Clarify that installation must run on GPU node
Remove manual installation instructions as PyTorch CUDA dependencies require GPU node
This commit is contained in:
parent
2be39c4f2e
commit
add21c7019
@ -64,15 +64,12 @@ The DPPO repository has been adapted to run on the HoReKa cluster. The original
|
|||||||
|
|
||||||
3. **Install the package and dependencies:**
|
3. **Install the package and dependencies:**
|
||||||
```bash
|
```bash
|
||||||
# Use the provided installation script
|
# Submit installation job (runs on dev node with GPU)
|
||||||
sbatch install_dppo.sh
|
sbatch install_dppo.sh
|
||||||
|
|
||||||
# Or install manually:
|
|
||||||
pip install --upgrade pip
|
|
||||||
pip install -e .
|
|
||||||
pip install -e .[gym] # For Gym environments
|
|
||||||
```
|
```
|
||||||
|
|
||||||
|
Note: Installation must run on a GPU node due to PyTorch CUDA dependencies. The installation script automatically requests appropriate resources.
|
||||||
|
|
||||||
### Running on HoReKa
|
### Running on HoReKa
|
||||||
|
|
||||||
The repository includes pre-configured SLURM scripts for job submission:
|
The repository includes pre-configured SLURM scripts for job submission:
|
||||||
|
Loading…
Reference in New Issue
Block a user