Use ZeroGPU-compatible PyTorch

README.md

@@ -42,10 +42,7 @@ cd physformer
 conda create -n physformer python=3.10 -y
 conda activate physformer
 
-# Install PyTorch (CUDA 12.4)
-pip install torch==2.5.1 torchvision==0.20.1 --index-url https://download.pytorch.org/whl/cu124
-
-# Install remaining requirements
+# Install requirements. The Hugging Face Space uses the ZeroGPU-supported PyTorch version pinned here.
 pip install -r requirements.txt
 ```
 
@@ -245,10 +242,10 @@ Use the project inference environment with PyTorch installed. The copied code ex
 - matplotlib
 - imageio and imageio-ffmpeg only if saving GIF/MP4 renders
 
-`requirements.txt` includes PyTorch so Hugging Face Spaces can build directly from this repository.
-For local installs, you may still install PyTorch explicitly first so pip chooses the intended CUDA
-wheel. The copied model uses PyTorch scaled-dot-product attention and requires a CUDA fast-attention
-backend; it does not directly import the external `flash-attn` package.
+`requirements.txt` includes a ZeroGPU-supported PyTorch version so Hugging Face Spaces can build
+directly from this repository. The copied model uses PyTorch scaled-dot-product attention and
+requires a CUDA fast-attention backend; it does not directly import the external `flash-attn`
+package.
 
 The packaged `src/official_demo_inference/configs/vertex_counts_multiobj_all.json` replaces the
 checkpoint's original training-machine absolute vertex-count path. A legacy copy is also kept under

requirements.txt

@@ -1,8 +1,6 @@
-# Hugging Face Spaces installs this file directly. For local installs, it is still fine to install
-# PyTorch explicitly first with the CUDA wheel index; pip will then treat it as already satisfied.
---extra-index-url https://download.pytorch.org/whl/cu124
-torch==2.5.1
-torchvision==0.20.1
+# Hugging Face ZeroGPU currently accepts only selected PyTorch versions.
+# The demo does not import torchvision.
+torch==2.8.0
 
 # Core runtime dependencies for the copied inference/render code.
 gradio>=4