482 Commits

Author	SHA1	Message	Date
Zhongjie Duan	0b4b337e9a	Merge pull request #933 from lzws/main ... update wan2.2-VACE-Fun-A14B	2025-09-24 09:56:37 +08:00
Zhongjie Duan	32b8b9b51e	Merge pull request #910 from ldiex/main ... Fix gradient checkpointing in WAN VACE blocks	2025-09-23 12:23:12 +08:00
Zhongjie Duan	f6534a5b63	Merge pull request #909 from huarzone/fix_bug ... fix load gif	2025-09-23 12:22:00 +08:00
Zhongjie Duan	034c9b6c60	Qwen-Image-Edit-2509 (#937) ... * qwen-image-edit-2509	2025-09-22 20:37:11 +08:00
lzws	76335e0fe5	uodate wan2.2-VACE-Fun	2025-09-22 02:14:20 +08:00
Artiprocher	7a5974d964	update vram management strategy	2025-09-18 16:51:53 +08:00
Zhongjie Duan	b0abdaffb4	Qwen image split training Bug Fix (#926) ... * bugfix	2025-09-17 20:53:46 +08:00
Artiprocher	1a7f482fbd	support qwen-image-distill-dmd2-lora	2025-09-16 19:43:07 +08:00
Tianlin Pan	3a0d51d100	Fix gradient checkpointing in WAN VACE blocks	2025-09-14 16:21:46 +08:00
Kared	bffdb901ed	fix load gif	2025-09-13 21:01:44 +08:00
xycdx	7e5ce5d5c9	Update diffsynth/extensions/FastBlend/patch_match.py ... Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>	2025-09-10 20:48:54 +08:00
xycdx	7aef554d83	add torch implementation for interpolation ... - Implement bilinear interpolation kernel using Numba - Benchmark shows 2x speedup compared to CPU version - Closes #817	2025-09-10 20:39:35 +08:00
Artiprocher	d6cf20ef33	support direct distill	2025-09-09 16:12:31 +08:00
Zhongjie Duan	efdd6a59b6	Merge pull request #892 from modelscope/dev2-dzj ... refine training framework	2025-09-04 15:53:52 +08:00
Artiprocher	144365b07d	merge data process to training script	2025-09-04 15:18:56 +08:00
Artiprocher	cb8de6be1b	move training code to base trainer	2025-09-03 12:03:49 +08:00
Zhongjie Duan	8c13362dcf	Merge pull request #884 from modelscope/dev2-dzj ... Unified Dataset & Splited Training	2025-09-03 09:50:23 +08:00
Artiprocher	b6da77e468	qwen-image splited training	2025-09-02 16:44:14 +08:00
Artiprocher	260e32217f	unified dataset	2025-09-02 13:14:08 +08:00
mi804	5cee326f92	support s2v framepack	2025-09-01 16:48:46 +08:00
Zhongjie Duan	826bab5962	Merge pull request #859 from krahets/main ... Fix batch decoding for Wan-Video-VAE	2025-08-29 12:45:49 +08:00
Zhongjie Duan	5b6d112c15	Merge pull request #843 from wuutiing/main ... add read gifs as video support	2025-08-29 12:36:24 +08:00
Zhongjie Duan	0a78bb9d38	Merge pull request #864 from modelscope/wans2v ... Support Wan-S2V	2025-08-28 10:21:12 +08:00
mi804	9cea10cc69	minor fix	2025-08-28 10:13:52 +08:00
mi804	fdeb363fa2	wans2v usp	2025-08-27 19:50:33 +08:00
mi804	4147473c81	wans2v refactor	2025-08-27 16:18:22 +08:00
mi804	8a0bd7c377	wans2v lowvram	2025-08-27 13:05:53 +08:00
mi804	b541b9bed2	wans2v inference	2025-08-27 11:51:56 +08:00
Yudong Jin	419d47c195	Remove unnecessary newline in encode method	2025-08-27 02:24:29 +08:00
Yudong Jin	ac2e859960	Fix batch decoding for Wan VAE.	2025-08-27 02:24:00 +08:00
Zhongjie Duan	8fcfa1dd2d	bugfix	2025-08-26 17:22:25 +08:00
Artiprocher	ce0b948655	support qwen-image fp8 lora training	2025-08-25 20:32:36 +08:00
lzws	c795e35142	add wan2.2-fun-A14B inp, control and control-camera (#839) ... * update wan2.2-fun * update wan2.2-fun * update wan2.2-fun * add examples * update wan2.2-fun * update wan2.2-fun * Rename Wan2.2-Fun-A14B-Inp.py to Wan2.2-Fun-A14B-InP.py --------- Co-authored-by: lzw478614@alibaba-inc.com <lzw478614@alibaba-inc.com>	2025-08-22 14:20:31 +08:00
mi804	8dae8d7bc8	fix blockwise controlnet training by avoid inplace	2025-08-22 11:28:57 +08:00
twu	f6418004bb	as numframe limit is impled in reader, add that	2025-08-22 03:00:35 +00:00
twu	e3f47a799b	make it more efficient to locate where to sample the frame	2025-08-21 09:13:45 +00:00
twu	e014cad820	add read gifs as video support	2025-08-21 09:01:48 +00:00
Zhongjie Duan	89bf3ce5cf	Merge pull request #841 from modelscope/qwen-image-lora-hotload ... support qwen-image lora hotload	2025-08-21 15:14:46 +08:00
Artiprocher	7f719cefe6	refine code	2025-08-21 14:25:17 +08:00
Artiprocher	613dafbd09	rename model	2025-08-21 13:35:47 +08:00
Artiprocher	6ab426e641	support qwen-image lora hotload	2025-08-21 10:12:52 +08:00
mi804	5e6f9f89f1	support eligenv2 and context_control	2025-08-20 22:48:34 +08:00
Zhongjie Duan	9ec0652339	Merge pull request #829 from mi804/qwen-image-edit-autoresize ... support edit_image_auto_resize	2025-08-20 13:40:02 +08:00
mi804	29b12b2f4e	support edit_image_auto_resize	2025-08-20 12:36:26 +08:00
mi804	c9fea729d8	fix qwen-image-edit-lowvram	2025-08-20 10:31:43 +08:00
mi804	838b8109b1	support qwen-image-edit lowres fix	2025-08-19 20:15:36 +08:00
mi804	fff2c89360	fix bug	2025-08-19 12:38:33 +08:00
mi804	123f6dbadb	update lora and full train	2025-08-18 19:09:19 +08:00
Hong Zhang	f9ce261a0e	Merge branch 'main' into qwen-image-edit	2025-08-18 18:56:26 +08:00
mi804	d93de98a21	fix qwen_rope	2025-08-18 17:31:18 +08:00