12月1月2月3月4月5月6月7月8月9月10月11月
周一周三周五
一年内 265876 次贡献
更少更多
- frostbyte_neo 推送了仓库 github-featured/invoke-ai--invokeai 的 ryan/extension-manager-cleanup-suggestions 分支51750e2060 Suggestion for how to clean up the ExtensionManager.
- frostbyte_neo 推送了仓库 github-featured/invoke-ai--invokeai 的 ryan/dora-peft 分支比较 3 提交 »dcf11a01ce Add TODO comment about peformance bottleneck in LoRA loading code.6e4de001f1 Remove line that was intended to save memory, but wasn't actually having any effect.4af258615f Improve the robustness of the logic for determining the PEFT model type. In particular, so that it doesn't incorrectly detect DoRA models as LoRA models.
- frostbyte_neo 推送了仓库 github-featured/invoke-ai--invokeai 的 ryan/dora 分支比较 10 提交 »6bfb4927c7 WIP - not workingc15e9e23ca Tidy types in LoraModelPatcher.e1aa1ed6af Split LoraModelPatcher out from ModelPatcher monolith.4b68050c9b Minor cleanup of LoRAModelRaw.9e68a5c851 Use load_state_dict() util in LoRAModelRaw.
- frostbyte_neo 推送了仓库 github-featured/invoke-ai--invokeai 的 ryan/dense-diffusion 分支比较 4 提交 »ff950bc5cd Add support for mask weights, and only mask the tokens associated with the prompts (not eh entire 77-token embedding).969982b789 Fixup some details of densediffusion for testing.b8cbff828b wipd3a40c5b2b Rough hacky implementation of DenseDiffusion.
- frostbyte_neo 推送了仓库 github-featured/invoke-ai--invokeai 的 ryan/dont-use-safetensors-mmap 分支03157a35ec Don't use safetensors mmap for some FLUX models to see what happens.
- frostbyte_neo 推送了仓库 github-featured/invoke-ai--invokeai 的 ryan/conservative-ram 分支109cbb8532 Update the default Model Cache behavior to be more conservative with RAM usage.
- frostbyte_neo 推送了仓库 github-featured/invoke-ai--invokeai 的 ryan/clothing-workflow-sam 分支比较 4 提交 »d5e824e782 Add mask dilation.5dd0aeb1ab Add VTO lighting controls.6fc3c959e0 Add VTO invocation.09046e811f Merge branch 'blessedcoolant-grounding_segment_anything' into ryan/clothing-workflow-sam