File size: 29,436 Bytes
a9eed45
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
2025-04-29,17:42:33 | INFO | No latest resume checkpoint found in /mnt/personal/zhudongy/datacomp_results/medium/low_inter_only/checkpoints.
2025-04-29,17:42:35 | INFO | Running in distributed mode with multiple processes. Device: cuda:0.Process (global: 0, local 0), total 2.
2025-04-29,17:42:35 | INFO | Loaded ViT-B-32 model config.
2025-04-29,17:42:36 | INFO | Model:
2025-04-29,17:42:36 | INFO | CLIP(
  (visual): VisionTransformer(
    (patchnorm_pre_ln): Identity()
    (conv1): Conv2d(3, 768, kernel_size=(32, 32), stride=(32, 32), bias=False)
    (patch_dropout): Identity()
    (ln_pre): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
    (transformer): Transformer(
      (resblocks): ModuleList(
        (0): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (1): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (2): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (3): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (4): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (5): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (6): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (7): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (8): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (9): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (10): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
        (11): ResidualAttentionBlock(
          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): MultiheadAttention(
            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
          )
          (ls_1): Identity()
          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Sequential(
            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
            (gelu): GELU(approximate='none')
            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
          )
          (ls_2): Identity()
        )
      )
    )
    (ln_post): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
  )
  (transformer): Transformer(
    (resblocks): ModuleList(
      (0): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (1): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (2): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (3): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (4): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (5): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (6): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (7): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (8): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (9): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (10): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
      (11): ResidualAttentionBlock(
        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (attn): MultiheadAttention(
          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
        )
        (ls_1): Identity()
        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (mlp): Sequential(
          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
          (gelu): GELU(approximate='none')
          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
        )
        (ls_2): Identity()
      )
    )
  )
  (token_embedding): Embedding(49408, 512)
  (ln_final): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
)
2025-04-29,17:42:36 | INFO | Params:
2025-04-29,17:42:36 | INFO |   accum_freq: 1
2025-04-29,17:42:36 | INFO |   aug_cfg: {}
2025-04-29,17:42:36 | INFO |   batch_size: 2048
2025-04-29,17:42:36 | INFO |   beta1: 0.9
2025-04-29,17:42:36 | INFO |   beta2: 0.98
2025-04-29,17:42:36 | INFO |   checkpoint_path: /mnt/personal/zhudongy/datacomp_results/medium/low_inter_only/checkpoints
2025-04-29,17:42:36 | INFO |   coca_caption_loss_weight: 2.0
2025-04-29,17:42:36 | INFO |   coca_contrastive_loss_weight: 1.0
2025-04-29,17:42:36 | INFO |   copy_codebase: False
2025-04-29,17:42:36 | INFO |   csv_caption_key: title
2025-04-29,17:42:36 | INFO |   csv_img_key: filepath
2025-04-29,17:42:36 | INFO |   csv_separator: 	
2025-04-29,17:42:36 | INFO |   dataset_resampled: True
2025-04-29,17:42:36 | INFO |   dataset_type: webdataset
2025-04-29,17:42:36 | INFO |   ddp_static_graph: True
2025-04-29,17:42:36 | INFO |   debug: False
2025-04-29,17:42:36 | INFO |   delete_previous_checkpoint: False
2025-04-29,17:42:36 | INFO |   device: cuda:0
2025-04-29,17:42:36 | INFO |   dist_backend: nccl
2025-04-29,17:42:36 | INFO |   dist_url: env://
2025-04-29,17:42:36 | INFO |   distill: False
2025-04-29,17:42:36 | INFO |   distill_model: None
2025-04-29,17:42:36 | INFO |   distill_pretrained: None
2025-04-29,17:42:36 | INFO |   distributed: True
2025-04-29,17:42:36 | INFO |   epochs: 8
2025-04-29,17:42:36 | INFO |   epochs_cooldown: None
2025-04-29,17:42:36 | INFO |   eps: 1e-06
2025-04-29,17:42:36 | INFO |   force_custom_text: False
2025-04-29,17:42:36 | INFO |   force_image_size: None
2025-04-29,17:42:36 | INFO |   force_patch_dropout: None
2025-04-29,17:42:36 | INFO |   force_quick_gelu: False
2025-04-29,17:42:36 | INFO |   gather_with_grad: True
2025-04-29,17:42:36 | INFO |   grad_checkpointing: True
2025-04-29,17:42:36 | INFO |   grad_clip_norm: None
2025-04-29,17:42:36 | INFO |   horovod: False
2025-04-29,17:42:36 | INFO |   image_mean: None
2025-04-29,17:42:36 | INFO |   image_std: None
2025-04-29,17:42:36 | INFO |   imagenet_v2: None
2025-04-29,17:42:36 | INFO |   imagenet_val: None
2025-04-29,17:42:36 | INFO |   local_loss: True
2025-04-29,17:42:36 | INFO |   local_rank: 0
2025-04-29,17:42:36 | INFO |   lock_image: False
2025-04-29,17:42:36 | INFO |   lock_image_freeze_bn_stats: False
2025-04-29,17:42:36 | INFO |   lock_image_unlocked_groups: 0
2025-04-29,17:42:36 | INFO |   lock_text: False
2025-04-29,17:42:36 | INFO |   lock_text_freeze_layer_norm: False
2025-04-29,17:42:36 | INFO |   lock_text_unlocked_layers: 0
2025-04-29,17:42:36 | INFO |   log_every_n_steps: 100
2025-04-29,17:42:36 | INFO |   log_level: 20
2025-04-29,17:42:36 | INFO |   log_local: False
2025-04-29,17:42:36 | INFO |   log_path: /mnt/personal/zhudongy/datacomp_results/medium/low_inter_only/out.log
2025-04-29,17:42:36 | INFO |   logs: /mnt/personal/zhudongy/datacomp_results/medium
2025-04-29,17:42:36 | INFO |   lr: 0.0005
2025-04-29,17:42:36 | INFO |   lr_cooldown_end: 0.0
2025-04-29,17:42:36 | INFO |   lr_cooldown_power: 1.0
2025-04-29,17:42:36 | INFO |   lr_scheduler: cosine
2025-04-29,17:42:36 | INFO |   model: ViT-B-32
2025-04-29,17:42:36 | INFO |   name: low_inter_only
2025-04-29,17:42:36 | INFO |   no_set_device_rank: False
2025-04-29,17:42:36 | INFO |   precision: amp_bfloat16
2025-04-29,17:42:36 | INFO |   pretrained: 
2025-04-29,17:42:36 | INFO |   pretrained_image: False
2025-04-29,17:42:36 | INFO |   rank: 0
2025-04-29,17:42:36 | INFO |   remote_sync: None
2025-04-29,17:42:36 | INFO |   remote_sync_frequency: 300
2025-04-29,17:42:36 | INFO |   remote_sync_protocol: s3
2025-04-29,17:42:36 | INFO |   report_to: 
2025-04-29,17:42:36 | INFO |   resume: None
2025-04-29,17:42:36 | INFO |   save_frequency: 0
2025-04-29,17:42:36 | INFO |   save_most_recent: True
2025-04-29,17:42:36 | INFO |   seed: 0
2025-04-29,17:42:36 | INFO |   skip_scheduler: False
2025-04-29,17:42:36 | INFO |   tensorboard: False
2025-04-29,17:42:36 | INFO |   tensorboard_path: 
2025-04-29,17:42:36 | INFO |   torchscript: False
2025-04-29,17:42:36 | INFO |   trace: False
2025-04-29,17:42:36 | INFO |   train_data: /mnt/personal/zhudongy/datacomp-medium/shards/0000{0000..6126}.tar
2025-04-29,17:42:36 | INFO |   train_data_upsampling_factors: None
2025-04-29,17:42:36 | INFO |   train_num_samples: 16000000
2025-04-29,17:42:36 | INFO |   use_bn_sync: False
2025-04-29,17:42:36 | INFO |   val_data: None
2025-04-29,17:42:36 | INFO |   val_frequency: 1
2025-04-29,17:42:36 | INFO |   val_num_samples: None
2025-04-29,17:42:36 | INFO |   wandb: False
2025-04-29,17:42:36 | INFO |   wandb_notes: 
2025-04-29,17:42:36 | INFO |   wandb_project_name: open-clip
2025-04-29,17:42:36 | INFO |   warmup: 500
2025-04-29,17:42:36 | INFO |   wd: 0.2
2025-04-29,17:42:36 | INFO |   workers: 16
2025-04-29,17:42:36 | INFO |   world_size: 2
2025-04-29,17:42:36 | INFO |   zeroshot_frequency: 2
2025-04-29,17:42:36 | INFO | Start epoch 0
2025-04-29,17:43:01 | INFO | Train Epoch: 0 [    4096/16056320 (0%)] Data (t): 21.925 Batch (t): 24.947, 164.187/s, 82.0936/s/gpu LR: 0.000001 Logit Scale: 14.286 Contrastive_loss: 8.3837 (8.3837) Loss: 8.3837 (8.3837)
2025-04-29,17:43:04 | INFO | Reducer buckets have been rebuilt in this iteration.
2025-04-29,17:47:03 | INFO | Train Epoch: 0 [  413696/16056320 (3%)] Data (t): 0.367 Batch (t): 2.418, 1701.28/s, 850.642/s/gpu LR: 0.000101 Logit Scale: 14.261 Contrastive_loss: 8.1890 (8.2863) Loss: 8.1890 (8.2863)
2025-04-29,17:51:07 | INFO | Train Epoch: 0 [  823296/16056320 (5%)] Data (t): 0.400 Batch (t): 2.442, 1697.43/s, 848.716/s/gpu LR: 0.000201 Logit Scale: 14.237 Contrastive_loss: 8.0558 (8.2095) Loss: 8.0558 (8.2095)
2025-04-29,17:55:15 | INFO | Train Epoch: 0 [ 1232896/16056320 (8%)] Data (t): 0.436 Batch (t): 2.479, 1693.55/s, 846.774/s/gpu LR: 0.000301 Logit Scale: 14.210 Contrastive_loss: 7.9360 (8.1411) Loss: 7.9360 (8.1411)
2025-04-29,17:59:16 | INFO | Train Epoch: 0 [ 1642496/16056320 (10%)] Data (t): 0.372 Batch (t): 2.411, 1693.62/s, 846.809/s/gpu LR: 0.000401 Logit Scale: 14.185 Contrastive_loss: 7.8394 (8.0808) Loss: 7.8394 (8.0808)
2025-04-29,18:03:17 | INFO | Train Epoch: 0 [ 2052096/16056320 (13%)] Data (t): 0.364 Batch (t): 2.407, 1698.74/s, 849.369/s/gpu LR: 0.000500 Logit Scale: 14.182 Contrastive_loss: 7.7843 (8.0314) Loss: 7.7843 (8.0314)
2025-04-29,18:07:18 | INFO | Train Epoch: 0 [ 2461696/16056320 (15%)] Data (t): 0.369 Batch (t): 2.413, 1692.95/s, 846.474/s/gpu LR: 0.000500 Logit Scale: 14.201 Contrastive_loss: 7.6211 (7.9728) Loss: 7.6211 (7.9728)
2025-04-29,18:11:19 | INFO | Train Epoch: 0 [ 2871296/16056320 (18%)] Data (t): 0.365 Batch (t): 2.407, 1707.98/s, 853.988/s/gpu LR: 0.000500 Logit Scale: 14.260 Contrastive_loss: 7.5299 (7.9174) Loss: 7.5299 (7.9174)
2025-04-29,18:15:20 | INFO | Train Epoch: 0 [ 3280896/16056320 (20%)] Data (t): 0.369 Batch (t): 2.410, 1690.42/s, 845.211/s/gpu LR: 0.000500 Logit Scale: 14.340 Contrastive_loss: 7.4407 (7.8644) Loss: 7.4407 (7.8644)
2025-04-29,18:19:24 | INFO | Train Epoch: 0 [ 3690496/16056320 (23%)] Data (t): 0.393 Batch (t): 2.436, 1701.12/s, 850.559/s/gpu LR: 0.000500 Logit Scale: 14.441 Contrastive_loss: 7.4649 (7.8245) Loss: 7.4649 (7.8245)
2025-04-29,18:23:25 | INFO | Train Epoch: 0 [ 4100096/16056320 (26%)] Data (t): 0.378 Batch (t): 2.417, 1712.22/s, 856.110/s/gpu LR: 0.000500 Logit Scale: 14.582 Contrastive_loss: 7.3469 (7.7811) Loss: 7.3469 (7.7811)
2025-04-29,18:27:27 | INFO | Train Epoch: 0 [ 4509696/16056320 (28%)] Data (t): 0.372 Batch (t): 2.413, 1699.89/s, 849.943/s/gpu LR: 0.000500 Logit Scale: 14.727 Contrastive_loss: 7.2976 (7.7408) Loss: 7.2976 (7.7408)
2025-04-29,18:31:29 | INFO | Train Epoch: 0 [ 4919296/16056320 (31%)] Data (t): 0.378 Batch (t): 2.418, 1680.76/s, 840.382/s/gpu LR: 0.000499 Logit Scale: 14.942 Contrastive_loss: 7.2187 (7.7006) Loss: 7.2187 (7.7006)
2025-04-29,18:35:30 | INFO | Train Epoch: 0 [ 5328896/16056320 (33%)] Data (t): 0.375 Batch (t): 2.417, 1677.33/s, 838.665/s/gpu LR: 0.000499 Logit Scale: 15.128 Contrastive_loss: 7.1773 (7.6632) Loss: 7.1773 (7.6632)
2025-04-29,18:39:32 | INFO | Train Epoch: 0 [ 5738496/16056320 (36%)] Data (t): 0.373 Batch (t): 2.416, 1698.13/s, 849.067/s/gpu LR: 0.000499 Logit Scale: 15.344 Contrastive_loss: 7.0983 (7.6256) Loss: 7.0983 (7.6256)
2025-04-29,18:43:35 | INFO | Train Epoch: 0 [ 6148096/16056320 (38%)] Data (t): 0.390 Batch (t): 2.433, 1681.07/s, 840.537/s/gpu LR: 0.000499 Logit Scale: 15.595 Contrastive_loss: 7.0212 (7.5878) Loss: 7.0212 (7.5878)
2025-04-29,18:47:38 | INFO | Train Epoch: 0 [ 6557696/16056320 (41%)] Data (t): 0.390 Batch (t): 2.430, 1683.60/s, 841.802/s/gpu LR: 0.000498 Logit Scale: 15.826 Contrastive_loss: 6.9587 (7.5508) Loss: 6.9587 (7.5508)
2025-04-29,18:51:40 | INFO | Train Epoch: 0 [ 6967296/16056320 (43%)] Data (t): 0.378 Batch (t): 2.418, 1698.91/s, 849.455/s/gpu LR: 0.000498 Logit Scale: 16.115 Contrastive_loss: 6.8671 (7.5128) Loss: 6.8671 (7.5128)
2025-04-29,18:55:42 | INFO | Train Epoch: 0 [ 7376896/16056320 (46%)] Data (t): 0.380 Batch (t): 2.423, 1679.70/s, 839.848/s/gpu LR: 0.000498 Logit Scale: 16.413 Contrastive_loss: 6.9072 (7.4809) Loss: 6.9072 (7.4809)
2025-04-29,18:59:46 | INFO | Train Epoch: 0 [ 7786496/16056320 (48%)] Data (t): 0.390 Batch (t): 2.432, 1697.74/s, 848.868/s/gpu LR: 0.000497 Logit Scale: 16.734 Contrastive_loss: 6.8222 (7.4480) Loss: 6.8222 (7.4480)
2025-04-29,19:03:50 | INFO | Train Epoch: 0 [ 8196096/16056320 (51%)] Data (t): 0.392 Batch (t): 2.440, 1684.68/s, 842.340/s/gpu LR: 0.000497 Logit Scale: 17.017 Contrastive_loss: 6.7487 (7.4147) Loss: 6.7487 (7.4147)
2025-04-29,19:07:55 | INFO | Train Epoch: 0 [ 8605696/16056320 (54%)] Data (t): 0.414 Batch (t): 2.455, 1699.34/s, 849.669/s/gpu LR: 0.000497 Logit Scale: 17.283 Contrastive_loss: 6.6749 (7.3811) Loss: 6.6749 (7.3811)
2025-04-29,19:11:58 | INFO | Train Epoch: 0 [ 9015296/16056320 (56%)] Data (t): 0.384 Batch (t): 2.425, 1704.23/s, 852.116/s/gpu LR: 0.000496 Logit Scale: 17.559 Contrastive_loss: 6.5998 (7.3471) Loss: 6.5998 (7.3471)
2025-04-29,19:16:01 | INFO | Train Epoch: 0 [ 9424896/16056320 (59%)] Data (t): 0.387 Batch (t): 2.438, 1666.65/s, 833.323/s/gpu LR: 0.000496 Logit Scale: 17.832 Contrastive_loss: 6.5839 (7.3153) Loss: 6.5839 (7.3153)
2025-04-29,19:20:05 | INFO | Train Epoch: 0 [ 9834496/16056320 (61%)] Data (t): 0.393 Batch (t): 2.432, 1699.78/s, 849.890/s/gpu LR: 0.000495 Logit Scale: 18.053 Contrastive_loss: 6.4805 (7.2819) Loss: 6.4805 (7.2819)
2025-04-29,19:24:07 | INFO | Train Epoch: 0 [10244096/16056320 (64%)] Data (t): 0.382 Batch (t): 2.425, 1685.49/s, 842.747/s/gpu LR: 0.000495 Logit Scale: 18.360 Contrastive_loss: 6.5788 (7.2549) Loss: 6.5788 (7.2549)
2025-04-29,19:28:10 | INFO | Train Epoch: 0 [10653696/16056320 (66%)] Data (t): 0.391 Batch (t): 2.431, 1693.68/s, 846.839/s/gpu LR: 0.000494 Logit Scale: 18.648 Contrastive_loss: 6.4210 (7.2240) Loss: 6.4210 (7.2240)
2025-04-29,19:32:14 | INFO | Train Epoch: 0 [11063296/16056320 (69%)] Data (t): 0.396 Batch (t): 2.438, 1692.32/s, 846.162/s/gpu LR: 0.000494 Logit Scale: 18.913 Contrastive_loss: 6.4979 (7.1981) Loss: 6.4979 (7.1981)
2025-04-29,19:36:17 | INFO | Train Epoch: 0 [11472896/16056320 (71%)] Data (t): 0.390 Batch (t): 2.429, 1678.49/s, 839.246/s/gpu LR: 0.000493 Logit Scale: 19.224 Contrastive_loss: 6.4005 (7.1706) Loss: 6.4005 (7.1706)
2025-04-29,19:40:20 | INFO | Train Epoch: 0 [11882496/16056320 (74%)] Data (t): 0.387 Batch (t): 2.430, 1678.26/s, 839.130/s/gpu LR: 0.000493 Logit Scale: 19.531 Contrastive_loss: 6.3682 (7.1438) Loss: 6.3682 (7.1438)
2025-04-29,19:44:23 | INFO | Train Epoch: 0 [12292096/16056320 (77%)] Data (t): 0.396 Batch (t): 2.436, 1677.17/s, 838.587/s/gpu LR: 0.000492 Logit Scale: 19.819 Contrastive_loss: 6.1071 (7.1104) Loss: 6.1071 (7.1104)
2025-04-29,19:48:27 | INFO | Train Epoch: 0 [12701696/16056320 (79%)] Data (t): 0.398 Batch (t): 2.438, 1696.85/s, 848.426/s/gpu LR: 0.000491 Logit Scale: 20.042 Contrastive_loss: 6.3618 (7.0870) Loss: 6.3618 (7.0870)
2025-04-29,19:52:31 | INFO | Train Epoch: 0 [13111296/16056320 (82%)] Data (t): 0.398 Batch (t): 2.439, 1606.99/s, 803.494/s/gpu LR: 0.000491 Logit Scale: 20.290 Contrastive_loss: 6.2285 (7.0610) Loss: 6.2285 (7.0610)
2025-04-29,19:56:35 | INFO | Train Epoch: 0 [13520896/16056320 (84%)] Data (t): 0.402 Batch (t): 2.443, 1668.95/s, 834.473/s/gpu LR: 0.000490 Logit Scale: 20.522 Contrastive_loss: 6.0085 (7.0300) Loss: 6.0085 (7.0300)
2025-04-29,20:00:38 | INFO | Train Epoch: 0 [13930496/16056320 (87%)] Data (t): 0.389 Batch (t): 2.429, 1705.41/s, 852.704/s/gpu LR: 0.000489 Logit Scale: 20.775 Contrastive_loss: 6.2236 (7.0070) Loss: 6.2236 (7.0070)
2025-04-29,20:04:42 | INFO | Train Epoch: 0 [14340096/16056320 (89%)] Data (t): 0.397 Batch (t): 2.439, 1616.17/s, 808.085/s/gpu LR: 0.000488 Logit Scale: 21.015 Contrastive_loss: 6.1521 (6.9832) Loss: 6.1521 (6.9832)
2025-04-29,20:08:46 | INFO | Train Epoch: 0 [14749696/16056320 (92%)] Data (t): 0.396 Batch (t): 2.438, 1700.78/s, 850.392/s/gpu LR: 0.000488 Logit Scale: 21.230 Contrastive_loss: 6.1918 (6.9618) Loss: 6.1918 (6.9618)
2025-04-29,20:12:49 | INFO | Train Epoch: 0 [15159296/16056320 (94%)] Data (t): 0.390 Batch (t): 2.432, 1695.14/s, 847.571/s/gpu LR: 0.000487 Logit Scale: 21.468 Contrastive_loss: 6.1201 (6.9397) Loss: 6.1201 (6.9397)
2025-04-29,20:16:52 | INFO | Train Epoch: 0 [15568896/16056320 (97%)] Data (t): 0.390 Batch (t): 2.432, 1704.81/s, 852.404/s/gpu LR: 0.000486 Logit Scale: 21.638 Contrastive_loss: 6.0887 (6.9179) Loss: 6.0887 (6.9179)