fix(hpc): enable wandb tracking by default on compute nodes
This commit is contained in:
parent
205b284b96
commit
f2139fb419
2 changed files with 0 additions and 7 deletions
|
|
@ -39,9 +39,3 @@ Submit long-running experiments via the provided Slurm script:
|
|||
```bash
|
||||
sbatch scripts/hpc_train.slurm
|
||||
```
|
||||
|
||||
### Syncing WandB
|
||||
Since compute nodes are offline, sync your WandB logs after the job finishes:
|
||||
```bash
|
||||
wandb sync runs/<run_name>/wandb
|
||||
```
|
||||
|
|
|
|||
|
|
@ -25,7 +25,6 @@ source "$VSC_DATA/venvs/sel3_${VSC_INSTITUTE_CLUSTER}/bin/activate"
|
|||
|
||||
# -- HPC-specific environment flags -------------------------------------------
|
||||
export MUJOCO_GL=egl # Headless OpenGL via EGL (no display required)
|
||||
export WANDB_MODE=offline # Sync to WandB after the job (no outbound internet needed)
|
||||
|
||||
# -- Run from the project root -------------------------------------------------
|
||||
cd "$SLURM_SUBMIT_DIR"
|
||||
|
|
|
|||
Reference in a new issue