comfyanonymous
da3d136cb7
Use empty t5 of size 128 for hidream, seems to give closer results.
2025-04-19 19:49:40 -04:00
power88
6f35198397
Hidream: Allow loading hidream text encoders in CLIPLoader and DualCLIPLoader ( #7676 )
...
* Hidream: Allow partial loading text encoders
* reformat code for ruff check.
2025-04-19 19:47:30 -04:00
comfyanonymous
da12bcd594
Fix hard crash when the spiece tokenizer path is bad.
2025-04-19 15:55:43 -04:00
comfyanonymous
f26cf7d5fa
Limit T5 to 128 tokens for HiDream: #7620
2025-04-16 18:07:55 -04:00
comfyanonymous
720a954e00
Basic support for hidream i1 model.
2025-04-15 17:35:05 -04:00
comfyanonymous
921181d22b
Cleanup.
2025-04-15 12:13:28 -04:00
comfyanonymous
df67f25c28
More flexible long clip support.
...
Add clip g long clip support.
Text encoder refactor.
Support llama models with different vocab sizes.
2025-04-15 10:32:21 -04:00
comfyanonymous
46b766bc49
Add an image_interleave option to the Hunyuan image to video encode node.
...
See the tooltip for what it does.
2025-03-07 19:56:26 -05:00
comfyanonymous
8e2c6489fe
Support HunyuanVideo image to video model.
2025-03-06 03:07:15 -05:00
comfyanonymous
4857e9b863
Make applying embeddings more efficient.
...
Adding new tokens no longer makes a whole copy of the embeddings weight
which can be massive on certain models.
2025-03-05 17:34:38 -05:00
comfyanonymous
843b5c22b3
Make it easier to set a custom template for hunyuan video.
2025-03-04 09:26:05 -05:00
comfyanonymous
9397b02169
Fix wan issues when prompt length is long.
2025-02-26 20:34:02 -05:00
comfyanonymous
3c67f6155a
WIP support for Wan t2v model.
2025-02-25 17:20:35 -05:00
comfyanonymous
49a76acccf
Cleanup some lumina te code.
2025-02-25 04:10:26 -05:00
comfyanonymous
e03a640306
Support Lumina 2 model.
2025-02-04 04:16:30 -05:00
comfyanonymous
fc1a24c402
Use maximum negative value instead of -inf for masks in text encoders.
...
This is probably more correct.
2025-02-02 09:46:00 -05:00
comfyanonymous
6d9f172aeb
WIP support for Nvidia Cosmos 7B and 14B text to world (video) models.
2025-01-10 09:14:16 -05:00
comfyanonymous
6b210b9e9b
Properly calculate inner dim for t5 model.
...
This is required to support some different types of t5 models.
2025-01-07 17:33:03 -05:00
City
a3a893abeb
Add PixArt model support ( #6055 )
...
* PixArt initial version
* PixArt Diffusers convert logic
* pos_emb and interpolation logic
* Reduce duplicate code
* Formatting
* Use optimized attention
* Edit empty token logic
* Basic PixArt LoRA support
* Fix aspect ratio logic
* PixArtAlpha text encode with conds
* Use same detection key logic for PixArt diffusers
2024-12-20 15:25:00 -05:00
comfyanonymous
fd5d3147dc
Pick attention implementation based on device in llama code.
2024-12-18 01:30:20 -05:00
comfyanonymous
5ea21ff2b2
Properly tokenize the template for hunyuan video.
2024-12-17 16:22:02 -05:00
comfyanonymous
aa5005ff35
Support llama hunyuan video text encoder in scaled fp8 format.
2024-12-17 04:19:22 -05:00
comfyanonymous
2b49852ca2
Basic Hunyuan Video model support.
2024-12-16 19:35:40 -05:00
Chenlei Hu
efca88cfb3
Lint all unused variables ( #5989 )
...
* Enable F841
* Autofix
* Remove all unused variable assignment
2024-12-12 17:59:16 -05:00
Chenlei Hu
9bffd67fc9
Lint unused import ( #5973 )
...
* Lint unused import
* nit
* Remove unused imports
* revert fix_torch import
* nit
2024-12-09 15:24:39 -05:00
comfyanonymous
5396a5266b
Make CLIP set last layer node work with t5 models.
2024-12-09 03:57:14 -05:00
comfyanonymous
8f3fabbfe5
Support Lightricks LTX-Video model.
2024-11-22 08:46:39 -05:00
comfyanonymous
cf71379da3
Fix mochi all in one checkpoint t5xxl key names.
2024-11-03 01:40:42 -05:00
comfyanonymous
9f8e7912ef
Basic Genmo Mochi video model support.
...
To use:
"Load CLIP" node with t5xxl + type mochi
"Load Diffusion Model" node with the mochi dit file.
"Load VAE" with the mochi vae file.
EmptyMochiLatentVideo node for the latent.
euler + linear_quadratic in the KSampler node.
2024-10-26 06:54:00 -04:00
comfyanonymous
567711396b
Support scaled fp8 t5xxl model.
2024-10-20 22:27:00 -04:00
comfyanonymous
3e7430aaa6
Make clip loader nodes support loading sd3 t5xxl in lower precision.
...
Add attention mask support in the SD3 text encoder code.
2024-10-10 15:06:15 -04:00
comfyanonymous
a05692d331
Fix flux TE not loading t5 embeddings.
2024-09-24 22:57:22 -04:00
comfyanonymous
28b12fb344
Long CLIP L support for SDXL, SD3 and Flux.
...
Use the *CLIPLoader nodes.
2024-09-15 07:59:38 -04:00
comfyanonymous
d30678a05e
Support loading long clipl model with the CLIP loader node.
2024-08-20 10:46:36 -04:00
comfyanonymous
2062f06e50
Properly set if clip text pooled projection instead of using hack.
2024-08-20 10:46:36 -04:00
comfyanonymous
03c874ec6e
Add model_options for text encoder.
2024-08-17 11:17:20 -04:00
Silver
b87a10e5f1
Correct spelling 'token_weight_pars_t5' to 'token_weight_pairs_t5' ( #4200 )
2024-08-04 17:10:02 -04:00
Alexander Brown
90267ffb3a
Fix clip_g/clip_l mixup ( #4168 )
2024-08-01 21:40:56 -04:00
comfyanonymous
d6d48306d8
Load flux t5 in fp8 if weights are in fp8.
2024-08-01 11:05:56 -04:00
comfyanonymous
fc6a27f36c
Basic Flux Schnell and Flux Dev model implementation.
2024-08-01 09:49:29 -04:00
comfyanonymous
a4b37c2939
Fix to get fp8 working on T5 base.
2024-07-31 02:00:19 -04:00
comfyanonymous
c5b754fbb6
Fix hunyuan dit text encoder weights always being in fp32.
2024-07-31 01:34:57 -04:00
comfyanonymous
4e9abfb0d5
Lower CLIP memory usage by a bit.
2024-07-31 01:32:35 -04:00
comfyanonymous
7f0000bc5a
Lower T5 memory usage by a few hundred MB.
2024-07-31 00:52:34 -04:00
comfyanonymous
671fba8c0d
Fix potential issue with non clip text embeddings.
2024-07-30 14:41:13 -04:00
comfyanonymous
79ea2d3d12
Refactor: Move sd2_clip.py to text_encoders folder.
2024-07-28 01:19:20 -04:00
comfyanonymous
f4a8f8db50
Don't treat Bert model like CLIP.
...
Bert can accept up to 512 tokens so any prompt with more than 77 should
just be passed to it as is instead of splitting it up like CLIP.
2024-07-26 13:08:12 -04:00
comfyanonymous
0a7029fabf
Own BertModel implementation that works with lowvram.
2024-07-26 04:47:17 -04:00
comfyanonymous
06fb93c0d1
Basic hunyuan dit implementation. ( #4102 )
...
* Let tokenizers return weights to be stored in the saved checkpoint.
* Basic hunyuan dit implementation.
* Fix some resolutions not working.
* Support hydit checkpoint save.
* Init with right dtype.
* Switch to optimized attention in pooler.
* Fix black images on hunyuan dit.
2024-07-25 18:21:08 -04:00
comfyanonymous
15a95668ae
Let tokenizers return weights to be stored in the saved checkpoint.
2024-07-25 10:52:09 -04:00
