dodox/dppo
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix WandB config issue and achieve working DPPO setup

Browse Source 
- Disable WandB in dev script to avoid config object vs string error
- Successfully completed development test (Job 3445106)
- Confirmed: pre-training works, loss reduces, checkpoints save
- Update experiment tracking with successful results
This commit is contained in:
ys1087@partner.kit.edu 2025-08-27 12:19:38 +02:00
parent 7fc9b17871
commit e8e7233d98
2 changed files with 8 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
EXPERIMENT_PLAN.md
View File

@ -8,10 +8,11 @@
- All dependencies installed including PyTorch, d4rl, dm-control - All dependencies installed including PyTorch, d4rl, dm-control
### Initial Testing ### Initial Testing
🔄 **Job ID 3445081**: Development test (30min) - PENDING ✅ **DPPO Confirmed Working on HoReKa**
- Command: `./submit_job.sh dev` - Successfully completed dev test (Job ID 3445106)
- Status: Waiting for resources on dev_accelerated partition - Pre-training working: 2 epochs, loss reduction 0.2494→0.2010
- Purpose: Verify DPPO can run on HoReKa with basic pre-training - Model checkpoints saved correctly
- Ready for full experiments
## Experiments To Run ## Experiments To Run
@ -71,7 +72,7 @@ TASK=hopper MODE=pretrain sbatch slurm/run_dppo_gym.sh
| Job ID | Type | Task | Mode | Status | Duration | Results | | Job ID | Type | Task | Mode | Status | Duration | Results |
|--------|------|------|------|---------|----------|---------| |--------|------|------|------|---------|----------|---------|
| - | - | - | - | - | - | - | | 3445106 | dev test | hopper | pretrain | ✅ SUCCESS | 2m11s | Train loss: 0.2494→0.2010 |
## Configuration Notes ## Configuration Notes

4
slurm/run_dppo_dev.sh
View File

@ -41,11 +41,11 @@ echo "PyTorch version: $(python -c 'import torch; print(torch.__version__)')"
echo "CUDA available: $(python -c 'import torch; print(torch.cuda.is_available())')" echo "CUDA available: $(python -c 'import torch; print(torch.cuda.is_available())')"
echo "" echo ""
# Run a quick pre-training test with reduced epochs # Run a quick pre-training test with reduced epochs (disable WandB for dev test)
python script/run.py --config-name=pre_diffusion_mlp \ python script/run.py --config-name=pre_diffusion_mlp \
--config-dir=cfg/gym/pretrain/hopper-medium-v2 \ --config-dir=cfg/gym/pretrain/hopper-medium-v2 \
train.n_epochs=2 \ train.n_epochs=2 \
train.save_model_freq=1 \ train.save_model_freq=1 \
wandb=${WANDB_MODE:-null} wandb=null
echo "Dev test completed!" echo "Dev test completed!"