2764 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
Ross Wightman	ea728f67fa	Improve several typing issues for flex vit, can (almost) work with jit if we bash h,w key into an int or str	2025-04-14 11:01:56 -07:00
Ross Wightman	681be882e8	Fix arg merging of sknet, old seresnet. Fix #2470	2025-04-14 10:32:26 -07:00
Ross Wightman	97341fec51	A much faster resample_patch_embed, can be used at train/validation time	2025-04-10 15:58:24 -07:00
Ross Wightman	b4bb0f452a	Exclude embeds module and mask attn functions from tracing	2025-04-09 15:34:15 -07:00
Ross Wightman	13e0f3a4a3	Add loss scale arg, initial distributed loss scale. Maybe fix FX for the model.	2025-04-08 20:47:57 -07:00
Ross Wightman	6675590264	Fix ParallelThingsBlock w/ attn_mask	2025-04-08 09:35:34 -07:00
Ross Wightman	9b23d6dea2	Exclude naflex models from jit tests	2025-04-08 07:59:19 -07:00
Ross Wightman	825edccf19	Type fixes, remove old comments	2025-04-07 21:35:03 -07:00
Ross Wightman	0893f5d296	Initial NaFlex ViT model and training support	2025-04-07 21:27:10 -07:00
Ross Wightman	e44f14d7d2	Update README v1.0.15	2025-02-22 21:04:13 -08:00
Ross Wightman	98e9651952	Update version.py ... Version 1.0.15, prep for a release	2025-02-22 10:50:21 -08:00
Ross Wightman	e76ea5474d	Update README.md	2025-02-21 16:09:42 -08:00
Adam J. Stewart	92682d8d4d	timm.models: explicitly export attributes	2025-02-21 14:19:39 -08:00
Ross Wightman	a667d3d8f0	siglip2 weights on hub, fix forward_intermediates when no prefix tokens (& return prefix selected)	2025-02-21 13:10:51 -08:00
Ross Wightman	f63a11cf81	Remove duplicate so400m/16 @ 256 model def	2025-02-21 13:10:51 -08:00
Ross Wightman	9758e0b8b0	Prep for siglip2 release	2025-02-21 13:10:51 -08:00
Adam J. Stewart	c68d724e9c	adapt_input_conv: add type hints	2025-02-21 12:28:22 -08:00
Ross Wightman	105a667baa	Dev version 1.0.15.dev0	2025-02-17 15:50:12 -08:00
Ross Wightman	7234f5c6c5	Add 448 so150m2 weight/model, add updated internvit 300m weight	2025-02-17 12:59:10 -08:00
Ross Wightman	9ce824c39a	Add vit so150m2 weights	2025-02-14 15:55:51 -08:00
Ross Wightman	a49b020eff	Merge branch 'ClashLuke-patch-1'	2025-01-31 12:53:29 -08:00
Ross Wightman	490d222dd8	Fix issue taking device from V before V exists	2025-01-31 12:52:47 -08:00
Ross Wightman	875c19d0c9	Merge branch 'patch-1' of github.com:ClashLuke/pytorch-image-models into ClashLuke-patch-1	2025-01-31 12:43:28 -08:00
Ross Wightman	8b3c07a841	Update README.md	2025-01-31 10:37:32 -08:00
Lucas Nestler	e025328f96	simplify RNG	2025-01-31 17:26:14 +01:00
Lucas Nestler	6367267298	unify RNG	2025-01-31 17:23:53 +01:00
Ross Wightman	872978ccfe	Fix comment, add 'stochastic weight decay' idea because why not	2025-01-30 18:22:36 -08:00
Ross Wightman	510bbd5389	Change start/end args	2025-01-30 18:22:36 -08:00
Ross Wightman	31831f5948	Change flattening behaviour in Kron	2025-01-30 18:22:36 -08:00
Ross Wightman	cdbafd9057	Try to force numpy<2.0 for torch 1.13 tests, update newest tested torch to 2.5.1	2025-01-28 20:56:30 -08:00
Ross Wightman	b1752eefb5	Fix missing model key in bulk validate results on error	2025-01-28 13:20:40 -08:00
Ross Wightman	b3a83b81d6	Prep Kron for merge, add detail to attributions note, README.	2025-01-27 21:02:26 -08:00
Ross Wightman	67ef6f0a92	Move opt_einsum import back out of class __init__	2025-01-27 21:02:26 -08:00
Ross Wightman	9ab5464e4d	More additions to Kron	2025-01-27 21:02:26 -08:00
Ross Wightman	5f10450235	Some more kron work. Figured out why some tests fail, implemented a deterministic rng state load but too slow so skipping some tests for now.	2025-01-27 21:02:26 -08:00
Ross Wightman	cd21e80d03	Fiddling with Kron (PSGD)	2025-01-27 21:02:26 -08:00
Adam J. Stewart	d81da93c16	Use import alias	2025-01-22 10:27:17 -08:00
Adam J. Stewart	4de1abf837	timm: add __all__ to __init__	2025-01-22 10:27:17 -08:00
Ryan	bda46f8e6f	Add num_classes assertion after reset_classifier	2025-01-21 11:52:05 -08:00
Ryan	17eabaad17	Fix RDNet forward call	2025-01-21 11:52:05 -08:00
Ryan	80a4877376	Fix self.reset_classifier num_classes update	2025-01-21 11:52:05 -08:00
Collin McCarthy	84631cb5c6	Add missing training flag to convert_sync_batchnorm	2025-01-21 11:51:55 -08:00
Josua Rieder	cb4cea561a	add arguments to the respective argument groups	2025-01-20 10:54:35 -08:00
Josua Rieder	634b68ae50	Fix metavar for --input-size	2025-01-20 10:53:46 -08:00
Ross Wightman	5d535d7a2d	Version 1.0.14, update README & changelog v1.0.14	2025-01-19 13:53:09 -08:00
Ross Wightman	c6b74eb5bd	Remove numpy ver constraint, not relevant for latest PyTorch version	2025-01-19 13:41:16 -08:00
Ross Wightman	aa333079da	Tweak so150m2 def	2025-01-19 13:40:53 -08:00
Josua Rieder	8d81fdf3d9	Fix typos	2025-01-19 13:39:40 -08:00
Ross Wightman	3677f67902	Add the 256x256 in1k ft of the so150m, add an alternate so150m def	2025-01-18 15:51:57 -08:00
Ross Wightman	2a84d68d02	Add some so150m vit w/ sbb recipe weights, and a ese_vovnet57b model with RA4 recipe	2025-01-18 15:51:57 -08:00