| { |
| "student": "outputs/release/unhealed/winnow-olmoe-math-keep50", |
| "teacher": "allenai/OLMoE-1B-7B-0125-Instruct", |
| "training_mode": "on-policy", |
| "kl_direction": "reverse", |
| "dataset": "allenai/Dolci-Instruct-RL", |
| "dataset_sources": null, |
| "max_difficulty": null, |
| "trajectories": "outputs/teacher_trajectories/dolci_math_curated.jsonl", |
| "trajectory_dataset": "allenai/Dolci-Instruct-RL", |
| "off_policy_frames": "chat", |
| "off_policy_max_seq_len": 2048, |
| "topk_targets": null, |
| "max_loss_tokens": null, |
| "loss_tokens_per_step": null, |
| "teacher_device": "cuda:0", |
| "student_device": "cuda:1", |
| "lr": 1e-05, |
| "optimizer": "adamw8bit", |
| "weight_decay": 0.1, |
| "epochs": 2, |
| "prompts_per_step": 256, |
| "group_size": 4, |
| "rollout_batch": 64, |
| "micro_batch": 4, |
| "max_new_tokens": 2048, |
| "max_prompt_len": 1024, |
| "warmup_steps": 10, |
| "max_grad_norm": 1.0, |
| "eval_every": 10, |
| "gsm8k_every": 50, |
| "gsm8k_n": 256, |
| "gsm8k_batch": 16, |
| "gsm8k_max_new_tokens": 512, |
| "gsm8k_frames": "chat", |
| "save_every": 50, |
| "out_dir": "outputs/redo_policy/cold_on_reverse", |
| "sweep": 150, |
| "wandb": true, |
| "wandb_project": "winnow-olmoe-release", |
| "wandb_run_name": "cold_on_reverse", |
| "wandb_run_id": null, |
| "wandb_resume": null, |
| "wandb_mode": "online", |
| "no_wandb_sync": false, |
| "debug": false, |
| "resume_from": null, |
| "start_step": 0, |
| "no_grad_checkpointing": false, |
| "seed": 1223, |
| "no_teacher_overlap": false, |
| "sync_checkpoints": false, |
| "rollout_engine": "vllm", |
| "vllm_gpu": "2", |
| "vllm_port": 8377, |
| "vllm_refresh_every": 1, |
| "vllm_serve_bin": "vllm-plugin/.venv/bin/python", |
| "vllm_gpu_mem_util": 0.85, |
| "liger_loss": false, |
| "gold_mix_lambda": 0.0, |
| "gold_topk_targets": null, |
| "gold_loss": "ce", |
| "gold_mix_decay": 0.0, |
| "fast_teacher": false, |
| "reference_kl_beta": 0.0, |
| "drop_truncated_rollouts": false, |
| "vllm_max_model_len": null, |
| "vllm_refresh_mode": "reload", |
| "vllm_live_dir": null, |
| "resolved_kl_direction": "reverse" |
| } |
