15e2a15692
reduce batch size
2026-05-16 15:49:26 +02:00
e0bd9fc7ce
import dist
2026-05-16 12:47:41 +02:00
79fc9564ea
DDP guard if single GPUgio
2026-05-16 11:52:10 +02:00
bad00f306e
reduce batch size because of GPU overload
2026-05-16 11:20:41 +02:00
7101f6c8e1
wilds 16x16/448 config without crop_scale
2026-05-16 11:14:45 +02:00
1094b8e7bc
remove imagenet configs
2026-05-16 11:09:47 +02:00
4df162180e
disable random resize crop + ft config
2026-05-16 11:04:00 +02:00
06f0c3fa78
change eval path
2026-05-16 10:32:15 +02:00
ceeb57fc52
extend readme with eval
2026-05-14 10:55:45 +02:00
d22a3aba90
evaluation pipeline
2026-05-14 10:52:40 +02:00
4a7110ebf2
remove wilds keyword
2026-05-13 08:40:31 +02:00
a4512456e8
delete
2026-05-11 10:04:45 +02:00
11bf44aeed
delete
2026-05-11 10:03:01 +02:00
a315d00fad
readme
2026-05-11 09:56:00 +02:00
75cf127d74
add checkpoint_folder + tweak params for supervised fine tuning + remove default values of params
2026-04-27 11:16:26 +02:00
e0daf4b1aa
Merge branch 'main' of github.com:YannAhlgrim/ijepa
2026-04-27 09:22:12 +02:00
4d2e62c07e
results of ijepa ssl job 73125
2026-04-27 09:21:47 +02:00
77d94c1b50
load checkpoint = true
2026-04-25 11:04:07 +02:00
459d998b20
add early stopping callback
2026-04-25 11:03:05 +02:00
c54816aa93
add evaluation transform method
2026-04-25 11:01:38 +02:00
02991a7eb5
no grad if frozen in forward pass
2026-04-25 11:00:14 +02:00
417319457a
remove image folder
2026-04-25 10:14:22 +02:00
7e32fc994f
return tuple (img, target) even on unlabeled data
2026-04-18 15:02:17 +02:00
24de5b3078
add data and logs
2026-04-18 12:35:21 +02:00
f2129dc28a
Fixed conflicts between cluster and github versions
2026-04-18 12:32:22 +02:00
69d811ecb4
slurm version of main_distributed.py
2026-04-18 12:28:40 +02:00
f513bc97b4
cuda 12.1
2026-04-18 12:21:45 +02:00
4c4d6713e0
add save checkpoint logic
2026-04-15 13:43:05 +02:00
ebad8ee627
experiment logs and venv
2026-04-15 13:23:24 +02:00
0a38a264f7
add supervised training pipeline
2026-04-15 13:22:44 +02:00
e2910cb458
Header for Linear Probing
2026-04-12 11:19:06 +02:00
4ec019f650
submitit for main_distributed.py
2026-04-05 18:46:55 +02:00
8956531fe5
change explicitly for slurm cluster
2026-04-05 18:44:54 +02:00
5da7b0f521
requirements.txt
2026-04-04 12:10:07 +02:00
e49c793cea
wilds vit huge yaml params
2026-04-04 11:23:04 +02:00
4d2fd1021b
integrate make_iwildcam in train.py
2026-04-02 14:15:06 +02:00
5165607f55
wilds dataset loader
2026-04-01 16:26:35 +02:00
f1656deab3
Update README for WILDS-IJEPA project
...
This fork of the official I-JEPA repository focuses on applying the learning method to wild animal images to enhance the generalization capabilities of wildlife image classification models in various locations with limited data.
2026-03-25 12:55:05 +01:00
52c1ae95d0
Initial commit
2023-06-13 13:03:30 +00:00
YannAhlgrim