instantX-research / InstantID

InstantID: Zero-shot Identity-Preserving Generation in Seconds 🔥
https://instantid.github.io/
Apache License 2.0
11.05k stars 805 forks source link

error1: libcublasLt.so.12 ;error 2: libcublasLt.so.12 #262

Closed zeezou-isee closed 2 months ago

zeezou-isee commented 2 months ago

This is a wonderful project!

When I create an envs of python 3.10.0 in my linux-server(Driver Version: 550.107.02, CUDA Version: 12.4 ). I download pytorch conda install pytorch==2.0.0 torchvision==0.15.0 torchaudio==2.0.0 pytorch-cuda=11.7 -c pytorch -c nvidia

I organized my filefolder as author said.

When I run infer_full.py , problems occurred as follows: 2024-08-22 10:38:52.670883465 [E:onnxruntime:Default, provider_bridge_ort.cc:1992 TryGetProviderInfo_CUDA] /onnxruntime_src/onnxruntime/core/session/provider_bridge_ort.cc:1637 onnxruntime::Provider& onnxruntime::ProviderLibrary::Get() [ONNXRuntimeError] : 1 : FAIL : Failed to load library libonnxruntime_providers_cuda.so with error: libcublasLt.so.12: cannot open shared object file: No such file or directory

and

OSError: diffusers/controlnet-depth-sdxl-1.0-small does not appear to have a file named config.json.

what can I do to solve these problems?

zeezou-isee commented 2 months ago

I have solved them.

error 1: Due to my conda atomatically installing onnxruntime V1.19.0, it is not compatible with my cuda-11.7 envs. So i reinstall by instruct pip install onnxruntime-gpu==1.17.0, solved!

error 2: I download the whole folder in huggingface.co diffusers/controlnet-depth-sdxl-1.0-small,and organized them in a folder in the same name, putting this folder under project root, solved!