-
Notifications
You must be signed in to change notification settings - Fork 0
/
command.txt
552 lines (472 loc) · 69.3 KB
/
command.txt
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
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
python train.py --config-file /home/ch7858/vpt/configs/prompt/*.yaml SFER_TYPE "prompt" MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "50" MODEL.PROMPT.DROPOUT "0.0"
python train.py --config-file /home/ch7858/vpt/configs/prompt/cub.yaml SFER_TYPE "prompt" MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "50" MODEL.PROMPT.DROPOUT "0.0"
python tune_fgvc.py --config-file configs/prompt/*.yaml SFER_TYPE "prompt" MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "50" MODEL.PROMPT.DROPOUT "0.0"
# vpt example
python tune_vtab.py --train-type "prompt" --config-file configs/prompt/cifar100_forVPT.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.1"
python train.py \
--config-file configs/prompt/cub.yaml \
MODEL.TYPE "vit" \
DATA.BATCH_SIZE "64" \
MODEL.PROMPT.NUM_TOKENS "100" \
MODEL.PROMPT.DEEP "True" \
MODEL.PROMPT.DROPOUT "0.1" \
DATA.FEATURE "sup_vitb16_imagenet21k" \
DATA.NAME "vtab-dmlab" \
DATA.NUMBER_CLASSES "6" \
SOLVER.BASE_LR "0.25" \
SOLVER.WEIGHT_DECAY "0.001" \
SEED ${seed} \
MODEL.MODEL_ROOT "${model_root}" \
DATA.DATAPATH "${data_path}" \
OUTPUT_DIR "${output_dir}/seed${seed}"
python tune_fgvc.py \
--train-type "prompt" \
--config-file configs/prompt/cub.yaml \
MODEL.TYPE "vit" \
DATA.BATCH_SIZE "128" \
MODEL.PROMPT.DEEP "True" \
MODEL.PROMPT.DROPOUT "0.1" \
MODEL.PROMPT.NUM_TOKENS "10" \
DATA.FEATURE "sup_vitb16_imagenet21k" \
DATA.DATAPATH \
MODEL.MODEL_ROOT \
OUTPUT_DIR
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file /home/ch7858/vpt/configs/prompt/*.yaml MODEL.TRANSFER_TYPE "prompt" MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "50" MODEL.PROMPT.DROPOUT "0.0"
src里面提供了config的baseline 但是configs里面的文件的优先级在其之上(修改对应configs文件内的内容就行)
tune_fgvc 和 tune_vtab 是用来一次性跑很多实验的 (find the best lr and wd) wd as WEIGHT_DECAY
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file /home/ch7858/vpt/configs/prompt/prompt_test/VTAB-1k/Natural/cifar100.yaml MODEL.TRANSFER_TYPE "prompt" MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.0"
prompt_proj 写着一层的目的是区分训练部分 方便选择对应的部分进行train
DATASETS = [
'caltech101',
'cifar(num_classes=100)',
'dtd',
'oxford_flowers102',
'oxford_iiit_pet',
'patch_camelyon',
'sun397',
'svhn',
'resisc45',
'eurosat',
'dmlab',
'kitti(task="closest_vehicle_distance")',
'smallnorb(predicted_attribute="label_azimuth")',
'smallnorb(predicted_attribute="label_elevation")',
'dsprites(predicted_attribute="label_x_position",num_classes=16)',
'dsprites(predicted_attribute="label_orientation",num_classes=16)',
'clevr(task="closest_object_distance")',
'clevr(task="count_all")',
'diabetic_retinopathy(config="btgraham-300")'
]
##### P_VK approach separate commands
# FGVC
Cars:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/FGVC/cars.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
CUB:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/FGVC/cub.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Dogs:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/FGVC/dogs.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Flowers:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/FGVC/flowers.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Nabirds:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/FGVC/nabirds.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
# VTAB-1k
## Natural
Caltech101(Failed):
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Natural/caltech101.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Cifar100:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Natural/cifar100.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
dtd:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Natural/dtd.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Oxford_flowers102:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Natural/oxford_flowers102.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Oxford_iiit_pet:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Natural/oxford_iiit_pet.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
SUN397:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Natural/sun397.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Svhn_cropped:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Natural/svhn_cropped.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
## Specialized
Diabetic_retinopathy_detection (Failed (Due to memory)):
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/diabetic_retinopathy_detection.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Eurosat:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/eurosat.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Patch_camelyon:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/patch_camelyon.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Resisc45 (Failed):
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/resisc45.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
## Structured
Clevr_count:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Structured/clevr_count.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Clevr_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Structured/clevr_distance.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Dmlab:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dmlab.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Dsprites_location (Failed (Due to memory)):
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dsprites_location.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Dsprites_orientation (Failed (Due to memory)):
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dsprites_orientation.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Kitti_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Structured/kitti_distance.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Smallnorb_azimuth:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Structured/smallnorb_azimuth.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
Smallnorb_evaluation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file configs/prompt/prompt_test/VTAB-1k/Structured/smallnorb_evaluation.yaml MODEL.TRANSFER_TYPE "P_VK" MODEL.PROMPT.NUM_TOKENS "10"
# SELF_QKV + bias approach
CUDA_VISIBLE_DEVICES=1 PORT=20000 python train.py --config-file /home/ch7858/vpt/configs/prompt/prompt_test/VTAB-1k/Natural/cifar100.yaml MODEL.TRANSFER_TYPE "SELF_QKV+bias" MODEL.PROMPT.NUM_TOKENS "10"
%%bash
# Tune VTAB-caltech101 with VPT:
python tune_vtab.py \
--train-type "QKV" \
--config-file configs/prompt/cub.yaml \
MODEL.TYPE "vit" \
DATA.BATCH_SIZE "128" \
MODEL.PROMPT.DEEP "True" \
MODEL.PROMPT.DROPOUT "0.1" \
MODEL.PROMPT.NUM_TOKENS "10" \
DATA.FEATURE "sup_vitb16_imagenet21k" \
DATA.NAME "vtab-caltech101" \
DATA.NUMBER_CLASSES "102" \
DATA.DATAPATH \
MODEL.MODEL_ROOT \
OUTPUT_DIR
# Tune CUB with VPT:
python tune_fgvc.py \
--train-type "QKV" \
--config-file configs/prompt/prompt_test/VTAB-1k/Natural/cifar100.yaml \
MODEL.TYPE "vit" \
DATA.BATCH_SIZE "128" \
MODEL.QKV_insert.DEEP "True" \
MODEL.QKV_insert.DROPOUT "0.1" \
MODEL.QKV_insert.NUM_TOKENS "10" \
DATA.FEATURE "sup_vitb16_imagenet21k" \
DATA.DATAPATH \
MODEL.MODEL_ROOT \
OUTPUT_DIR
##### P_VK approach tune_command commands
# FGVC
Cars:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/cars.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "20" MODEL.P_VK.NUM_TOKENS "20" MODEL.P_VK.SHARE_PARAM_KV "True"
CUB:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/cub.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.SHARED_ACCROSS "True"
Dogs:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/dogs.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "100" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True"
Flowers:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/flowers.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True"
Nabirds:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/nabirds.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True"
# VTAB-1k
## Natural
Caltech101:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/caltech101.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "32" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Cifar100:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/cifar100.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
dtd:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/dtd.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_flowers102:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/oxford_flowers102.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "1" MODEL.P_VK.NUM_TOKENS "1" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_iiit_pet:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/oxford_iiit_pet.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
SUN397:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/sun397.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Svhn_cropped:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/svhn_cropped.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
## Specialized
Diabetic_retinopathy_detection:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/diabetic_retinopathy_detection.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Eurosat:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/eurosat.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Patch_camelyon:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/patch_camelyon.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "10" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Resisc45:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/resisc45.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
## Structured
Clevr_count:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/clevr_count.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Clevr_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/clevr_distance.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dmlab:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dmlab.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_location:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dsprites_location.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_orientation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dsprites_orientation.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "20" MODEL.P_VK.NUM_TOKENS "20" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Kitti_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/kitti_distance.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_azimuth:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/smallnorb_azimuth.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_evaluation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/smallnorb_evaluation.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
# FGVC final runs(contain 5 runs)
# cub
CUDA_VISIBLE_DEVICES=1 PORT=20000 python five_runs_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/cub.yaml
# cars
CUDA_VISIBLE_DEVICES=0 PORT=20000 python five_runs_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/cars.yaml
# dogs
CUDA_VISIBLE_DEVICES=0 PORT=20000 python five_runs_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/dogs.yaml
# flowers
CUDA_VISIBLE_DEVICES=1 PORT=20000 python five_runs_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/flowers.yaml
# nabirds
CUDA_VISIBLE_DEVICES=1 PORT=20000 python five_runs_fgvc.py --train-type "P_VK" --config-file configs/prompt/prompt_test/FGVC/nabirds.yaml
****************************************************************
########## Documented Version (Origin 5runs on VTAB-1k)
# VTAB-1k
## Natural
Caltech101:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/caltech101.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "32" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Cifar100:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/cifar100.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
dtd:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/dtd.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_flowers102:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/oxford_flowers102.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "1" MODEL.P_VK.NUM_TOKENS "1" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_iiit_pet:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/oxford_iiit_pet.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
SUN397:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/sun397.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Svhn_cropped:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/svhn_cropped.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
## Specialized
Diabetic_retinopathy_detection:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/diabetic_retinopathy_detection.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Eurosat:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/eurosat.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Patch_camelyon:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/patch_camelyon.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "10" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Resisc45:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/resisc45.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
## Structured
Clevr_count:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/clevr_count.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Clevr_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/clevr_distance.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dmlab:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dmlab.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_location:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dsprites_location.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_orientation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dsprites_orientation.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Kitti_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/kitti_distance.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_azimuth:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/smallnorb_azimuth.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_evaluation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/smallnorb_evaluation.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
########## Documented Version (Origin 5runs on FGVC)
CUB
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_fiveRuns.py --config-file configs/prompt/prompt_test/FGVC/cub.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False"
NA Birds
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_fiveRuns.py --config-file configs/prompt/prompt_test/FGVC/nabirds.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False"
Oxford_flowers102
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_fiveRuns.py --config-file configs/prompt/prompt_test/FGVC/flowers.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False"
Stanford Dogs
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_fiveRuns.py --config-file configs/prompt/prompt_test/FGVC/dogs.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False"
Stanford Cars
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_fiveRuns.py --config-file configs/prompt/prompt_test/FGVC/cars.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.MASK_CLS_TOKEN "False"
***************************************************************************************
########## Documented Version (Pruning and Rewind on VTAB-1k)
# VTAB-1k
## Natural
Caltech101:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/caltech101.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "32" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Cifar100:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/cifar100.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
dtd:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/dtd.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_flowers102:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/oxford_flowers102.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "1" MODEL.P_VK.NUM_TOKENS "1" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_iiit_pet:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/oxford_iiit_pet.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
SUN397:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/sun397.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Svhn_cropped:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/svhn_cropped.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
## Specialized
Diabetic_retinopathy_detection:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/diabetic_retinopathy_detection.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Eurosat:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/eurosat.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Patch_camelyon:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/patch_camelyon.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "10" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Resisc45:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Specialized/resisc45.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
## Structured
Clevr_count:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/clevr_count.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Clevr_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/clevr_distance.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dmlab:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dmlab.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_location:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dsprites_location.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_orientation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/dsprites_orientation.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Kitti_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/kitti_distance.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_azimuth:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/smallnorb_azimuth.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_evaluation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Structured/smallnorb_evaluation.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
########## Documented Version (Pruning and Rewind on FGVC)
CUB
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_test/FGVC/cub.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
NA Birds
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_test/FGVC/nabirds.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Oxford_flowers102
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_test/FGVC/flowers.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Stanford Dogs
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_test/FGVC/dogs.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Stanford Cars
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_test/FGVC/cars.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
**********************************************************************************
########## Documented Version (ssl mae VTAB-1k)
## Natural
Caltech101:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Natural/caltech101.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "32" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Cifar100:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Natural/cifar100.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
dtd:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Natural/dtd.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_flowers102:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Natural/oxford_flowers102.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "1" MODEL.P_VK.NUM_TOKENS "1" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_iiit_pet:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Natural/oxford_iiit_pet.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
SUN397:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Natural/sun397.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Svhn_cropped:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Natural/svhn_cropped.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
## Specialized
Diabetic_retinopathy_detection:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Specialized/diabetic_retinopathy_detection.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Eurosat:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Specialized/eurosat.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Patch_camelyon:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Specialized/patch_camelyon.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "10" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Resisc45:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Specialized/resisc45.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
## Structured
Clevr_count:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Structured/clevr_count.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Clevr_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Structured/clevr_distance.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dmlab:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Structured/dmlab.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_location:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Structured/dsprites_location.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_orientation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Structured/dsprites_orientation.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Kitti_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Structured/kitti_distance.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_azimuth:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Structured/smallnorb_azimuth.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_evaluation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_mae/VTAB-1k/Structured/smallnorb_evaluation.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
########## Documented Version (ssl mae FGVC)
CUB
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_mae/FGVC/cub.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
NA Birds
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_mae/FGVC/nabirds.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Oxford_flowers102
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_mae/FGVC/flowers.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Stanford Dogs
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_mae/FGVC/dogs.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Stanford Cars
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_mae/FGVC/cars.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
**********************************************************************************
########## Documented Version (ssl moco VTAB-1k)
## Natural
Caltech101:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Natural/caltech101.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "32" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Cifar100:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Natural/cifar100.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
dtd:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Natural/dtd.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_flowers102:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Natural/oxford_flowers102.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "1" MODEL.P_VK.NUM_TOKENS "1" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Oxford_iiit_pet:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Natural/oxford_iiit_pet.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
SUN397:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Natural/sun397.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
Svhn_cropped:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Natural/svhn_cropped.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Natural/"
## Specialized
Diabetic_retinopathy_detection:
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Specialized/diabetic_retinopathy_detection.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Eurosat:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Specialized/eurosat.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Patch_camelyon:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Specialized/patch_camelyon.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "10" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
Resisc45:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Specialized/resisc45.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Specialized/"
## Structured
Clevr_count:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Structured/clevr_count.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Clevr_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Structured/clevr_distance.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dmlab:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Structured/dmlab.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_location:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Structured/dsprites_location.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Dsprites_orientation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Structured/dsprites_orientation.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Kitti_distance:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Structured/kitti_distance.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_azimuth:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Structured/smallnorb_azimuth.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
Smallnorb_evaluation:
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_ssl_moco/VTAB-1k/Structured/smallnorb_evaluation.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "128" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "10" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" DATA.DATAPATH "/home/zhiwenc/data/VTAB-1k/Structured/"
########## Documented Version (ssl moco FGVC)
CUB
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_moco/FGVC/cub.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "64" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
NA Birds
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_moco/FGVC/nabirds.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "64" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Oxford_flowers102
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_moco/FGVC/flowers.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "64" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Stanford Dogs
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_moco/FGVC/dogs.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "64" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
Stanford Cars
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_fgvc_PruningRewind.py --config-file configs/prompt/prompt_ssl_moco/FGVC/cars.yaml MODEL.TYPE "ssl-vit" DATA.BATCH_SIZE "64" MODEL.TRANSFER_TYPE "P_VK" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "5" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True"
## for 384 resolution
add DATA_CROPSIZE "384" to the end of the command, notice for OOM problem (By default, all settings are under batchsize=32)
## for full fine-tune (FGVC for visualization propose)
CUB
python train.py --config-file configs/finetune/cub.yaml DATA.BATCH_SIZE "128" SOLVER.BASE_LR "0.0001" SOLVER.WEIGHT_DECAY "0.0001"
NA Birds
python train.py --config-file configs/finetune/nabirds.yaml DATA.BATCH_SIZE "128" SOLVER.BASE_LR "0.0001" SOLVER.WEIGHT_DECAY "0.0001"
Oxford_flowers102
python train.py --config-file configs/finetune/flowers.yaml DATA.BATCH_SIZE "128" SOLVER.BASE_LR "0.001" SOLVER.WEIGHT_DECAY "0.0001"
Stanford Dogs
python train.py --config-file configs/finetune/dogs.yaml DATA.BATCH_SIZE "128" SOLVER.BASE_LR "0.005" SOLVER.WEIGHT_DECAY "0.0"
Stanford Cars
python train.py --config-file configs/finetune/cars.yaml DATA.BATCH_SIZE "128" SOLVER.BASE_LR "0.0005" SOLVER.WEIGHT_DECAY "0.01"
## get results
python get_results_PruningRewind_CVK.py
## finetune command
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab.py --train-type "finetune" --config-file configs/finetune/VTAB-1k/Structured/clevr_distance.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "128"
## get results for imagenet prompt tuning (saved for NIPS)
python tune_vtab_PruningRewind_CVK_imagenet1k.py --train-type "P_VK" --config-file configs/prompt/prompt_test/imagenet1k.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "1" MODEL.P_VK.NUM_TOKENS "1" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.P_VK.QUERY_PROMPT_MODE "0"
(under construction)
# python train.py --config-file configs/finetune/imagenet1k.yaml DATA.BATCH_SIZE "128" SOLVER.BASE_LR "0.0001" SOLVER.WEIGHT_DECAY "0.0001"
# calculate IG
python tune_vtab_AS.py --train-type "prompt" --config-file configs/prompt/cifar100_forVPT.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "A_test_as" DATA.BATCH_SIZE "64"
python tune_vtab_AS.py --train-type "prompt" --config-file configs/prompt/prompt_vpt/Natural/cifar100_forVPT.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "A_test_as" DATA.BATCH_SIZE "64"
python tune_vtab_AS.py --train-type "finetune" --config-file configs/finetune/VTAB-1k/Natural/cifar100.yaml OUTPUT_DIR "A_test_as_finetune" DATA.BATCH_SIZE "128"
CUDA_VISIBLE_DEVICES=0 PORT=20000 python tune_vtab_AS.py --train-type "finetune" --config-file configs/finetune/VTAB-1k/Natural/cifar100.yaml OUTPUT_DIR "A_test_as_finetune_general" DATA.BATCH_SIZE "64" ATTRIBUTION_TYPE "general"
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_AS.py --train-type "prompt" --config-file configs/prompt/prompt_vpt/Natural/cifar100_forVPT.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "IG_VIS" DATA.BATCH_SIZE "64" ATTRIBUTION_TYPE "general" ATTRIBUTION_INTEGRATED_METHOD "ig"
# example command for visualization
CUDA_VISIBLE_DEVICES=1 PORT=20000 python tune_vtab_AS.py --train-type "prompt" --config-file configs/prompt/prompt_vpt/Specialized/diabetic_retinopathy_detection.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "Retino_vis_prompt" DATA.BATCH_SIZE "64" ATTRIBUTION_TYPE "general" ATTRIBUTION_INTEGRATED_METHOD "pytorch_gradcam"
CUDA_VISIBLE_DEVICES=0 PORT=30000 python tune_vtab_AS.py --train-type "finetune" --config-file configs/finetune/VTAB-1k/Specialized/diabetic_retinopathy_detection.yaml OUTPUT_DIR "Retino_vis_finetune" DATA.BATCH_SIZE "128" ATTRIBUTION_TYPE "general" ATTRIBUTION_INTEGRATED_METHOD "pytorch_gradcam"
# example command for few-shot learning
CUDA_VISIBLE_DEVICES=0 PORT=40000 python tune_vtab.py --train-type "prompt" --config-file configs/prompt/prompt_vpt/Structured/dsprites_orientation.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "50" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "fewshot_prompt" DATA.BATCH_SIZE "64" DATA.EVEN_SEPARETE "True"
CUDA_VISIBLE_DEVICES=1 PORT=50000 python tune_vtab.py --train-type "finetune" --config-file configs/finetune/VTAB-1k/Structured/dsprites_orientation.yaml OUTPUT_DIR "fewshot_finetune" DATA.BATCH_SIZE "128" DATA.EVEN_SEPARETE "True"
# example command for swin finetune
python tune_vtab.py --train-type "finetune" --config-file configs/finetune/VTAB-1k_swin/Natural/cifar100.yaml MODEL.TYPE "swin" DATA.BATCH_SIZE "128" OUTPUT_DIR "DatasetChanged_swin8000-2000FineTune"
# example command for swin vpt origin
python tune_vtab.py --train-type "prompt" --config-file configs/prompt/prompt_swin/VTAB-1k_forVPT/Natural/cifar100.yaml MODEL.TYPE "swin" DATA.BATCH_SIZE "80" MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "test_dataset"
# ft + pt(two steps)
python tune_vtab.py --train-type "prompt" --config-file configs/prompt/prompt_vpt/Structured/clevr_distance.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "200" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "ft_pt" DATA.BATCH_SIZE "64"
# ft + pt(one step) ## run!
python tune_vtab.py --train-type "prompt" --config-file configs/prompt/prompt_vpt/Natural/cifar100_forVPT.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "ft_pt_mixed" DATA.BATCH_SIZE "64" MODEL.PROMPT.FT_PT_MIXED "True"
# pt (original) ## run!
python tune_vtab.py --train-type "prompt" --config-file configs/prompt/prompt_vpt/Natural/cifar100_forVPT.yaml MODEL.PROMPT.DEEP "True" MODEL.PROMPT.NUM_TOKENS "10" MODEL.PROMPT.DROPOUT "0.1" OUTPUT_DIR "pt_origin" DATA.BATCH_SIZE "64"
# ft + pt(one step) for E2VPT
python tune_vtab_PruningRewind_CVK.py --train-type "P_VK" --config-file configs/prompt/prompt_test/VTAB-1k/Natural/caltech101.yaml MODEL.TYPE "vit" DATA.BATCH_SIZE "64" MODEL.P_VK.DEEP "True" MODEL.P_VK.DEEP_P "True" MODEL.P_VK.DROPOUT_P "0.1" MODEL.P_VK.NUM_TOKENS_P "32" MODEL.P_VK.NUM_TOKENS "5" MODEL.P_VK.ORIGIN_INIT "2" MODEL.P_VK.SHARE_PARAM_KV "True" MODEL.PROMPT.FT_PT_MIXED "True"