damian0815 compel issues

damian0815 / compel

A prompting enhancement library for transformers-type text embedding systems

MIT License

519 stars 47 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

TypeError: unsupported operand type(s) for +: 'NoneType' and 'int' in get_embeddings_for_weighted_prompt_fragments

#105 Clement-Lelievre opened 2 days ago
0
Stable Cascade support, new ReturnedEmbeddingsType

#104 Teriks opened 2 weeks ago
0
Multiple unclosed brackets slow down/crash Prompt Parser

#103 sssemil opened 3 weeks ago
0
Paralism in generating embeddings

#102 Daniel-SicSo-Edinburgh opened 1 month ago
1
Update README.md

#101 GerbertBless closed 3 weeks ago
1
Question: Do I still need to set negative_prompt to pipe() or I should just combine them to prompts with "-"?

#100 tclzcja opened 1 month ago
0
Support Flux.1

#99 MohamedAliRashad opened 1 month ago
6
Support for AuraFlow

#98 waadarsh opened 2 months ago
0
ELLA & Compel

#97 andupotorac opened 2 months ago
0
Incorrect pooled embeds

#96 BitPhinix opened 2 months ago
1
Stable Cascade?

#95 Teriks opened 3 months ago
5
[Reporting a Bug] Error when prompt contains "!"

#94 sk-uma opened 3 months ago
3
Compel to text

#93 HatmanStack opened 3 months ago
3
Support for SD3

#92 adhikjoshi opened 3 months ago
11
License in pyproject.toml shows AGPL

#91 thnake closed 3 months ago
0
Error when using enable_sequential_cpu_offload() reopen

#90 xalteropsx closed 3 months ago
0
[SDXL] Compel can't batch prompt (mat1 and mat2 shapes cannot be multiplied)

#89 Fqlox opened 3 months ago
1
Fix and() for SDXL #78

#88 kno10 opened 3 months ago
0
When using compel it changes results drastically with same prompts

#87 alexblattner closed 4 months ago
1
when use compel with inpaint controlnet, the result mask area is blue

#86 crapthings opened 5 months ago
2
"build_conditioning_tensor_for_conjunction shape has length 3" - Using .and() on ControlNet + SDXL causing mismatch size error. 154) must match the size of tensor b (77) at non-singleton dimension 1

#85 rndm-jpg closed 5 months ago
0
Using fp16 text_encoder seems to cause computational error

#84 GongXinyuu closed 2 months ago
5
SD XL Inpainting - 'NoneType' object has no attribute 'repeat'

#83 WaterKnight1998 opened 7 months ago
3
Compel fails with TypeError: 'NoneType' object cannot be interpreted as integer when using T5Tokenizer

#82 davemssavage opened 7 months ago
2
Add clip_skip to EmbeddingsProvider

#81 Disty0 closed 4 months ago
3
fix small error in call to get_token_ids

#80 spezialspezial closed 5 months ago
0
still get The size of tensor a (154) must match the size of tensor b (77) after using the code in readme

#79 Deaddawn opened 8 months ago
6
sdxl with .and() with weights?

#78 rb-synth opened 9 months ago
5
Performance deteriorates at some prompts

#77 sk-uma closed 9 months ago
1
when I try the code in compel-demo-sdxl.ipynb, a OverflowError returns

#76 phoenixGRS opened 9 months ago
0
ValueError: `prompt_embeds` and `negative_prompt_embeds` must have the same shape when passed directly, but got: `prompt_embeds` torch.Size([1, 154, 2048]) != `negative_prompt_embeds` torch.Size([1, 77, 2048]).

#75 ynie closed 9 months ago
1
Issue with negative prompt when using non truncated long prompts

#74 o5faruk opened 10 months ago
2
Is there anyway to debug the parsed prompt weight?

#73 ynie closed 10 months ago
1
Token indices sequence length is longer than the specified maximum sequence length for this model

#72 mlinmg closed 10 months ago
4
compel vs auto1111 prompt weight

#71 loboere opened 11 months ago
1
fix: on error parsing, fall back to "raw" prompt

#70 psychedelicious opened 11 months ago
2
Invalid blend syntax causes error, does not fallback on treating prompt as raw prompt

#69 psychedelicious opened 11 months ago
0
How to use multiple textual inversions?

#68 sarmientoj24 opened 1 year ago
1
Blend for Interpolation

#67 aduchon opened 1 year ago
2
Why does the following error occur? TypeError: delimited_list() got an unexpected keyword argument 'min'

#66 Eduardishion closed 1 year ago
2
Inconsistent parse errors when using .blend() incorrectly

#65 JeLuF opened 1 year ago
1
allow padding mixed 2d and 3d conditionings

#64 jameshurst closed 1 year ago
0
Supporting separate SDXL CLIP G/L weighted prompts

#63 subpanic closed 1 year ago
2
Is it ok to use compel only for prompt whle using the original negative prompt in SDXL1.0 text to image?

#62 leyangjin closed 1 year ago
1
load negative_embeddings

#61 cacko closed 1 year ago
1
Compel ails with SDXL Refiner

#60 hosseinsarshar opened 1 year ago
1
improve long prompt handling logic to split >75 token prompts at previous punctuation

#59 damian0815 opened 1 year ago
13
Error when running compel-demo-sdxl.py example

#58 yegorsw closed 1 year ago
1
Add device parameter to EmbeddingsProvider

#57 Disty0 closed 1 year ago
5
Clip Skip SDXL

#56 adhikjoshi closed 1 year ago
1