Upload PolyGuard artifact: promoted qwen-qwen2-5-3b-instruct
Browse files
outputs/reports/grpo_trl_run.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"status": "ok",
|
| 3 |
+
"backend": "trl_transformers",
|
| 4 |
+
"model_id": "Qwen/Qwen2.5-3B-Instruct",
|
| 5 |
+
"records": 2000,
|
| 6 |
+
"prompts_path": "/app/data/processed/training_corpus_grpo_prompts.jsonl",
|
| 7 |
+
"reward_summary": {
|
| 8 |
+
"count": 4000,
|
| 9 |
+
"avg_reward": 0.767,
|
| 10 |
+
"avg_reward_components": {
|
| 11 |
+
"format_compliance_score": 0.999,
|
| 12 |
+
"candidate_alignment_score": 0.999,
|
| 13 |
+
"legality_score": 0.929,
|
| 14 |
+
"safety_delta_score": 0.497,
|
| 15 |
+
"burden_improvement_score": 0.469,
|
| 16 |
+
"disease_stability_score": 0.861,
|
| 17 |
+
"dosing_quality_score": 0.526,
|
| 18 |
+
"abstention_quality_score": 0.56,
|
| 19 |
+
"efficiency_score": 0.849,
|
| 20 |
+
"process_fidelity_score": 0.856,
|
| 21 |
+
"explanation_grounding_score": 0.795,
|
| 22 |
+
"anti_cheat_score": 0.589,
|
| 23 |
+
"uncertainty_calibration_score": 0.747
|
| 24 |
+
},
|
| 25 |
+
"avg_primary_reward_channels": {
|
| 26 |
+
"safety_legality": 0.816,
|
| 27 |
+
"clinical_improvement": 0.609,
|
| 28 |
+
"dosing_quality": 0.543,
|
| 29 |
+
"process_integrity": 0.875
|
| 30 |
+
}
|
| 31 |
+
},
|
| 32 |
+
"reward_log": "/app/checkpoints/sweeps/qwen-qwen2-5-3b-instruct/grpo_reward_components.jsonl",
|
| 33 |
+
"train_metrics": {
|
| 34 |
+
"train_runtime": 6873.9375,
|
| 35 |
+
"train_samples_per_second": 0.291,
|
| 36 |
+
"train_steps_per_second": 0.291,
|
| 37 |
+
"total_flos": 0.0,
|
| 38 |
+
"train_loss": 2.665005830824185e-06
|
| 39 |
+
},
|
| 40 |
+
"history_path": "/app/checkpoints/sweeps/qwen-qwen2-5-3b-instruct/grpo_history.json",
|
| 41 |
+
"artifact_path": "/app/checkpoints/sweeps/qwen-qwen2-5-3b-instruct/grpo_adapter",
|
| 42 |
+
"unsloth_available": false
|
| 43 |
+
}
|