Commits · 72188dffc3d331be41e366c4f0fa6883645f669a · chenpangpang / ComfyUI

06 Oct, 2023 1 commit
- load_checkpoint_guess_config can now optionally output the model. · 72188dff
  comfyanonymous authored Oct 06, 2023
  
  72188dff
04 Oct, 2023 1 commit
- Fix quality loss due to low precision · 9bfec2bd
  City authored Oct 04, 2023
  
  9bfec2bd
29 Sep, 2023 2 commits
- fix: typo in extra sampler · 0f17993d
  badayvedat authored Sep 29, 2023
  
  0f17993d
- Add SamplerDPMPP_2M_SDE node. · 66756de1
  comfyanonymous authored Sep 28, 2023
  
  66756de1
28 Sep, 2023 3 commits
- Print missing VAE keys. · 71713888
  comfyanonymous authored Sep 28, 2023
  
  71713888
- Add missing samplers to KSamplerSelect. · d234ca55
  comfyanonymous authored Sep 28, 2023
  
  d234ca55
- Add a SamplerCustom Node. · 1adcc4c3
  comfyanonymous authored Sep 27, 2023
```
This node takes a list of sigmas and a sampler object as input.

This lets people easily implement custom schedulers and samplers as nodes.

More nodes will be added to it in the future.
```
  1adcc4c3
27 Sep, 2023 2 commits
- Refactor sampling related code. · bf3fc2f1
  comfyanonymous authored Sep 27, 2023
  
  bf3fc2f1
- Model patches can now know which batch is positive and negative. · fff491b0
  comfyanonymous authored Sep 27, 2023
  
  fff491b0
26 Sep, 2023 2 commits
- Scheduler code refactor. · 1d6dd831
  comfyanonymous authored Sep 26, 2023
  
  1d6dd831
- Sampling code refactor. · 446caf71
  comfyanonymous authored Sep 26, 2023
  
  446caf71
23 Sep, 2023 3 commits
- Support more controlnet models. · 76cdc809
  comfyanonymous authored Sep 23, 2023
  
  76cdc809
- Allow Intel GPUs to LoRA cast on GPU since it supports BF16 natively. · eec449ca
  Simon Lui authored Sep 22, 2023
  
  eec449ca
- Add a way to set output block patches to modify the h and hsp. · afa2399f
  comfyanonymous authored Sep 22, 2023
  
  afa2399f
21 Sep, 2023 1 commit
- Allow having a different pooled output for each image in a batch. · 492db2de
  comfyanonymous authored Sep 21, 2023
  
  492db2de
20 Sep, 2023 1 commit
- Only do the cast on the device if the device supports it. · 1cdfb3db
  comfyanonymous authored Sep 20, 2023
  
  1cdfb3db
19 Sep, 2023 3 commits
- Don't depend on torchvision. · 7c9a92f5
  comfyanonymous authored Sep 19, 2023
  
  7c9a92f5
- Added support for lanczos scaling · 2b6b1781
  MoonRide303 authored Sep 19, 2023
  
  2b6b1781
- Do lora cast on GPU instead of CPU for higher performance. · b92bf819
  comfyanonymous authored Sep 18, 2023
  
  b92bf819
17 Sep, 2023 1 commit
- Enable pytorch attention by default on xpu. · 321c5fa2
  comfyanonymous authored Sep 17, 2023
  
  321c5fa2
16 Sep, 2023 2 commits
- Support models without previews. · 61b1f677
  comfyanonymous authored Sep 16, 2023
  
  61b1f677
- Add cond_or_uncond array to transformer_options so hooks can check what is · 43d4935a
  comfyanonymous authored Sep 15, 2023
```
cond and what is uncond.
```
  43d4935a
15 Sep, 2023 4 commits
- Add DDPM sampler. · 415abb27
  comfyanonymous authored Sep 15, 2023
  
  415abb27
- This isn't used anywhere. · 94e4fe39
  comfyanonymous authored Sep 15, 2023
  
  94e4fe39
- Support for text encoder models that need attention_mask. · 44361f63
  comfyanonymous authored Sep 15, 2023
  
  44361f63
- Set last layer on SD2.x models uses the proper indexes now. · 0d8f3764
  comfyanonymous authored Sep 14, 2023
```
Before I had made the last layer the penultimate layer because some
checkpoints don't have them but it's not consistent with the others models.

TLDR: for SD2.x models only: CLIPSetLastLayer -1 is now -2.
```
  0d8f3764
14 Sep, 2023 1 commit
- Don't run text encoders on xpu because there are issues. · 0966d3ce
  comfyanonymous authored Sep 14, 2023
  
  0966d3ce
13 Sep, 2023 1 commit
- Only parse command line args when main.py is called. · 3039b08e
  comfyanonymous authored Sep 13, 2023
  
  3039b08e
12 Sep, 2023 2 commits
- Don't leave very large hidden states in the clip vision output. · ed587306
  comfyanonymous authored Sep 12, 2023
  
  ed587306
- Fix issue where autocast fp32 CLIP gave different results from regular. · fb3b7282
  comfyanonymous authored Sep 11, 2023
  
  fb3b7282
11 Sep, 2023 1 commit
- Add ldm format support to UNETLoader. · 7d401ed1
  comfyanonymous authored Sep 11, 2023
  
  7d401ed1
08 Sep, 2023 3 commits
- Add a penultimate_hidden_states to the clip vision output. · e85be36b
  comfyanonymous authored Sep 08, 2023
  
  e85be36b
- Support diffusers format t2i adapters. · 1e6b6710
  comfyanonymous authored Sep 08, 2023
  
  1e6b6710
- Allow cancelling of everything with a progress bar. · 326577d0
  comfyanonymous authored Sep 07, 2023
  
  326577d0
06 Sep, 2023 1 commit
- Add a ConditioningSetAreaPercentage node. · f88f7f41
  comfyanonymous authored Sep 06, 2023
  
  f88f7f41
04 Sep, 2023 1 commit
- Add a force argument to soft_empty_cache to force a cache empty. · 1938f5c5
  comfyanonymous authored Sep 04, 2023
  
  1938f5c5
03 Sep, 2023 3 commits
- Revert changes in comfy/ldm/modules/diffusionmodules/util.py, which is unused. · 2da73b70
  Simon Lui authored Sep 02, 2023
  
  2da73b70
- Move some functions to utils.py · a74c5dbf
  comfyanonymous authored Sep 02, 2023
  
  a74c5dbf
- Some fixes to generalize CUDA specific functionality to Intel or other GPUs. · 4a0c4ce4
  Simon Lui authored Sep 02, 2023
  
  4a0c4ce4
02 Sep, 2023 1 commit
- Use common function to reshape batch to. · 77a176f9
  comfyanonymous authored Sep 02, 2023
  
  77a176f9