Datasets:
| { | |
| "checkpoint": "/net/work/wangq/CycAMP/Scripts/Generation_Model/Models_ckpt/260226_content/GPT/MEDIUM/best.pt", | |
| "tokenizer_path": "/net/work/wangq/CycAMP/Scripts/Generation_Model/Tokenizer/content_vocab.json", | |
| "test_data_path": "/net/work/wangq/CycAMP/Scripts/Generation_Model/Data/processed/hierarchical/test_content.jsonl", | |
| "config": { | |
| "arch": "gpt", | |
| "d_model": 512, | |
| "num_layers": 8, | |
| "max_len": 512, | |
| "dropout": 0.35, | |
| "label_smoothing": 0.0 | |
| }, | |
| "evaluation": { | |
| "device": "cuda", | |
| "batch_size": 384, | |
| "num_samples": 75413, | |
| "timestamp": "2026-02-27T21:34:54.593737" | |
| }, | |
| "metrics": { | |
| "loss": 1.91179130749414, | |
| "perplexity": 6.76519649817366 | |
| } | |
| } |