Update README.md
Browse files
README.md
CHANGED
@@ -26,18 +26,13 @@ No extra meta-data was used for training these models (as was the case for the c
|
|
26 |
## Fine-tune hparams
|
27 |
|
28 |
```
|
29 |
-
./distributed_train.sh 4 --data-dir /tfds/ --dataset tfds/i_naturalist2021 --amp -j 8 --model vit_large_patch14_clip_224 --img-size 336 --model-kwargs img_size=336 --val-split val --opt adamw --opt-eps 1e-6 --weight-decay .01 --lr 5e-5 -
|
30 |
-
-warmup-lr 0 --sched-on-updates --clip-grad 1.0 --pretrained -b 48 --num-classes 10000 --grad-accum-steps 8 --layer-decay 0.
|
31 |
-
```
|
32 |
-
|
33 |
-
```
|
34 |
-
./distributed_train.sh 4 --data-dir /tfds/ --dataset tfds/i_naturalist2021 --amp -j 8 --model eva02_large_patch14_clip_336 --val-split val --opt adamw --opt-eps 1e-6 --weight-decay .01 --lr 5e-5 --warmup-lr 0 --sched-on-updates --clip-gra
|
35 |
-
d 1.0 --pretrained -b 40 --num-classes 10000 --grad-accum-steps 10 --layer-decay 0.8 --torchcompile
|
36 |
```
|
37 |
|
38 |
## Run Validation
|
39 |
```
|
40 |
-
python validate.py /tfds/ --dataset tfds/i_naturalist2021 --model hf-hub:timm/
|
41 |
```
|
42 |
|
43 |
## Citation
|
|
|
26 |
## Fine-tune hparams
|
27 |
|
28 |
```
|
29 |
+
./distributed_train.sh 4 --data-dir /tfds/ --dataset tfds/i_naturalist2021 --amp -j 8 --model vit_large_patch14_clip_224.datacompxl --img-size 336 --model-kwargs img_size=336 --val-split val --opt adamw --opt-eps 1e-6 --weight-decay .01 --lr 5e-5 -
|
30 |
+
-warmup-lr 0 --sched-on-updates --clip-grad 1.0 --pretrained -b 48 --num-classes 10000 --grad-accum-steps 8 --layer-decay 0.85 --reprob 0.3 --drop-path 0.1
|
|
|
|
|
|
|
|
|
|
|
31 |
```
|
32 |
|
33 |
## Run Validation
|
34 |
```
|
35 |
+
python validate.py /tfds/ --dataset tfds/i_naturalist2021 --model hf-hub:timm/vit_large_patch14_clip_336.datacompxl_ft_augreg_inat21 --split val --amp
|
36 |
```
|
37 |
|
38 |
## Citation
|