comfyanonymous
28b12fb344
Long CLIP L support for SDXL, SD3 and Flux.
...
Use the *CLIPLoader nodes.
2024-09-15 07:59:38 -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
comfyanonymous
7c94219393
Fix bundled embed.
2024-08-07 13:30:45 -04:00
comfyanonymous
04f70b853e
Support format for embeddings bundled in loras.
2024-08-07 03:45:25 -04:00
comfyanonymous
4e9abfb0d5
Lower CLIP memory usage by a bit.
2024-07-31 01:32:35 -04:00
comfyanonymous
671fba8c0d
Fix potential issue with non clip text embeddings.
2024-07-30 14:41:13 -04:00
comfyanonymous
15a95668ae
Let tokenizers return weights to be stored in the saved checkpoint.
2024-07-25 10:52:09 -04:00
comfyanonymous
33e584b3bc
Make it possible to load tokenizer data from checkpoints.
2024-07-24 16:43:53 -04:00
comfyanonymous
238efd021c
More flexibility with text encoder return values.
...
Text encoders can now return other values to the CONDITIONING than the cond
and pooled output.
2024-07-10 20:06:50 -04:00
comfyanonymous
27226fb151
Support returning text encoder attention masks.
2024-07-10 19:31:22 -04:00
comfyanonymous
002954b124
Rename clip_t5base to t5base for stable audio text encoder.
2024-07-08 08:53:55 -04:00
comfyanonymous
b9b3986c45
Allow specifying the padding token for the tokenizer.
2024-07-06 00:06:49 -04:00
comfyanonymous
fc982c3e43
Allow zeroing out of embeds with unused attention mask.
2024-07-05 23:48:17 -04:00
Mario Klingemann
c39be91219
Update sd1_clip.py ( #3684 )
...
Made token instance check more flexible so it also works with integers from numpy arrays or long tensors
2024-06-19 16:42:41 -04:00
comfyanonymous
917810a2c6
Load the SD3 T5xxl model in the same dtype stored in the checkpoint.
2024-06-11 17:03:26 -04:00
comfyanonymous
a2a20c27f4
Support zeroing out text embeddings with the attention mask.
2024-06-09 16:51:58 -04:00
comfyanonymous
add86d4eb5
Use the end token for the text encoder attention mask.
2024-06-07 03:05:23 -04:00
comfyanonymous
35d0fd9a5d
Replace prints with logging and add --verbose argument.
2024-03-10 12:14:23 -04:00
comfyanonymous
13fb082928
Add a min_length property to tokenizer class.
2024-02-26 21:36:37 -05:00
comfyanonymous
c8556665f0
Make return_projected_pooled setable from the __init__
2024-02-25 14:49:13 -05:00
comfyanonymous
766b45c2c1
Always return unprojected pooled output for gligen.
2024-02-25 07:33:13 -05:00
comfyanonymous
9e5c5a0ad3
Move text projection into the CLIP model code.
...
Fix issue with not loading the SSD1B clip correctly.
2024-02-25 01:41:08 -05:00
comfyanonymous
b102d65c41
StableCascade CLIP model support.
2024-02-16 13:29:04 -05:00
comfyanonymous
2f33a09886
Cleanup some unused imports.
2024-01-21 21:51:22 -05:00
comfyanonymous
58b39151a4
Switch text encoder to manual cast.
...
Use fp16 text encoder weights for CPU inference to lower memory usage.
2023-12-10 23:00:54 -05:00
comfyanonymous
63349484b8
Make --gpu-only put intermediate values in GPU memory instead of cpu.
2023-12-08 02:35:45 -05:00
comfyanonymous
782cf4f295
Cleaner CLIP text encoder implementation.
...
Use a simple CLIP model implementation instead of the one from
transformers.
This will allow some interesting things that would too hackish to implement
using the transformers implementation.
2023-12-06 23:50:03 -05:00
comfyanonymous
9002e58ce5
Less useless downcasting.
2023-12-04 12:53:46 -05:00
comfyanonymous
35b7304ac5
Fix last pr.
2023-11-14 14:41:31 -05:00
Jianqi Pan
58a23fc106
fix: adaptation to older versions of pytroch
2023-11-14 14:32:05 +09:00
comfyanonymous
842ac7fb1b
CLIP code refactor and improvements.
...
More generic clip model class that can be used on more types of text
encoders.
Don't apply weighting algorithm when weight is 1.0
Don't compute an empty token output when it's not needed.
2023-11-06 14:17:41 -05:00
comfyanonymous
01e37204ed
Make SDTokenizer class work with more types of tokenizers.
2023-11-06 01:09:18 -05:00
comfyanonymous
9533904e39
Fix checkpoint loader with config.
2023-10-27 22:13:55 -04:00
comfyanonymous
3ad424ff47
SD1 and SD2 clip and tokenizer code is now more similar to the SDXL one.
2023-10-27 15:54:04 -04:00
comfyanonymous
817a182bac
Restrict loading embeddings from embedding folders.
2023-10-27 02:54:13 -04:00
comfyanonymous
11df5713a0
Support for text encoder models that need attention_mask.
2023-09-15 02:02:05 -04:00
comfyanonymous
36cc11edbd
Fix issue where autocast fp32 CLIP gave different results from regular.
2023-09-11 21:49:56 -04:00
comfyanonymous
145e279e6c
Move text_projection to base clip model.
2023-08-24 23:43:48 -04:00
comfyanonymous
5dbbb2c93c
Fix potential issue with text projection matrix multiplication.
2023-08-24 00:54:16 -04:00
comfyanonymous
1aff0360c3
Initialize text encoder to target dtype.
2023-08-23 21:01:15 -04:00
comfyanonymous
e7fc7fb557
Save memory by storing text encoder weights in fp16 in most situations.
...
Do inference in fp32 to make sure quality stays the exact same.
2023-08-23 01:08:51 -04:00
comfyanonymous
bbd5052ed0
Make sure the pooled output stays at the EOS token with added embeddings.
2023-08-03 20:27:50 -04:00
comfyanonymous
f67f1c99b8
Fix CLIPSetLastLayer not reverting when removed.
2023-07-15 01:41:21 -04:00
comfyanonymous
28bf6d49da
Fix potential tensors being on different devices issues.
2023-07-12 19:29:27 -04:00
comfyanonymous
6e99974161
Support SDXL embedding format with 2 CLIP.
2023-07-10 10:34:59 -04:00
comfyanonymous
d3b3c94616
Fix bug with weights when prompt is long.
2023-07-06 02:43:40 -04:00
comfyanonymous
5ace1146c5
Lower latency by batching some text encoder inputs.
2023-07-01 15:07:39 -04:00
comfyanonymous
d5a7abe10d
Try to keep text encoders loaded and patched to increase speed.
...
load_model_gpu() is now used with the text encoder models instead of just
the unet.
2023-07-01 13:28:07 -04:00
comfyanonymous
e946dca0e1
Make highvram and normalvram shift the text encoders to vram and back.
...
This is faster on big text encoder models than running it on the CPU.
2023-07-01 12:37:23 -04:00
