popov/neuron-tai
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Load recommended model metadata from JSON

Browse Source
This commit is contained in:
Dobromir Popov
2026-07-01 12:57:23 +02:00
parent bc760c1694
commit b035338e58
8 changed files with 402 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
QUICKSTART.md
View File

@@ -86,12 +86,18 @@ python -c "import torch; print(torch.__version__, torch.cuda.is_available())"
If you get `ModuleNotFoundError: No module named 'torch'` even though `pip install torch`
says "already satisfied", the `torch/` package directory is missing while the metadata
stub remains (can happen after a conda-managed install). Force-reinstall via pip:
stub remains (can happen after a conda-managed install). Force-reinstall all three
PyTorch packages together so their versions stay in sync:
```powershell
pip install --force-reinstall torch --index-url https://download.pytorch.org/whl/cu118
pip install --force-reinstall torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
```
> **Important:** always reinstall `torch`, `torchvision`, and `torchaudio` as a group.
> Upgrading only `torch` leaves `torchvision` on an incompatible version, which causes
> `RuntimeError: operator torchvision::nms does not exist` and makes transformers fail
> to import any model class (the error surfaces as `Could not import module 'Qwen2ForCausalLM'`).
Then re-run the verify step above.
If that prints `True` but `meshnet-node` still can't find torch, the venv entry point