Skip to content
代码片段 群组 项目

仓库图

可以使用方向键移动图形。
选择Git版本
  • add_codeowner
  • add_custom_node_i18n
  • add_kf_interp
  • add_owners
  • add_sample_sigmas
  • add_undefined_lint
  • all_flake
  • annoate_get_input_info
  • base-path-env-var
  • calc_weight_api
  • clean-folder-paths
  • clean_3.8
  • comfyanonymous-patch-1
  • defaultconds_fix
  • desktop-release-feb172025
  • desktop-release-feb192025
  • desktop-release-mar112025
  • document_get_attr
  • explicit_control_option
  • fast_options
  • v0.3.27
  • v0.3.26
  • v0.3.25
  • v0.3.24
  • v0.3.23
  • v0.3.22
  • v0.3.21
  • v0.3.20
  • v0.3.19
  • v0.3.18
  • v0.3.17
  • v0.3.16
  • v0.3.15
  • v0.3.14
  • v0.3.13
  • v0.3.12
  • v0.3.11
  • v0.3.10
  • v0.3.9
  • v0.3.8
40 个结果
Created with Raphaël 2.2.015Jul141312111098765432130Jun292827262524232221201918171615141312111098765432131May30292827262524232221201918171615141312111098765432130Apr292827262524232221201918171615141312111098765432131Mar30292827262524232221201918171615141312111098765432129Feb2827262524232221201918171615141312111098765432131Jan30292827262524232221201918171615141312111098765432131Dec3029282726252423222120191817161516151413121110111098765432130Nov292827262524Fix bug with custom nodes on other drives.Fix bug with workflow not registering change.Refactor: Move some code to the comfy/text_encoders folder.Quick fix for the promax controlnet.v0.0.2v0.0.2New menu fixes - fix send to workflow (#3909)Support AuraFlow Lora and loading model weights in diffusers format.Add link to AuraFlow example in Readme.Add Github Workflow for releasing stable versions and standalone bundle. (#3949)v0.0.1v0.0.1Better tokenizing code for AuraFlow.Add sentencepiece dependency.Update artifact name to be ComfyUI_windows_portable_nvidia.7zrh-releaserh-releaseAdd a ModelSamplingAuraFlow node to change the shift value.AuraFlow model implementation.Fix error message never being shown.Cleaner support for loading different diffusion model types.Remove useless code.More flexibility with text encoder return values.Support returning text encoder attention masks.Update bug issue template (#3996)Check unhandled exception in test log in test action (#3987)Add __module__ to node info (#3936)Add sampler_pre_cfg_function (#3979)Fix ConditioningZeroOut when there is no pooled output.Fix loadGraphData call during restore (#3976)Remove useless import.Controlnet union model basic implementation.lzma2Upgrade checkout and setup-python to latest version.Only build CUDA 12.1 + 3.11 Python.Rename clip_t5base to t5base for stable audio text encoder.Move audio nodes out of _for_testing.Add a way to set the timestep multiplier in the flow sampling.Support other types of T5 models.Better T5xxl detection.Cleanup T5 code a bit.Allow specifying the padding token for the tokenizer.Allow zeroing out of embeds with unused attention mask.Remove duplicate import.Readme changes.Forgot this in last commit.