ce9ac2fe05
Fix clip_g/clip_l mixup ( #4168 )
7 months ago
5f98de7697
Load flux t5 in fp8 if weights are in fp8.
7 months ago
1589b58d3e
Basic Flux Schnell and Flux Dev model implementation.
7 months ago
c24f897352
Fix to get fp8 working on T5 base.
7 months ago
a5991a7aa6
Fix hunyuan dit text encoder weights always being in fp32.
7 months ago
2c038ccef0
Lower CLIP memory usage by a bit.
7 months ago
b85216a3c0
Lower T5 memory usage by a few hundred MB.
7 months ago
82cae45d44
Fix potential issue with non clip text embeddings.
7 months ago
4ba7fa0244
Refactor: Move sd2_clip.py to text_encoders folder.
7 months ago
cf4418b806
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.
7 months ago
a9ac56fc0d
Own BertModel implementation that works with lowvram.
7 months ago
a5f4292f9f
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.
7 months ago
f87810cd3e
Let tokenizers return weights to be stored in the saved checkpoint.
7 months ago
10c919f4c7
Make it possible to load tokenizer data from checkpoints.
7 months ago
0a4c49c57c
Support MT5.
7 months ago
88ed893034
Allow SPieceTokenizer to load model from a byte string.
7 months ago
14764aa2e2
Rename LLAMATokenizer to SPieceTokenizer.
7 months ago
1305fb294c
Refactor: Move some code to the comfy/text_encoders folder.
7 months ago
29c2e26724
Better tokenizing code for AuraFlow.
8 months ago
9f291d75b3
AuraFlow model implementation.
8 months ago
Alexander Brown
comfyanonymous
comfyanonymous