search

ToTheBeginning / PuLID

[NeurIPS 2024] Official code for PuLID: Pure and Lightning ID Customization via Contrastive Alignment
Apache License 2.0
2.12k stars 138 forks source link

Can you provide support for flux1.dev-fp8? #63

Open JohnsonXi opened 2 weeks ago

JohnsonXi commented 2 weeks ago
image
ToTheBeginning commented 2 weeks ago

Provided now.

NeoAnthropocene commented 2 weeks ago

Does this fp8 implementation drop the VRAM usage under 16 Gigs? The consumer grade GPUs have a maximum of 24 gigs, but not everyone got it for sure.

Do you have any plans to support GGUF models?

JohnsonXi commented 2 weeks ago

Provided now.

This is a much better job compared to sdxl, thanks for your efforts

ToTheBeginning commented 2 weeks ago

@NeoAnthropocene We have further optimized the codes, now it supports 16GB cards!

DrakenZA commented 2 weeks ago

File "/usr/local/lib/python3.10/dist-packages/pydantic/_internal/_generate_schema.py", line 558, in _unknown_type_schema raise PydanticSchemaGenerationError( pydantic.errors.PydanticSchemaGenerationError: Unable to generate pydantic-core schema for <class 'starlette.requests.Request'>. Set arbitrary_types_allowed=True in the model_config to ignore this error or implement __get_pydantic_core_schema__ on your type to fully support it.

If you got this error by calling handler() within __get_pydantic_core_schema__ then you likely need to call handler.generate_schema(<some type>) since we do not call __get_pydantic_core_schema__ on <some type> otherwise to avoid infinite recursion.

For further information visit https://errors.pydantic.dev/2.9/u/schema-for-unknown-type

I get this error when following the instructions on google colab, it gets all the way to loading up the gradio UI, but when pressing generate i get this.

ToTheBeginning commented 2 weeks ago

File "/usr/local/lib/python3.10/dist-packages/pydantic/_internal/_generate_schema.py", line 558, in _unknown_type_schema raise PydanticSchemaGenerationError( pydantic.errors.PydanticSchemaGenerationError: Unable to generate pydantic-core schema for <class 'starlette.requests.Request'>. Set arbitrary_types_allowed=True in the model_config to ignore this error or implement __get_pydantic_core_schema__ on your type to fully support it.

If you got this error by calling handler() within __get_pydantic_core_schema__ then you likely need to call handler.generate_schema(<some type>) since we do not call __get_pydantic_core_schema__ on <some type> otherwise to avoid infinite recursion.

For further information visit https://errors.pydantic.dev/2.9/u/schema-for-unknown-type

I get this error when following the instructions on google colab, it gets all the way to loading up the gradio UI, but when pressing generate i get this.

@DrakenZA Maybe you can refer to this issue https://github.com/ToTheBeginning/PuLID/issues/61

NeoAnthropocene commented 2 weeks ago

@NeoAnthropocene We have further optimized the codes, now it supports 16GB cards!

Thanks for letting me know!

I guess I will wait to test it on ConfyUI once it becomes really optimized. Waiting for the feedback from cubic and ZhoZhoZho 😀