| { |
| "num_samples": 128, |
| "accuracy": 0.0, |
| "macro_f1": 0.0, |
| "labels": [ |
| "Close bottle cap", |
| "Grasp coffee scoop", |
| "Grasp gooseneck kettle", |
| "Hold coffee carafe", |
| "Hold gooseneck kettle", |
| "Lift gooseneck kettle", |
| "Move kettle", |
| "Move kettle away", |
| "Pick up kettle", |
| "Pick up white bottle", |
| "Place item on table", |
| "Place kettle on table", |
| "Position kettle to pour", |
| "Pour coffee", |
| "Pour liquid from white bottle", |
| "Pour milk into coffee", |
| "Transfer coffee to dripper", |
| "Wait/Prepare for pouring", |
| "unknown" |
| ], |
| "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", |
| "adapter_dir": "checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", |
| "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", |
| "eval_split": "train", |
| "train_split": "train", |
| "num_eval_episodes": 1, |
| "unseen_eval_labels": [], |
| "num_unseen_label_samples": 0, |
| "seen_label_accuracy": 0.0, |
| "unseen_label_accuracy": null, |
| "eval_label_counts": { |
| "Close bottle cap": 9, |
| "Pick up kettle": 8, |
| "Position kettle to pour": 8, |
| "Move kettle": 8, |
| "Hold coffee carafe": 8, |
| "Grasp coffee scoop": 8, |
| "Transfer coffee to dripper": 8, |
| "Hold gooseneck kettle": 8, |
| "Lift gooseneck kettle": 8, |
| "Move kettle away": 8, |
| "Wait/Prepare for pouring": 8, |
| "Pour coffee": 8, |
| "Grasp gooseneck kettle": 6, |
| "Place kettle on table": 6, |
| "Pick up white bottle": 6, |
| "Pour liquid from white bottle": 6, |
| "Place item on table": 6, |
| "Pour milk into coffee": 1 |
| }, |
| "json_validity_rate": 1.0, |
| "action_macro_f1": 0.0, |
| "subtask_accuracy": 0.015625, |
| "transition_accuracy": 0.0, |
| "next_action_accuracy": 0.0078125, |
| "contact_accuracy": 0.0, |
| "object_micro_f1": 0.031496062992125984, |
| "caption_window_grounding": { |
| "mrr": null, |
| "recall_at_5": null, |
| "note": "Grounding ranking requires a retrieval candidate set; JSON evidence_window is stored for later scoring." |
| } |
| } |