3853988af7
- Fixed PSO budget calculation bug in ml/src/hyperopt/optimizer.rs - Root cause: Division by n_particles in sequential execution - Now correctly calculates max_iters = remaining_trials (no division) - Result: 50 trials complete instead of 23 (100% vs 46%) - Added comprehensive DQN hyperopt results analysis - 39/50 trials analyzed across 2 RunPod deployments - Best hyperparameters identified: LR 4.89e-5 (ultra-low) - Created DQN_HYPEROPT_RESULTS_SUMMARY.md with expert validation - GitLab CI/CD pipeline operational (48 lines fixed) - Fixed YAML syntax errors (unquoted colons) - All 7 jobs validated and working - Warning cleanup complete (136 → 0 warnings) - Removed 143 lines dead code - Fixed visibility, unused imports, Debug traits - Archived Wave D reports to docs/archive/ - 8 early stopping reports moved - Root directory cleaned up 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
39 lines
1.7 KiB
Markdown
39 lines
1.7 KiB
Markdown
# Deprecated Deployment Scripts
|
|
|
|
**Archive Date**: 2025-11-02
|
|
**Reason**: Consolidated duplicate deployment scripts to reduce confusion
|
|
|
|
## Why These Scripts Were Deprecated
|
|
|
|
During the hyperopt optimization phase (Oct-Nov 2025), multiple versions of deployment scripts were created:
|
|
- `*_corrected.sh`: Fixed objective functions (validation loss → episode rewards)
|
|
- `*_batch256.sh`: Experimental batch size configurations
|
|
- `*_rtx4090_corrected.sh`: GPU-specific optimizations
|
|
- `*_completion.sh`: Intermediate test scripts
|
|
|
|
After hyperopt completion, the canonical versions were consolidated into:
|
|
- `deploy_{model}_hyperopt.sh`: Standard hyperparameter optimization
|
|
- `deploy_{model}_production.sh`: Production training with optimized hyperparameters
|
|
|
|
## Migration Guide
|
|
|
|
| Deprecated Script | Canonical Replacement |
|
|
|---|---|
|
|
| `deploy_dqn_hyperopt_corrected.sh` | `deploy_dqn_hyperopt.sh` |
|
|
| `deploy_ppo_hyperopt_corrected.sh` | `deploy_ppo_hyperopt.sh` |
|
|
| `deploy_ppo_hyperopt_completion.sh` | `deploy_ppo_hyperopt.sh` |
|
|
| `deploy_ppo_production_corrected.sh` | `deploy_ppo_production.sh` |
|
|
| `deploy_mamba2_hyperopt_batch256.sh` | `deploy_mamba2_hyperopt.sh` |
|
|
| `deploy_mamba2_hyperopt_rtx4090_corrected.sh` | `deploy_mamba2_hyperopt.sh` |
|
|
|
|
## Key Changes in Canonical Scripts
|
|
|
|
1. **DQN Hyperopt**: Uses episode rewards as objective (not validation loss)
|
|
2. **PPO Hyperopt**: Optimizes dual learning rates (policy + value)
|
|
3. **MAMBA2 Hyperopt**: Standardized batch size (no hardcoded 256)
|
|
4. **All Scripts**: Updated with historical results and best hyperparameters
|
|
|
|
## Archived Scripts
|
|
|
|
All deprecated scripts preserved here for historical reference. Do not use in production.
|