Conda failed to create environment

[benczejrobert]

Machine: ASUS TUF Dash F15 FX517ZE_FX517ZE
OS: Windows 11 Education 23H2, 64 bit
Version: 10.0.22631 Build 22631
GPU: Nvidia RTX 3050

D:\AudioSep>conda env create -f environment.yml
Retrieving notices: ...working... done
Collecting package metadata (repodata.json): done
Solving environment: failed
Error:
ResolvePackageNotFound:

• urllib3==1.26.14=py310h06a4308_0
• jupyter_core==5.3.0=py310h06a4308_0
• libcusolver==11.3.4.124=h33c3c4e_0
• cuda-nvrtc==11.6.124=h020bade_0
• ld_impl_linux-64==2.38=h1181459_1
• libdeflate==1.17=h5eee18b_0
• tornado==6.2=py310h5eee18b_0
• cuda-nvml-dev==11.6.55=haa9ef22_0
• gmp==6.2.1=h295c915_3
• openh264==2.1.1=h4ff587b_0
• pytorch==1.13.1=py3.10_cuda11.6_cudnn8.3.2_0
• cuda-cuxxfilt==11.6.124=hecbf4f6_0
• libcusparse-dev==11.7.2.124=hbbe9722_0
• libcufft-dev==10.7.1.112=ha5ce4c0_0
• zstd==1.5.4=hc292b87_0
• freetype==2.12.1=h4a9f257_0
• gnutls==3.6.15=he1e5248_0
• tk==8.6.12=h1ccaba5_0
• lz4-c==1.9.4=h6a678d5_0
• cuda-nsight==12.1.55=0
• libcusparse==11.7.2.124=h7538f96_0
• cuda-nvrtc-dev==11.6.124=h249d397_0
• brotlipy==0.7.0=py310h7f8727e_1002
• psutil==5.9.0=py310h5eee18b_0
• ruamel.yaml.clib==0.2.6=py310h5eee18b_1
• _openmp_mutex==5.1=1_gnu
• cryptography==38.0.4=py310h9ce1e76_0
• libcublas-dev==11.9.2.110=h5c901ab_0
• libcufile-dev==1.6.0.25=0
• cuda-nvprune==11.6.124=he22ec0a_0
• ca-certificates==2023.01.10=h06a4308_0
• libiconv==1.16=h7f8727e_2
• libcublas==11.9.2.110=h5e84587_0
• libwebp-base==1.2.4=h5eee18b_1
• tqdm==4.64.1=py310h06a4308_0
• libsodium==1.0.18=h7b6447c_0
• libunistring==0.9.10=h27cfd23_0
• gds-tools==1.6.0.25=0
• libtiff==4.5.0=h6a678d5_2
• readline==8.2=h5eee18b_0
• cuda-gdb==12.1.55=0
• libidn2==2.3.2=h7f8727e_0
• libtasn1==4.19.0=h5eee18b_0
• lame==3.100=h7b6447c_0
• conda-content-trust==0.1.3=py310h06a4308_0
• bzip2==1.0.8=h7b6447c_0
• setuptools==65.6.3=py310h06a4308_0
• libffi==3.4.2=h6a678d5_6
• requests==2.28.1=py310h06a4308_0
• libnvjpeg-dev==11.6.2.124=hb5906b9_0
• nest-asyncio==1.5.6=py310h06a4308_0
• conda-package-handling==2.0.2=py310h06a4308_0
• debugpy==1.5.1=py310h295c915_0
• sqlite==3.40.1=h5082296_0
• mkl-service==2.4.0=py310h7f8727e_0
• numpy-base==1.23.5=py310h8e6c178_0
• conda==23.3.1=py310h06a4308_0
• libgcc-ng==11.2.0=h1234567_1
• pip==22.3.1=py310h06a4308_0
• intel-openmp==2021.4.0=h06a4308_3561
• libnpp-dev==11.6.3.124=h3c42840_0
• boltons==23.0.0=py310h06a4308_0
• pycosat==0.6.4=py310h5eee18b_0
• pyzmq==23.2.0=py310h6a678d5_0
• cuda-nvcc==11.6.124=hbba6d2d_0
• ipython==8.12.0=py310h06a4308_0
• ncurses==6.4=h6a678d5_0
• nettle==3.7.3=hbbd107a_1
• libwebp==1.2.4=h11a3e52_1
• ruamel.yaml==0.17.21=py310h5eee18b_0
• zstandard==0.18.0=py310h5eee18b_0
• cffi==1.15.1=py310h5eee18b_3
• jpeg==9e=h5eee18b_1
• xz==5.2.10=h5eee18b_1
• libuuid==1.41.5=h5eee18b_0
• certifi==2022.12.7=py310h06a4308_0
• mkl_random==1.2.2=py310h00e6091_0
• flit-core==3.8.0=py310h06a4308_0
• libcufile==1.6.0.25=0
• libgomp==11.2.0=h1234567_1
• giflib==5.2.1=h5eee18b_3
• libpng==1.6.39=h5eee18b_0
• lerc==3.0=h295c915_0
• typing_extensions==4.4.0=py310h06a4308_0
• cuda-cupti==11.6.124=h86345e5_0
• idna==3.4=py310h06a4308_0
• libstdcxx-ng==11.2.0=h1234567_1
• platformdirs==2.5.2=py310h06a4308_0
• ipykernel==6.19.2=py310h2f386ee_0
• matplotlib-inline==0.1.6=py310h06a4308_0
• pluggy==1.0.0=py310h06a4308_1
• zlib==1.2.13=h5eee18b_0
• lcms2==2.12=h3be6417_0
• numpy==1.23.5=py310hd5efca6_0
• cuda-cccl==11.6.55=hf6102b2_0
• mkl_fft==1.3.1=py310hd6ae3a3_0
• libnpp==11.6.3.124=hd2722f0_0
• libcufft==10.7.1.112=hf425ae0_0
• comm==0.1.2=py310h06a4308_0
• packaging==23.0=py310h06a4308_0
• pysocks==1.7.1=py310h06a4308_0
• cuda-driver-dev==11.6.55=0
• cuda-nvtx==11.6.124=h0630a44_0
• cuda-cudart==11.6.55=he381448_0
• libnvjpeg==11.6.2.124=hd473ad6_0
• conda-package-streaming==0.7.0=py310h06a4308_0
• mkl==2021.4.0=h06a4308_640
• openssl==1.1.1t=h7f8727e_0
• cuda-cuobjdump==11.6.124=h2eeebcb_0
• jupyter_client==8.1.0=py310h06a4308_0
• cuda-cudart-dev==11.6.55=h42ad0f4_0
• python==3.10.9=h7a1cb2a_0
• cuda-samples==11.6.101=h8efea70_0
• zeromq==4.3.4=h2531618_0
• toolz==0.12.0=py310h06a4308_0

[davyxx3]

Looks like It works only on Linux, not on Windows or MacOS.

[yukinoshitayuigahama]

I meet the same problem, any ideas on how to fix it?

[celvig]

Pls try out the environment_win64.yaml file in my fork. It's exported from my local environment which works for me
https://github.com/coldinjection/AudioSep/blob/main/environment_win64.yaml

[Senzen18]

@coldinjection tried installing using your env file got this error :
The conflict is caused by:
The user requested torch==2.1.2
dtk 0.2 depends on torch
lightning 2.0.0 depends on torch<4.0 and >=1.11.0
pytorch-ignite 0.3.0 depends on torch
pytorch-lightning 2.0.1.post0 depends on torch>=1.11.0
timm 0.3.2 depends on torch>=1.0
torchaudio 0.13.1 depends on torch==1.13.1

To fix this you could try to:

1. loosen the range of package versions you've specified
2. remove package versions to allow pip attempt to solve the dependency conflict

Pip subprocess error:
ERROR: Cannot install -r C:\Users\sendh\QBAS\qbas\condaenv.tquzlmha.requirements.txt (line 101), -r C:\Users\sendh\QBAS\qbas\condaenv.tquzlmha.requirements.txt (line 169), -r C:\Users\sendh\QBAS\qbas\condaenv.tquzlmha.requirements.txt (line 170), -r C:\Users\sendh\QBAS\qbas\condaenv.tquzlmha.requirements.txt (line 216), -r C:\Users\sendh\QBAS\qbas\condaenv.tquzlmha.requirements.txt (line 221), -r C:\Users\sendh\QBAS\qbas\condaenv.tquzlmha.requirements.txt (line 39) and torch==2.1.2 because these package versions have conflicting dependencies.
ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts

failed

CondaEnvException: Pip failed

[Brodvd]

@Senzen18 I had the same problem. @coldinjection @liuxubo717 ?

[Brodvd]

Or other that had resolve these problems?

[Brodvd]

Hey, I ran Audio Sep, the instruction are here: https://github.com/Brodvd/Audio-Sep---from-Huggin-Face---
Now It can be execute correctly locally.