VideoChatGPT
6kplus commited on
Commit
1a33f17
·
verified ·
1 Parent(s): 6e80c6e

Upload 5 files

Browse files
adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "/nfs/turbo/coe-chaijy-unreplicated/pre-trained-weights/LLaVA/LLaVA-7B-Lightening-v1-1/",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 256,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "o_proj",
19
+ "v_proj",
20
+ "k_proj",
21
+ "up_proj",
22
+ "gate_proj",
23
+ "down_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b5003734a30219966d3a5751cfd8eda85ac975bf4ae3a6ea1174ccc07589cbf
3
+ size 639787082
config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/nfs/turbo/coe-chaijy-unreplicated/pre-trained-weights/LLaVA/LLaVA-7B-Lightening-v1-1/",
3
+ "architectures": [
4
+ "LlavaLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 0,
9
+ "eos_token_id": 1,
10
+ "freeze_mm_mlp_adapter": false,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "max_position_embeddings": 2048,
16
+ "max_sequence_length": 2048,
17
+ "mm_hidden_size": 1024,
18
+ "mm_use_im_start_end": true,
19
+ "mm_use_vid_start_end": true,
20
+ "mm_vision_select_layer": -2,
21
+ "mm_vision_tower": "openai/clip-vit-large-patch14",
22
+ "model_type": "VideoChatGPT",
23
+ "num_attention_heads": 32,
24
+ "num_hidden_layers": 32,
25
+ "num_key_value_heads": 32,
26
+ "pad_token_id": -1,
27
+ "pretraining_tp": 1,
28
+ "rms_norm_eps": 1e-06,
29
+ "rope_scaling": null,
30
+ "rope_theta": 10000.0,
31
+ "sep_video_conv_front": false,
32
+ "tie_word_embeddings": false,
33
+ "torch_dtype": "float16",
34
+ "transformers_version": "4.38.1",
35
+ "tune_mm_mlp_adapter": false,
36
+ "use_cache": false,
37
+ "use_mm_proj": true,
38
+ "vocab_size": 32006
39
+ }
non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
3
+ size 912
trainer_state.json ADDED
@@ -0,0 +1,751 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 10317,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.03,
13
+ "grad_norm": 0.0,
14
+ "learning_rate": 1.6129032258064517e-05,
15
+ "loss": 1.0967,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.06,
20
+ "grad_norm": 0.0,
21
+ "learning_rate": 3.2258064516129034e-05,
22
+ "loss": 0.1342,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.09,
27
+ "grad_norm": 0.0,
28
+ "learning_rate": 4.8387096774193554e-05,
29
+ "loss": 0.1196,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.12,
34
+ "grad_norm": 0.0,
35
+ "learning_rate": 4.9990021664883965e-05,
36
+ "loss": 0.1065,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.15,
41
+ "grad_norm": 0.0,
42
+ "learning_rate": 4.995553888081917e-05,
43
+ "loss": 0.1011,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.17,
48
+ "grad_norm": 0.0,
49
+ "learning_rate": 4.989646243314479e-05,
50
+ "loss": 0.1021,
51
+ "step": 600
52
+ },
53
+ {
54
+ "epoch": 0.2,
55
+ "grad_norm": 0.0,
56
+ "learning_rate": 4.9812850541652764e-05,
57
+ "loss": 0.0961,
58
+ "step": 700
59
+ },
60
+ {
61
+ "epoch": 0.23,
62
+ "grad_norm": 0.0,
63
+ "learning_rate": 4.970478560579529e-05,
64
+ "loss": 0.089,
65
+ "step": 800
66
+ },
67
+ {
68
+ "epoch": 0.26,
69
+ "grad_norm": 0.0,
70
+ "learning_rate": 4.9572374123480235e-05,
71
+ "loss": 0.0853,
72
+ "step": 900
73
+ },
74
+ {
75
+ "epoch": 0.29,
76
+ "grad_norm": 0.0,
77
+ "learning_rate": 4.941574658611754e-05,
78
+ "loss": 0.0866,
79
+ "step": 1000
80
+ },
81
+ {
82
+ "epoch": 0.32,
83
+ "grad_norm": 0.0,
84
+ "learning_rate": 4.923505735002e-05,
85
+ "loss": 0.0824,
86
+ "step": 1100
87
+ },
88
+ {
89
+ "epoch": 0.35,
90
+ "grad_norm": 0.0,
91
+ "learning_rate": 4.90304844842853e-05,
92
+ "loss": 0.0846,
93
+ "step": 1200
94
+ },
95
+ {
96
+ "epoch": 0.38,
97
+ "grad_norm": 0.0,
98
+ "learning_rate": 4.880222959530902e-05,
99
+ "loss": 0.0837,
100
+ "step": 1300
101
+ },
102
+ {
103
+ "epoch": 0.41,
104
+ "grad_norm": 0.0,
105
+ "learning_rate": 4.85505176281017e-05,
106
+ "loss": 0.0792,
107
+ "step": 1400
108
+ },
109
+ {
110
+ "epoch": 0.44,
111
+ "grad_norm": 0.0,
112
+ "learning_rate": 4.8275596644605725e-05,
113
+ "loss": 0.0757,
114
+ "step": 1500
115
+ },
116
+ {
117
+ "epoch": 0.47,
118
+ "grad_norm": 0.0,
119
+ "learning_rate": 4.7977737579230465e-05,
120
+ "loss": 0.0769,
121
+ "step": 1600
122
+ },
123
+ {
124
+ "epoch": 0.49,
125
+ "grad_norm": 0.0,
126
+ "learning_rate": 4.765723397184665e-05,
127
+ "loss": 0.0725,
128
+ "step": 1700
129
+ },
130
+ {
131
+ "epoch": 0.52,
132
+ "grad_norm": 0.0,
133
+ "learning_rate": 4.731440167850305e-05,
134
+ "loss": 0.0772,
135
+ "step": 1800
136
+ },
137
+ {
138
+ "epoch": 0.55,
139
+ "grad_norm": 0.0,
140
+ "learning_rate": 4.6949578560150574e-05,
141
+ "loss": 0.0728,
142
+ "step": 1900
143
+ },
144
+ {
145
+ "epoch": 0.58,
146
+ "grad_norm": 0.0,
147
+ "learning_rate": 4.65631241496806e-05,
148
+ "loss": 0.066,
149
+ "step": 2000
150
+ },
151
+ {
152
+ "epoch": 0.61,
153
+ "grad_norm": 0.0,
154
+ "learning_rate": 4.615541929760561e-05,
155
+ "loss": 0.0715,
156
+ "step": 2100
157
+ },
158
+ {
159
+ "epoch": 0.64,
160
+ "grad_norm": 0.0,
161
+ "learning_rate": 4.572686579673125e-05,
162
+ "loss": 0.0678,
163
+ "step": 2200
164
+ },
165
+ {
166
+ "epoch": 0.67,
167
+ "grad_norm": 0.0,
168
+ "learning_rate": 4.527788598618992e-05,
169
+ "loss": 0.0663,
170
+ "step": 2300
171
+ },
172
+ {
173
+ "epoch": 0.7,
174
+ "grad_norm": 0.0,
175
+ "learning_rate": 4.480892233522584e-05,
176
+ "loss": 0.0691,
177
+ "step": 2400
178
+ },
179
+ {
180
+ "epoch": 0.73,
181
+ "grad_norm": 0.0,
182
+ "learning_rate": 4.432043700714202e-05,
183
+ "loss": 0.0688,
184
+ "step": 2500
185
+ },
186
+ {
187
+ "epoch": 0.76,
188
+ "grad_norm": 0.0,
189
+ "learning_rate": 4.381291140383861e-05,
190
+ "loss": 0.0633,
191
+ "step": 2600
192
+ },
193
+ {
194
+ "epoch": 0.79,
195
+ "grad_norm": 0.0,
196
+ "learning_rate": 4.3286845691391785e-05,
197
+ "loss": 0.0662,
198
+ "step": 2700
199
+ },
200
+ {
201
+ "epoch": 0.81,
202
+ "grad_norm": 0.0,
203
+ "learning_rate": 4.274275830714048e-05,
204
+ "loss": 0.061,
205
+ "step": 2800
206
+ },
207
+ {
208
+ "epoch": 0.84,
209
+ "grad_norm": 0.0,
210
+ "learning_rate": 4.218118544876673e-05,
211
+ "loss": 0.0627,
212
+ "step": 2900
213
+ },
214
+ {
215
+ "epoch": 0.87,
216
+ "grad_norm": 0.0,
217
+ "learning_rate": 4.160268054587338e-05,
218
+ "loss": 0.0637,
219
+ "step": 3000
220
+ },
221
+ {
222
+ "epoch": 0.9,
223
+ "grad_norm": 0.0,
224
+ "learning_rate": 4.100781371457958e-05,
225
+ "loss": 0.0643,
226
+ "step": 3100
227
+ },
228
+ {
229
+ "epoch": 0.93,
230
+ "grad_norm": 0.0,
231
+ "learning_rate": 4.039717119567185e-05,
232
+ "loss": 0.0606,
233
+ "step": 3200
234
+ },
235
+ {
236
+ "epoch": 0.96,
237
+ "grad_norm": 0.0,
238
+ "learning_rate": 3.9771354776864214e-05,
239
+ "loss": 0.0588,
240
+ "step": 3300
241
+ },
242
+ {
243
+ "epoch": 0.99,
244
+ "grad_norm": 0.0,
245
+ "learning_rate": 3.913098119973694e-05,
246
+ "loss": 0.0619,
247
+ "step": 3400
248
+ },
249
+ {
250
+ "epoch": 1.02,
251
+ "grad_norm": 0.0,
252
+ "learning_rate": 3.8476681551938166e-05,
253
+ "loss": 0.0584,
254
+ "step": 3500
255
+ },
256
+ {
257
+ "epoch": 1.05,
258
+ "grad_norm": 0.0,
259
+ "learning_rate": 3.7809100645247544e-05,
260
+ "loss": 0.0531,
261
+ "step": 3600
262
+ },
263
+ {
264
+ "epoch": 1.08,
265
+ "grad_norm": 0.0,
266
+ "learning_rate": 3.7128896380114734e-05,
267
+ "loss": 0.0501,
268
+ "step": 3700
269
+ },
270
+ {
271
+ "epoch": 1.1,
272
+ "grad_norm": 0.0,
273
+ "learning_rate": 3.643673909729906e-05,
274
+ "loss": 0.0528,
275
+ "step": 3800
276
+ },
277
+ {
278
+ "epoch": 1.13,
279
+ "grad_norm": 0.0,
280
+ "learning_rate": 3.5733310917249186e-05,
281
+ "loss": 0.0494,
282
+ "step": 3900
283
+ },
284
+ {
285
+ "epoch": 1.16,
286
+ "grad_norm": 0.0,
287
+ "learning_rate": 3.5019305067874016e-05,
288
+ "loss": 0.056,
289
+ "step": 4000
290
+ },
291
+ {
292
+ "epoch": 1.19,
293
+ "grad_norm": 0.0,
294
+ "learning_rate": 3.4295425201367106e-05,
295
+ "loss": 0.0484,
296
+ "step": 4100
297
+ },
298
+ {
299
+ "epoch": 1.22,
300
+ "grad_norm": 0.0,
301
+ "learning_rate": 3.356238470075799e-05,
302
+ "loss": 0.0508,
303
+ "step": 4200
304
+ },
305
+ {
306
+ "epoch": 1.25,
307
+ "grad_norm": 0.0,
308
+ "learning_rate": 3.28209059768738e-05,
309
+ "loss": 0.0466,
310
+ "step": 4300
311
+ },
312
+ {
313
+ "epoch": 1.28,
314
+ "grad_norm": 0.0,
315
+ "learning_rate": 3.207171975640394e-05,
316
+ "loss": 0.0427,
317
+ "step": 4400
318
+ },
319
+ {
320
+ "epoch": 1.31,
321
+ "grad_norm": 0.0,
322
+ "learning_rate": 3.1315564361769544e-05,
323
+ "loss": 0.0516,
324
+ "step": 4500
325
+ },
326
+ {
327
+ "epoch": 1.34,
328
+ "grad_norm": 0.0,
329
+ "learning_rate": 3.0553184983507316e-05,
330
+ "loss": 0.0512,
331
+ "step": 4600
332
+ },
333
+ {
334
+ "epoch": 1.37,
335
+ "grad_norm": 0.0,
336
+ "learning_rate": 2.9785332945884835e-05,
337
+ "loss": 0.0501,
338
+ "step": 4700
339
+ },
340
+ {
341
+ "epoch": 1.4,
342
+ "grad_norm": 0.0,
343
+ "learning_rate": 2.9012764966471118e-05,
344
+ "loss": 0.0465,
345
+ "step": 4800
346
+ },
347
+ {
348
+ "epoch": 1.42,
349
+ "grad_norm": 0.0,
350
+ "learning_rate": 2.8236242410392038e-05,
351
+ "loss": 0.0486,
352
+ "step": 4900
353
+ },
354
+ {
355
+ "epoch": 1.45,
356
+ "grad_norm": 0.0,
357
+ "learning_rate": 2.7456530540005604e-05,
358
+ "loss": 0.049,
359
+ "step": 5000
360
+ },
361
+ {
362
+ "epoch": 1.48,
363
+ "grad_norm": 0.0,
364
+ "learning_rate": 2.6674397760736534e-05,
365
+ "loss": 0.0446,
366
+ "step": 5100
367
+ },
368
+ {
369
+ "epoch": 1.51,
370
+ "grad_norm": 0.0,
371
+ "learning_rate": 2.589061486381331e-05,
372
+ "loss": 0.0445,
373
+ "step": 5200
374
+ },
375
+ {
376
+ "epoch": 1.54,
377
+ "grad_norm": 0.0,
378
+ "learning_rate": 2.5105954266654043e-05,
379
+ "loss": 0.043,
380
+ "step": 5300
381
+ },
382
+ {
383
+ "epoch": 1.57,
384
+ "grad_norm": 0.0,
385
+ "learning_rate": 2.432118925164972e-05,
386
+ "loss": 0.047,
387
+ "step": 5400
388
+ },
389
+ {
390
+ "epoch": 1.6,
391
+ "grad_norm": 0.0,
392
+ "learning_rate": 2.3537093204095038e-05,
393
+ "loss": 0.0451,
394
+ "step": 5500
395
+ },
396
+ {
397
+ "epoch": 1.63,
398
+ "grad_norm": 0.0,
399
+ "learning_rate": 2.27544388500178e-05,
400
+ "loss": 0.0432,
401
+ "step": 5600
402
+ },
403
+ {
404
+ "epoch": 1.66,
405
+ "grad_norm": 0.0,
406
+ "learning_rate": 2.197399749465805e-05,
407
+ "loss": 0.0426,
408
+ "step": 5700
409
+ },
410
+ {
411
+ "epoch": 1.69,
412
+ "grad_norm": 0.0,
413
+ "learning_rate": 2.1196538262347358e-05,
414
+ "loss": 0.0433,
415
+ "step": 5800
416
+ },
417
+ {
418
+ "epoch": 1.72,
419
+ "grad_norm": 0.0,
420
+ "learning_rate": 2.0422827338537416e-05,
421
+ "loss": 0.0451,
422
+ "step": 5900
423
+ },
424
+ {
425
+ "epoch": 1.74,
426
+ "grad_norm": 0.0,
427
+ "learning_rate": 1.965362721472488e-05,
428
+ "loss": 0.0476,
429
+ "step": 6000
430
+ },
431
+ {
432
+ "epoch": 1.77,
433
+ "grad_norm": 0.0,
434
+ "learning_rate": 1.888969593701663e-05,
435
+ "loss": 0.0418,
436
+ "step": 6100
437
+ },
438
+ {
439
+ "epoch": 1.8,
440
+ "grad_norm": 0.0,
441
+ "learning_rate": 1.8131786359075914e-05,
442
+ "loss": 0.0424,
443
+ "step": 6200
444
+ },
445
+ {
446
+ "epoch": 1.83,
447
+ "grad_norm": 0.0,
448
+ "learning_rate": 1.738064540018567e-05,
449
+ "loss": 0.0416,
450
+ "step": 6300
451
+ },
452
+ {
453
+ "epoch": 1.86,
454
+ "grad_norm": 0.0,
455
+ "learning_rate": 1.6637013309160194e-05,
456
+ "loss": 0.0436,
457
+ "step": 6400
458
+ },
459
+ {
460
+ "epoch": 1.89,
461
+ "grad_norm": 0.0,
462
+ "learning_rate": 1.5901622934830474e-05,
463
+ "loss": 0.0419,
464
+ "step": 6500
465
+ },
466
+ {
467
+ "epoch": 1.92,
468
+ "grad_norm": 0.0,
469
+ "learning_rate": 1.5175199003822325e-05,
470
+ "loss": 0.0405,
471
+ "step": 6600
472
+ },
473
+ {
474
+ "epoch": 1.95,
475
+ "grad_norm": 0.0,
476
+ "learning_rate": 1.445845740633881e-05,
477
+ "loss": 0.0392,
478
+ "step": 6700
479
+ },
480
+ {
481
+ "epoch": 1.98,
482
+ "grad_norm": 0.0,
483
+ "learning_rate": 1.3752104490651061e-05,
484
+ "loss": 0.0383,
485
+ "step": 6800
486
+ },
487
+ {
488
+ "epoch": 2.01,
489
+ "grad_norm": 0.0,
490
+ "learning_rate": 1.305683636699259e-05,
491
+ "loss": 0.0364,
492
+ "step": 6900
493
+ },
494
+ {
495
+ "epoch": 2.04,
496
+ "grad_norm": 0.0,
497
+ "learning_rate": 1.237333822154326e-05,
498
+ "loss": 0.0282,
499
+ "step": 7000
500
+ },
501
+ {
502
+ "epoch": 2.06,
503
+ "grad_norm": 0.0,
504
+ "learning_rate": 1.1702283641178782e-05,
505
+ "loss": 0.028,
506
+ "step": 7100
507
+ },
508
+ {
509
+ "epoch": 2.09,
510
+ "grad_norm": 0.0,
511
+ "learning_rate": 1.1044333949651476e-05,
512
+ "loss": 0.0296,
513
+ "step": 7200
514
+ },
515
+ {
516
+ "epoch": 2.12,
517
+ "grad_norm": 0.0,
518
+ "learning_rate": 1.0400137555856137e-05,
519
+ "loss": 0.0263,
520
+ "step": 7300
521
+ },
522
+ {
523
+ "epoch": 2.15,
524
+ "grad_norm": 0.0,
525
+ "learning_rate": 9.770329314823701e-06,
526
+ "loss": 0.028,
527
+ "step": 7400
528
+ },
529
+ {
530
+ "epoch": 2.18,
531
+ "grad_norm": 0.0,
532
+ "learning_rate": 9.15552990207201e-06,
533
+ "loss": 0.0279,
534
+ "step": 7500
535
+ },
536
+ {
537
+ "epoch": 2.21,
538
+ "grad_norm": 0.0,
539
+ "learning_rate": 8.556345201930674e-06,
540
+ "loss": 0.0257,
541
+ "step": 7600
542
+ },
543
+ {
544
+ "epoch": 2.24,
545
+ "grad_norm": 0.0,
546
+ "learning_rate": 7.97336571044251e-06,
547
+ "loss": 0.028,
548
+ "step": 7700
549
+ },
550
+ {
551
+ "epoch": 2.27,
552
+ "grad_norm": 0.0,
553
+ "learning_rate": 7.4071659534301424e-06,
554
+ "loss": 0.0271,
555
+ "step": 7800
556
+ },
557
+ {
558
+ "epoch": 2.3,
559
+ "grad_norm": 0.0,
560
+ "learning_rate": 6.858303920301321e-06,
561
+ "loss": 0.0254,
562
+ "step": 7900
563
+ },
564
+ {
565
+ "epoch": 2.33,
566
+ "grad_norm": 0.0,
567
+ "learning_rate": 6.327320514150742e-06,
568
+ "loss": 0.0252,
569
+ "step": 8000
570
+ },
571
+ {
572
+ "epoch": 2.36,
573
+ "grad_norm": 0.0,
574
+ "learning_rate": 5.8147390187005465e-06,
575
+ "loss": 0.027,
576
+ "step": 8100
577
+ },
578
+ {
579
+ "epoch": 2.38,
580
+ "grad_norm": 0.0,
581
+ "learning_rate": 5.3210645826045565e-06,
582
+ "loss": 0.026,
583
+ "step": 8200
584
+ },
585
+ {
586
+ "epoch": 2.41,
587
+ "grad_norm": 0.0,
588
+ "learning_rate": 4.846783721624764e-06,
589
+ "loss": 0.0286,
590
+ "step": 8300
591
+ },
592
+ {
593
+ "epoch": 2.44,
594
+ "grad_norm": 0.0,
595
+ "learning_rate": 4.392363839170357e-06,
596
+ "loss": 0.0245,
597
+ "step": 8400
598
+ },
599
+ {
600
+ "epoch": 2.47,
601
+ "grad_norm": 0.0,
602
+ "learning_rate": 3.9582527656720964e-06,
603
+ "loss": 0.0256,
604
+ "step": 8500
605
+ },
606
+ {
607
+ "epoch": 2.5,
608
+ "grad_norm": 0.0,
609
+ "learning_rate": 3.5448783172457367e-06,
610
+ "loss": 0.0253,
611
+ "step": 8600
612
+ },
613
+ {
614
+ "epoch": 2.53,
615
+ "grad_norm": 0.0,
616
+ "learning_rate": 3.152647874079628e-06,
617
+ "loss": 0.0244,
618
+ "step": 8700
619
+ },
620
+ {
621
+ "epoch": 2.56,
622
+ "grad_norm": 0.0,
623
+ "learning_rate": 2.781947978961827e-06,
624
+ "loss": 0.0224,
625
+ "step": 8800
626
+ },
627
+ {
628
+ "epoch": 2.59,
629
+ "grad_norm": 0.0,
630
+ "learning_rate": 2.4331439563425234e-06,
631
+ "loss": 0.0246,
632
+ "step": 8900
633
+ },
634
+ {
635
+ "epoch": 2.62,
636
+ "grad_norm": 0.0,
637
+ "learning_rate": 2.106579552307053e-06,
638
+ "loss": 0.0241,
639
+ "step": 9000
640
+ },
641
+ {
642
+ "epoch": 2.65,
643
+ "grad_norm": 0.0,
644
+ "learning_rate": 1.8025765958144453e-06,
645
+ "loss": 0.0242,
646
+ "step": 9100
647
+ },
648
+ {
649
+ "epoch": 2.68,
650
+ "grad_norm": 0.0,
651
+ "learning_rate": 1.5214346815351942e-06,
652
+ "loss": 0.0258,
653
+ "step": 9200
654
+ },
655
+ {
656
+ "epoch": 2.7,
657
+ "grad_norm": 0.0,
658
+ "learning_rate": 1.2634308746009932e-06,
659
+ "loss": 0.0226,
660
+ "step": 9300
661
+ },
662
+ {
663
+ "epoch": 2.73,
664
+ "grad_norm": 0.0,
665
+ "learning_rate": 1.0288194375572436e-06,
666
+ "loss": 0.0254,
667
+ "step": 9400
668
+ },
669
+ {
670
+ "epoch": 2.76,
671
+ "grad_norm": 0.0,
672
+ "learning_rate": 8.178315797875257e-07,
673
+ "loss": 0.0236,
674
+ "step": 9500
675
+ },
676
+ {
677
+ "epoch": 2.79,
678
+ "grad_norm": 0.0,
679
+ "learning_rate": 6.30675229656949e-07,
680
+ "loss": 0.0247,
681
+ "step": 9600
682
+ },
683
+ {
684
+ "epoch": 2.82,
685
+ "grad_norm": 0.0,
686
+ "learning_rate": 4.6753482959888406e-07,
687
+ "loss": 0.0263,
688
+ "step": 9700
689
+ },
690
+ {
691
+ "epoch": 2.85,
692
+ "grad_norm": 0.0,
693
+ "learning_rate": 3.285711543471265e-07,
694
+ "loss": 0.0237,
695
+ "step": 9800
696
+ },
697
+ {
698
+ "epoch": 2.88,
699
+ "grad_norm": 0.0,
700
+ "learning_rate": 2.1392115249250222e-07,
701
+ "loss": 0.0251,
702
+ "step": 9900
703
+ },
704
+ {
705
+ "epoch": 2.91,
706
+ "grad_norm": 0.0,
707
+ "learning_rate": 1.2369781152015425e-07,
708
+ "loss": 0.0253,
709
+ "step": 10000
710
+ },
711
+ {
712
+ "epoch": 2.94,
713
+ "grad_norm": 0.0,
714
+ "learning_rate": 5.799004646045436e-08,
715
+ "loss": 0.0246,
716
+ "step": 10100
717
+ },
718
+ {
719
+ "epoch": 2.97,
720
+ "grad_norm": 0.0,
721
+ "learning_rate": 1.6862612263324195e-08,
722
+ "loss": 0.0247,
723
+ "step": 10200
724
+ },
725
+ {
726
+ "epoch": 3.0,
727
+ "grad_norm": 0.0,
728
+ "learning_rate": 3.560399822716276e-10,
729
+ "loss": 0.0241,
730
+ "step": 10300
731
+ },
732
+ {
733
+ "epoch": 3.0,
734
+ "step": 10317,
735
+ "total_flos": 1.8160995702249882e+18,
736
+ "train_loss": 0.059974256984898895,
737
+ "train_runtime": 16482.6502,
738
+ "train_samples_per_second": 2.503,
739
+ "train_steps_per_second": 0.626
740
+ }
741
+ ],
742
+ "logging_steps": 100,
743
+ "max_steps": 10317,
744
+ "num_input_tokens_seen": 0,
745
+ "num_train_epochs": 3,
746
+ "save_steps": 500,
747
+ "total_flos": 1.8160995702249882e+18,
748
+ "train_batch_size": 1,
749
+ "trial_name": null,
750
+ "trial_params": null
751
+ }