1.2611,
3134
+ "step": 246500
3135
+ },
3136
+ {
3137
+ "epoch": 160.62,
3138
+ "learning_rate": 1.5010730701313625e-05,
3139
+ "loss": 1.2422,
3140
+ "step": 247000
3141
+ },
3142
+ {
3143
+ "epoch": 160.95,
3144
+ "learning_rate": 1.4735983564590783e-05,
3145
+ "loss": 1.2355,
3146
+ "step": 247500
3147
+ },
3148
+ {
3149
+ "epoch": 161.27,
3150
+ "learning_rate": 1.4463573983949341e-05,
3151
+ "loss": 1.2152,
3152
+ "step": 248000
3153
+ },
3154
+ {
3155
+ "epoch": 161.6,
3156
+ "learning_rate": 1.4193509427635543e-05,
3157
+ "loss": 1.2499,
3158
+ "step": 248500
3159
+ },
3160
+ {
3161
+ "epoch": 161.92,
3162
+ "learning_rate": 1.3925797299605647e-05,
3163
+ "loss": 1.2226,
3164
+ "step": 249000
3165
+ },
3166
+ {
3167
+ "epoch": 162.25,
3168
+ "learning_rate": 1.3660444939322836e-05,
3169
+ "loss": 1.2423,
3170
+ "step": 249500
3171
+ },
3172
+ {
3173
+ "epoch": 162.58,
3174
+ "learning_rate": 1.339745962155613e-05,
3175
+ "loss": 1.2512,
3176
+ "step": 250000
3177
+ },
3178
+ {
3179
+ "epoch": 162.58,
3180
+ "eval_runtime": 0.0013,
3181
+ "eval_samples_per_second": 0.0,
3182
+ "eval_steps_per_second": 0.0,
3183
+ "step": 250000
3184
+ },
3185
+ {
3186
+ "epoch": 162.9,
3187
+ "learning_rate": 1.3136848556180892e-05,
3188
+ "loss": 1.2637,
3189
+ "step": 250500
3190
+ },
3191
+ {
3192
+ "epoch": 163.23,
3193
+ "learning_rate": 1.2878618887981064e-05,
3194
+ "loss": 1.2572,
3195
+ "step": 251000
3196
+ },
3197
+ {
3198
+ "epoch": 163.55,
3199
+ "learning_rate": 1.262277769645348e-05,
3200
+ "loss": 1.2812,
3201
+ "step": 251500
3202
+ },
3203
+ {
3204
+ "epoch": 163.88,
3205
+ "learning_rate": 1.2369331995613665e-05,
3206
+ "loss": 1.2265,
3207
+ "step": 252000
3208
+ },
3209
+ {
3210
+ "epoch": 164.2,
3211
+ "learning_rate": 1.2118288733803473e-05,
3212
+ "loss": 1.2682,
3213
+ "step": 252500
3214
+ },
3215
+ {
3216
+ "epoch": 164.53,
3217
+ "learning_rate": 1.1869654793500784e-05,
3218
+ "loss": 1.2759,
3219
+ "step": 253000
3220
+ },
3221
+ {
3222
+ "epoch": 164.85,
3223
+ "learning_rate": 1.1623436991130654e-05,
3224
+ "loss": 1.2526,
3225
+ "step": 253500
3226
+ },
3227
+ {
3228
+ "epoch": 165.18,
3229
+ "learning_rate": 1.1379642076878527e-05,
3230
+ "loss": 1.2597,
3231
+ "step": 254000
3232
+ },
3233
+ {
3234
+ "epoch": 165.5,
3235
+ "learning_rate": 1.1138276734505104e-05,
3236
+ "loss": 1.203,
3237
+ "step": 254500
3238
+ },
3239
+ {
3240
+ "epoch": 165.83,
3241
+ "learning_rate": 1.0899347581163221e-05,
3242
+ "loss": 1.2535,
3243
+ "step": 255000
3244
+ },
3245
+ {
3246
+ "epoch": 166.15,
3247
+ "learning_rate": 1.0662861167216243e-05,
3248
+ "loss": 1.2394,
3249
+ "step": 255500
3250
+ },
3251
+ {
3252
+ "epoch": 166.48,
3253
+ "learning_rate": 1.042882397605871e-05,
3254
+ "loss": 1.2433,
3255
+ "step": 256000
3256
+ },
3257
+ {
3258
+ "epoch": 166.8,
3259
+ "learning_rate": 1.0197242423938446e-05,
3260
+ "loss": 1.2433,
3261
+ "step": 256500
3262
+ },
3263
+ {
3264
+ "epoch": 167.13,
3265
+ "learning_rate": 9.968122859780648e-06,
3266
+ "loss": 1.2309,
3267
+ "step": 257000
3268
+ },
3269
+ {
3270
+ "epoch": 167.45,
3271
+ "learning_rate": 9.74147156501396e-06,
3272
+ "loss": 1.2279,
3273
+ "step": 257500
3274
+ },
3275
+ {
3276
+ "epoch": 167.78,
3277
+ "learning_rate": 9.517294753398064e-06,
3278
+ "loss": 1.2313,
3279
+ "step": 258000
3280
+ },
3281
+ {
3282
+ "epoch": 168.1,
3283
+ "learning_rate": 9.295598570853514e-06,
3284
+ "loss": 1.2593,
3285
+ "step": 258500
3286
+ },
3287
+ {
3288
+ "epoch": 168.43,
3289
+ "learning_rate": 9.076389095293148e-06,
3290
+ "loss": 1.2374,
3291
+ "step": 259000
3292
+ },
3293
+ {
3294
+ "epoch": 168.75,
3295
+ "learning_rate": 8.85967233645547e-06,
3296
+ "loss": 1.2336,
3297
+ "step": 259500
3298
+ },
3299
+ {
3300
+ "epoch": 169.08,
3301
+ "learning_rate": 8.645454235739903e-06,
3302
+ "loss": 1.2567,
3303
+ "step": 260000
3304
+ },
3305
+ {
3306
+ "epoch": 169.08,
3307
+ "eval_runtime": 0.0013,
3308
+ "eval_samples_per_second": 0.0,
3309
+ "eval_steps_per_second": 0.0,
3310
+ "step": 260000
3311
+ },
3312
+ {
3313
+ "epoch": 169.4,
3314
+ "learning_rate": 8.433740666043898e-06,
3315
+ "loss": 1.2084,
3316
+ "step": 260500
3317
+ },
3318
+ {
3319
+ "epoch": 169.73,
3320
+ "learning_rate": 8.224537431601886e-06,
3321
+ "loss": 1.239,
3322
+ "step": 261000
3323
+ },
3324
+ {
3325
+ "epoch": 170.05,
3326
+ "learning_rate": 8.017850267826232e-06,
3327
+ "loss": 1.2482,
3328
+ "step": 261500
3329
+ },
3330
+ {
3331
+ "epoch": 170.38,
3332
+ "learning_rate": 7.81368484114996e-06,
3333
+ "loss": 1.2488,
3334
+ "step": 262000
3335
+ },
3336
+ {
3337
+ "epoch": 170.7,
3338
+ "learning_rate": 7.612046748871327e-06,
3339
+ "loss": 1.2325,
3340
+ "step": 262500
3341
+ },
3342
+ {
3343
+ "epoch": 171.03,
3344
+ "learning_rate": 7.412941519000527e-06,
3345
+ "loss": 1.2575,
3346
+ "step": 263000
3347
+ },
3348
+ {
3349
+ "epoch": 171.35,
3350
+ "learning_rate": 7.216374610108012e-06,
3351
+ "loss": 1.2599,
3352
+ "step": 263500
3353
+ },
3354
+ {
3355
+ "epoch": 171.68,
3356
+ "learning_rate": 7.022351411174866e-06,
3357
+ "loss": 1.2562,
3358
+ "step": 264000
3359
+ },
3360
+ {
3361
+ "epoch": 172.0,
3362
+ "learning_rate": 6.830877241445111e-06,
3363
+ "loss": 1.2147,
3364
+ "step": 264500
3365
+ },
3366
+ {
3367
+ "epoch": 172.33,
3368
+ "learning_rate": 6.6419573502798374e-06,
3369
+ "loss": 1.255,
3370
+ "step": 265000
3371
+ },
3372
+ {
3373
+ "epoch": 172.65,
3374
+ "learning_rate": 6.455596917013273e-06,
3375
+ "loss": 1.2347,
3376
+ "step": 265500
3377
+ },
3378
+ {
3379
+ "epoch": 172.98,
3380
+ "learning_rate": 6.2718010508108545e-06,
3381
+ "loss": 1.2591,
3382
+ "step": 266000
3383
+ },
3384
+ {
3385
+ "epoch": 173.31,
3386
+ "learning_rate": 6.090574790529091e-06,
3387
+ "loss": 1.2623,
3388
+ "step": 266500
3389
+ },
3390
+ {
3391
+ "epoch": 173.63,
3392
+ "learning_rate": 5.911923104577455e-06,
3393
+ "loss": 1.2462,
3394
+ "step": 267000
3395
+ },
3396
+ {
3397
+ "epoch": 173.96,
3398
+ "learning_rate": 5.735850890782157e-06,
3399
+ "loss": 1.2318,
3400
+ "step": 267500
3401
+ },
3402
+ {
3403
+ "epoch": 174.28,
3404
+ "learning_rate": 5.562362976251901e-06,
3405
+ "loss": 1.2481,
3406
+ "step": 268000
3407
+ },
3408
+ {
3409
+ "epoch": 174.61,
3410
+ "learning_rate": 5.39146411724547e-06,
3411
+ "loss": 1.2649,
3412
+ "step": 268500
3413
+ },
3414
+ {
3415
+ "epoch": 174.93,
3416
+ "learning_rate": 5.223158999041444e-06,
3417
+ "loss": 1.2464,
3418
+ "step": 269000
3419
+ },
3420
+ {
3421
+ "epoch": 175.26,
3422
+ "learning_rate": 5.057452235809624e-06,
3423
+ "loss": 1.2538,
3424
+ "step": 269500
3425
+ },
3426
+ {
3427
+ "epoch": 175.58,
3428
+ "learning_rate": 4.8943483704846475e-06,
3429
+ "loss": 1.2532,
3430
+ "step": 270000
3431
+ },
3432
+ {
3433
+ "epoch": 175.58,
3434
+ "eval_runtime": 0.0012,
3435
+ "eval_samples_per_second": 0.0,
3436
+ "eval_steps_per_second": 0.0,
3437
+ "step": 270000
3438
+ },
3439
+ {
3440
+ "epoch": 175.91,
3441
+ "learning_rate": 4.733851874641382e-06,
3442
+ "loss": 1.256,
3443
+ "step": 270500
3444
+ },
3445
+ {
3446
+ "epoch": 176.23,
3447
+ "learning_rate": 4.575967148372317e-06,
3448
+ "loss": 1.2083,
3449
+ "step": 271000
3450
+ },
3451
+ {
3452
+ "epoch": 176.56,
3453
+ "learning_rate": 4.420698520166988e-06,
3454
+ "loss": 1.2336,
3455
+ "step": 271500
3456
+ },
3457
+ {
3458
+ "epoch": 176.88,
3459
+ "learning_rate": 4.268050246793276e-06,
3460
+ "loss": 1.2793,
3461
+ "step": 272000
3462
+ },
3463
+ {
3464
+ "epoch": 177.21,
3465
+ "learning_rate": 4.118026513180695e-06,
3466
+ "loss": 1.2365,
3467
+ "step": 272500
3468
+ },
3469
+ {
3470
+ "epoch": 177.53,
3471
+ "learning_rate": 3.970631432305694e-06,
3472
+ "loss": 1.2496,
3473
+ "step": 273000
3474
+ },
3475
+ {
3476
+ "epoch": 177.86,
3477
+ "learning_rate": 3.825869045078867e-06,
3478
+ "loss": 1.2603,
3479
+ "step": 273500
3480
+ },
3481
+ {
3482
+ "epoch": 178.18,
3483
+ "learning_rate": 3.68374332023419e-06,
3484
+ "loss": 1.2342,
3485
+ "step": 274000
3486
+ },
3487
+ {
3488
+ "epoch": 178.51,
3489
+ "learning_rate": 3.5442581542201923e-06,
3490
+ "loss": 1.2548,
3491
+ "step": 274500
3492
+ },
3493
+ {
3494
+ "epoch": 178.83,
3495
+ "learning_rate": 3.40741737109318e-06,
3496
+ "loss": 1.2268,
3497
+ "step": 275000
3498
+ },
3499
+ {
3500
+ "epoch": 179.16,
3501
+ "learning_rate": 3.273224722412327e-06,
3502
+ "loss": 1.2424,
3503
+ "step": 275500
3504
+ },
3505
+ {
3506
+ "epoch": 179.48,
3507
+ "learning_rate": 3.1416838871368924e-06,
3508
+ "loss": 1.2274,
3509
+ "step": 276000
3510
+ },
3511
+ {
3512
+ "epoch": 179.81,
3513
+ "learning_rate": 3.012798471525324e-06,
3514
+ "loss": 1.251,
3515
+ "step": 276500
3516
+ },
3517
+ {
3518
+ "epoch": 180.13,
3519
+ "learning_rate": 2.8865720090364034e-06,
3520
+ "loss": 1.2266,
3521
+ "step": 277000
3522
+ },
3523
+ {
3524
+ "epoch": 180.46,
3525
+ "learning_rate": 2.7630079602323442e-06,
3526
+ "loss": 1.2417,
3527
+ "step": 277500
3528
+ },
3529
+ {
3530
+ "epoch": 180.78,
3531
+ "learning_rate": 2.6421097126839712e-06,
3532
+ "loss": 1.2536,
3533
+ "step": 278000
3534
+ },
3535
+ {
3536
+ "epoch": 181.11,
3537
+ "learning_rate": 2.5238805808778242e-06,
3538
+ "loss": 1.2233,
3539
+ "step": 278500
3540
+ },
3541
+ {
3542
+ "epoch": 181.43,
3543
+ "learning_rate": 2.4083238061252567e-06,
3544
+ "loss": 1.2278,
3545
+ "step": 279000
3546
+ },
3547
+ {
3548
+ "epoch": 181.76,
3549
+ "learning_rate": 2.295442556473637e-06,
3550
+ "loss": 1.2092,
3551
+ "step": 279500
3552
+ },
3553
+ {
3554
+ "epoch": 182.08,
3555
+ "learning_rate": 2.1852399266194314e-06,
3556
+ "loss": 1.2115,
3557
+ "step": 280000
3558
+ },
3559
+ {
3560
+ "epoch": 182.08,
3561
+ "eval_runtime": 0.0012,
3562
+ "eval_samples_per_second": 0.0,
3563
+ "eval_steps_per_second": 0.0,
3564
+ "step": 280000
3565
+ },
3566
+ {
3567
+ "epoch": 182.41,
3568
+ "learning_rate": 2.0777189378234143e-06,
3569
+ "loss": 1.2362,
3570
+ "step": 280500
3571
+ },
3572
+ {
3573
+ "epoch": 182.73,
3574
+ "learning_rate": 1.9728825378278246e-06,
3575
+ "loss": 1.1964,
3576
+ "step": 281000
3577
+ },
3578
+ {
3579
+ "epoch": 183.06,
3580
+ "learning_rate": 1.8707336007754873e-06,
3581
+ "loss": 1.2569,
3582
+ "step": 281500
3583
+ },
3584
+ {
3585
+ "epoch": 183.38,
3586
+ "learning_rate": 1.771274927131139e-06,
3587
+ "loss": 1.2593,
3588
+ "step": 282000
3589
+ },
3590
+ {
3591
+ "epoch": 183.71,
3592
+ "learning_rate": 1.6745092436045494e-06,
3593
+ "loss": 1.2745,
3594
+ "step": 282500
3595
+ },
3596
+ {
3597
+ "epoch": 184.04,
3598
+ "learning_rate": 1.580439203075812e-06,
3599
+ "loss": 1.2483,
3600
+ "step": 283000
3601
+ },
3602
+ {
3603
+ "epoch": 184.36,
3604
+ "learning_rate": 1.4890673845226133e-06,
3605
+ "loss": 1.2369,
3606
+ "step": 283500
3607
+ },
3608
+ {
3609
+ "epoch": 184.69,
3610
+ "learning_rate": 1.400396292949513e-06,
3611
+ "loss": 1.2195,
3612
+ "step": 284000
3613
+ },
3614
+ {
3615
+ "epoch": 185.01,
3616
+ "learning_rate": 1.3144283593192752e-06,
3617
+ "loss": 1.236,
3618
+ "step": 284500
3619
+ },
3620
+ {
3621
+ "epoch": 185.34,
3622
+ "learning_rate": 1.231165940486234e-06,
3623
+ "loss": 1.2671,
3624
+ "step": 285000
3625
+ },
3626
+ {
3627
+ "epoch": 185.66,
3628
+ "learning_rate": 1.1506113191316447e-06,
3629
+ "loss": 1.226,
3630
+ "step": 285500
3631
+ },
3632
+ {
3633
+ "epoch": 185.99,
3634
+ "learning_rate": 1.0727667037011668e-06,
3635
+ "loss": 1.2473,
3636
+ "step": 286000
3637
+ },
3638
+ {
3639
+ "epoch": 186.31,
3640
+ "learning_rate": 9.976342283442463e-07,
3641
+ "loss": 1.2191,
3642
+ "step": 286500
3643
+ },
3644
+ {
3645
+ "epoch": 186.64,
3646
+ "learning_rate": 9.252159528556403e-07,
3647
+ "loss": 1.2208,
3648
+ "step": 287000
3649
+ },
3650
+ {
3651
+ "epoch": 186.96,
3652
+ "learning_rate": 8.555138626189618e-07,
3653
+ "loss": 1.2619,
3654
+ "step": 287500
3655
+ },
3656
+ {
3657
+ "epoch": 187.29,
3658
+ "learning_rate": 7.885298685522235e-07,
3659
+ "loss": 1.2389,
3660
+ "step": 288000
3661
+ },
3662
+ {
3663
+ "epoch": 187.61,
3664
+ "learning_rate": 7.242658070554464e-07,
3665
+ "loss": 1.244,
3666
+ "step": 288500
3667
+ },
3668
+ {
3669
+ "epoch": 187.94,
3670
+ "learning_rate": 6.627234399603555e-07,
3671
+ "loss": 1.238,
3672
+ "step": 289000
3673
+ },
3674
+ {
3675
+ "epoch": 188.26,
3676
+ "learning_rate": 6.039044544820404e-07,
3677
+ "loss": 1.2443,
3678
+ "step": 289500
3679
+ },
3680
+ {
3681
+ "epoch": 188.59,
3682
+ "learning_rate": 5.478104631726711e-07,
3683
+ "loss": 1.2178,
3684
+ "step": 290000
3685
+ },
3686
+ {
3687
+ "epoch": 188.59,
3688
+ "eval_runtime": 0.0013,
3689
+ "eval_samples_per_second": 0.0,
3690
+ "eval_steps_per_second": 0.0,
3691
+ "step": 290000
3692
+ },
3693
+ {
3694
+ "epoch": 188.91,
3695
+ "learning_rate": 4.944430038773762e-07,
3696
+ "loss": 1.2326,
3697
+ "step": 290500
3698
+ },
3699
+ {
3700
+ "epoch": 189.24,
3701
+ "learning_rate": 4.438035396920004e-07,
3702
+ "loss": 1.2231,
3703
+ "step": 291000
3704
+ },
3705
+ {
3706
+ "epoch": 189.56,
3707
+ "learning_rate": 3.958934589230467e-07,
3708
+ "loss": 1.2526,
3709
+ "step": 291500
3710
+ },
3711
+ {
3712
+ "epoch": 189.89,
3713
+ "learning_rate": 3.50714075049563e-07,
3714
+ "loss": 1.2415,
3715
+ "step": 292000
3716
+ },
3717
+ {
3718
+ "epoch": 190.21,
3719
+ "learning_rate": 3.0826662668720364e-07,
3720
+ "loss": 1.2414,
3721
+ "step": 292500
3722
+ },
3723
+ {
3724
+ "epoch": 190.54,
3725
+ "learning_rate": 2.685522775541904e-07,
3726
+ "loss": 1.2515,
3727
+ "step": 293000
3728
+ },
3729
+ {
3730
+ "epoch": 190.86,
3731
+ "learning_rate": 2.315721164394713e-07,
3732
+ "loss": 1.238,
3733
+ "step": 293500
3734
+ },
3735
+ {
3736
+ "epoch": 191.19,
3737
+ "learning_rate": 1.973271571728441e-07,
3738
+ "loss": 1.2693,
3739
+ "step": 294000
3740
+ },
3741
+ {
3742
+ "epoch": 191.51,
3743
+ "learning_rate": 1.6581833859716788e-07,
3744
+ "loss": 1.2387,
3745
+ "step": 294500
3746
+ },
3747
+ {
3748
+ "epoch": 191.84,
3749
+ "learning_rate": 1.3704652454261668e-07,
3750
+ "loss": 1.2411,
3751
+ "step": 295000
3752
+ },
3753
+ {
3754
+ "epoch": 192.16,
3755
+ "learning_rate": 1.1101250380300965e-07,
3756
+ "loss": 1.2521,
3757
+ "step": 295500
3758
+ },
3759
+ {
3760
+ "epoch": 192.49,
3761
+ "learning_rate": 8.771699011416168e-08,
3762
+ "loss": 1.2672,
3763
+ "step": 296000
3764
+ },
3765
+ {
3766
+ "epoch": 192.81,
3767
+ "learning_rate": 6.71606221343768e-08,
3768
+ "loss": 1.259,
3769
+ "step": 296500
3770
+ },
3771
+ {
3772
+ "epoch": 193.14,
3773
+ "learning_rate": 4.934396342684e-08,
3774
+ "loss": 1.2573,
3775
+ "step": 297000
3776
+ },
3777
+ {
3778
+ "epoch": 193.46,
3779
+ "learning_rate": 3.4267502444274015e-08,
3780
+ "loss": 1.228,
3781
+ "step": 297500
3782
+ },
3783
+ {
3784
+ "epoch": 193.79,
3785
+ "learning_rate": 2.193165251545004e-08,
3786
+ "loss": 1.2397,
3787
+ "step": 298000
3788
+ },
3789
+ {
3790
+ "epoch": 194.11,
3791
+ "learning_rate": 1.2336751833941229e-08,
3792
+ "loss": 1.2125,
3793
+ "step": 298500
3794
+ },
3795
+ {
3796
+ "epoch": 194.44,
3797
+ "learning_rate": 5.483063448785686e-09,
3798
+ "loss": 1.2484,
3799
+ "step": 299000
3800
+ },
3801
+ {
3802
+ "epoch": 194.77,
3803
+ "learning_rate": 1.3707752573255405e-09,
3804
+ "loss": 1.2367,
3805
+ "step": 299500
3806
+ },
3807
+ {
3808
+ "epoch": 195.09,
3809
+ "learning_rate": 0.0,
3810
+ "loss": 1.2385,
3811
+ "step": 300000
3812
+ },
3813
+ {
3814
+ "epoch": 195.09,
3815
+ "eval_runtime": 0.0012,
3816
+ "eval_samples_per_second": 0.0,
3817
+ "eval_steps_per_second": 0.0,
3818
+ "step": 300000
3819
+ }
3820
+ ],
3821
+ "max_steps": 300000,
3822
+ "num_train_epochs": 196,
3823
+ "total_flos": 2.1823985264086426e+18,
3824
+ "trial_name": null,
3825
+ "trial_params": null
3826
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e99bad87055c8596f50b2f43b37b391f7401b000f7a6ab850e76b7833ce2eda
3
+ size 4152