yanis/LTX-Video
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
89 Commits 1 Branch 0 Tags
main
Commit Graph

89 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Yoav HaCohen
4b2d053057 Merge pull request #261 from Lightricks/release/ltx2-redirect-notice
Some checks failed
Ruff / build (3.10) (push) Has been cancelled
Details 
Readme: Link to LTX-2 with synchronized audio+video support
 2026-01-06 00:37:06 +02:00
Yoav HaCohen
e7e7fb4bd7 Readme: Add notice directing users to LTX-2 with synchronized audio+video support 2026-01-06 00:29:56 +02:00
Yoav HaCohen
5260738e17 Merge pull request #251 from rluxemburg/main 
Revise README with updated links and LTX-2 info
 2025-10-25 21:02:27 +03:00
Rachel Luxemburg
f5879832da Revise README with updated links and LTX-2 info 
Updated links and descriptions in the README for LTX-Video, including LTX-2 model announcements and feature enhancements.
 2025-10-23 09:18:58 -04:00
Yoav HaCohen
53d263f317 Merge pull request #224 from LightricksResearch/public/feature/fp8 
Inference: Integrate fp8 kernels
 2025-07-21 17:44:17 +03:00
Yaki
ccfc030932 Inference: Integrate fp8 kernels 2025-07-21 13:55:51 +03:00
Jacob Bitterman
bdc8f017f0 Merge pull request #221 from LightricksResearch/release/v0-9-8 
LTXV-0.9.8: Long shot generation
 2025-07-16 17:50:19 +03:00
Yoav HaCohen
02844cec63 LTXV-0.9.8: Long shot generation 2025-07-16 17:47:04 +03:00
Jacob Bitterman
20799e51cd Merge pull request #221 from Lightricks/readme/broken_link 
Readme: Fix broken link
 2025-07-08 17:45:54 +03:00
Jacob Bitterman
9f50566718 Readme: Fix broken link 2025-07-08 17:44:58 +03:00
Jacob Bitterman
e29ca10f82 Merge pull request #220 from Lightricks/feature/readme-release-ic-loras 
Readme: Add official Control Models
 2025-07-08 16:36:13 +03:00
Yaki
71d4b5f920 Readme: Add official Control Models 2025-07-08 12:25:56 +00:00
Yoav HaCohen
d63732d105 Merge pull request #209 from LightricksResearch/sync_public_main 
sync: Update tests, fp8 configs, and minor fixes
 2025-07-04 15:33:15 +03:00
Yaki
98a0ce6a69 sync: Update tests, fp8 configs, and minor fixes 2025-07-02 11:09:51 +00:00
Yoav HaCohen
34625c3a1f Merge pull request #185 from Lightricks/feature/readme-fix 
README: Fix links
 2025-05-15 18:04:16 +03:00
Yoav HaCohen
b6b5f11c3d Readme: Fix broken comfy links 2025-05-15 18:01:38 +03:00
Yoav HaCohen
d866cc2156 Readme: Update link to a comfy flow 2025-05-14 18:18:06 +03:00
Yoav HaCohen
003f0e5294 Merge pull request #181 from Lightricks/feature/13b-distilled 
Models: Release LTXV-13B-distilled
 2025-05-14 17:55:13 +03:00
Yoav HaCohen
6ff652c3c3 Models: Release LTXV-13B-distilled 2025-05-14 16:54:22 +03:00
Yoav HaCohen
832e043843 Merge pull request #180 from Lightricks/feature/enhanced-pipeline 
Pipeline: Enhance inference pipelines with new features
 2025-05-14 14:37:54 +03:00
Yoav HaCohen
93af6864b9 Pipeline: Enhance inference pipelines with new features 
* Adaptive normalization after latent upsampling
* CFG Star Rescale
* Varying STG/CFG parameters per step
* Support skipping the initial and/or the final diffusion steps
* CRF compression for image condition (useful for getting more motion in image-to-video)
 2025-05-14 12:25:21 +03:00
Yoav HaCohen
cb6f842770 Merge pull request #179 from Lightricks/feature/update-readme 
README: Update assets and training section
 2025-05-14 09:56:26 +03:00
Yoav HaCohen
39c98ce833 README: Update assets and training section 2025-05-14 09:30:06 +03:00
Yoav HaCohen
6e45836dda Merge pull request #157 from Lightricks/feature/fix-pyproject 
Install: Fix install to include ltx_video directory only
 2025-05-07 13:03:45 +03:00
Yoav HaCohen
b3c857a48e Install: Fix install to include ltx_video directory only 2025-05-07 13:02:28 +03:00
Yoav HaCohen
27de2ef3a2 Merge pull request #154 from Lightricks/feature/readme-fix-links 
Readme: Fix links to ComfyUI workflow
 2025-05-06 20:22:29 +03:00
Yoav HaCohen
2d1132fc3e Readme: Fix links to ComfyUI workflow 2025-05-06 20:19:56 +03:00
Yoav HaCohen
3460d35857 Merge pull request #153 from Lightricks/release/v0-9-7 
Inference: Support multiscale pipeline and new models
 2025-05-06 15:20:49 +03:00
Yoav HaCohen
69213dba5f Inference: Support multiscale pipeline and new models 2025-05-06 15:19:00 +03:00
Yoav HaCohen
8b0e71a7d8 Merge pull request #147 from Lightricks/feature/update-model-license 
Readme: Update license
 2025-04-24 00:35:38 +03:00
Yaki
c8206214c6 Readme: Update license 2025-04-24 00:34:07 +03:00
Yoav HaCohen
b1aeddd7cc Merge pull request #140 from Lightricks/release/distilled-model 
Inference: Support distilled model and update default params
 2025-04-17 17:58:18 +03:00
Yoav HaCohen
eab705c649 Inference: Support distilled model and update default params 
This commit adds supports for the distilled model.
It updates the default params for higher resolution and FPS
as the distilled model is still fast with these params.
 2025-04-17 17:55:04 +03:00
Yoav HaCohen
496dc5058f Merge pull request #123 from Lightricks/release/keyframe-and-sequence-conditioning 2025-03-05 15:59:20 +02:00
Yoav HaCohen
956a53b478 Readme: Add change log with new features 2025-03-05 15:12:10 +02:00
Yoav HaCohen
e7ec4d0445 Pipeline: Support prompt enhancement 
In inference.py prompt enhancement is configured as a function of the given prompt length
 2025-03-05 15:01:51 +02:00
Yoav HaCohen
7360d9463a Pipeline: Increase the maximum t5 sequence length 
This commit increases the maximum sequence length of the T5 embeddings
from 128 to 256.
 2025-03-03 23:32:02 +02:00
Yoav HaCohen
23e83b15b4 VAE: Support new architectures 
This commit adds support for:
* Custom spatial padding mode
* Constant latent_log_var
* Normalize latents using a BatchNorm layer
* Additoinal down/up layer options
 2025-03-03 23:25:52 +02:00
Yoav HaCohen
6189c45aa4 Pipeline: Support keyframe and sequence conditioning 2025-03-03 23:25:52 +02:00
Yoav HaCohen
84e36dbbee Merge pull request #115 from Lightricks/public/feature/stg-attention-values 
STG: Add a new STG strategy: AttentionValues (default)
 2025-02-16 16:57:31 +02:00
Yoav HaCohen
9d0df65a83 STG: Add a new STG strategy: AttentionValues (default) 
This new strategy returns V(x) from the skipped attention blocks.
 2025-02-16 16:31:36 +02:00
Yoav HaCohen
f5a27c94bf Merge pull request #114 from Lightricks/feature/more-inference-options 
LTXVideo: Add more inference options
 2025-02-16 00:12:37 +02:00
Yoav HaCohen
cec4fef279 README: Add a link to TeaCache 2025-02-16 00:09:22 +02:00
Yoav HaCohen
18575687ef README: Add a link to ltx-videoq8 2025-02-16 00:09:22 +02:00
Yoav HaCohen
b6da4d0d23 README: Add a link to Replicate 2025-02-16 00:09:22 +02:00
Yoav HaCohen
ef66abc8fb README: Update checkpoint link to v0.9.1 2025-02-16 00:09:22 +02:00
Yoav HaCohen
dfa9be0448 Tests: Add some tests 2025-02-16 00:09:22 +02:00
Yoav HaCohen
4ba32692bf inference: Refactor to expose a simple infer function 
This enables tests and other apps to call the infer function directly.

This commit also add support for a custom t5xxl checkpoint.
 2025-02-16 00:09:21 +02:00
Yoav HaCohen
770dd45d34 Inference: Accept PIL.Image in load_image_to_tensor_with_resize_and_crop 2025-02-15 23:53:21 +02:00
Yoav HaCohen
fa38d5bf62 Inference: Explictly set the logger name 2025-02-15 23:53:21 +02:00