diff options
| author | YurenHao0426 <blackhao0426@gmail.com> | 2026-02-10 09:50:33 -0600 |
|---|---|---|
| committer | YurenHao0426 <blackhao0426@gmail.com> | 2026-02-10 09:50:33 -0600 |
| commit | 039c12d3cf7178db6a7d80b02cf022d67231014e (patch) | |
| tree | b3104310bfaced0d992729f59f1a7ef2e769c6bd /scripts/slurm_train.sh | |
| parent | 80579d6cc254d337a23e71404ae7ecab1849d1e5 (diff) | |
Add auto-resume checkpointing, S1/S2 configs, and experiment results
- Auto-resume: find latest checkpoint in save_dir on startup
- SIGUSR1 handler: save checkpoint before SLURM timeout
- S1 config (constant tau=5, identity init verification)
- S2 config (constant tau=2, gradient flow check)
- Experiment results tracker with S0/S1 data
- Speed estimates and experiment plan
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Diffstat (limited to 'scripts/slurm_train.sh')
| -rw-r--r-- | scripts/slurm_train.sh | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/scripts/slurm_train.sh b/scripts/slurm_train.sh index e1df687..361ba94 100644 --- a/scripts/slurm_train.sh +++ b/scripts/slurm_train.sh @@ -1,4 +1,5 @@ #!/bin/bash +#SBATCH --signal=SIGUSR1@120 export HF_HOME=/projects/bfqt/users/yurenh2/hf_cache export TRANSFORMERS_CACHE=/projects/bfqt/users/yurenh2/hf_cache/transformers export HF_HUB_CACHE=/projects/bfqt/users/yurenh2/hf_cache/hub @@ -18,4 +19,5 @@ echo "GPU: $(nvidia-smi --query-gpu=name,memory.total --format=csv,noheader)" echo "" echo "=== Starting training ===" -python3 -u scripts/train.py --config configs/sanity_check.yaml +echo " Auto-resume enabled: will pick up from latest checkpoint" +python3 -u scripts/train.py --config ${CONFIG:-configs/sanity_check.yaml} |