| { | |
| "model_path": "/Users/duanyongyao/PythonProjects/SemanticGallery/.cache/mlx/siglip2-base-patch16-224-f32", | |
| "init_weights": null, | |
| "weights_file": "/Users/duanyongyao/PythonProjects/SemanticGallery/logs/semanticgallery_public_stage1/weights.safetensors", | |
| "precision": "bfloat16", | |
| "lr": 1e-05, | |
| "weight_decay": 0.01, | |
| "max_length": 64, | |
| "batch_size": 2, | |
| "epochs": 1, | |
| "seed": 20260403, | |
| "train_rows": 1587, | |
| "val_rows": 333, | |
| "policy": { | |
| "train": { | |
| "public": 1587, | |
| "private": 0, | |
| "unknown": 0, | |
| "sources": { | |
| "flickr30k": 819, | |
| "screen2words": 768 | |
| } | |
| }, | |
| "val": { | |
| "public": 0, | |
| "private": 0, | |
| "unknown": 0, | |
| "sources": {} | |
| } | |
| }, | |
| "trainable": { | |
| "text_last_n": 2, | |
| "vision_last_n": 2, | |
| "trainable_parameter_tensors": 83 | |
| }, | |
| "history": [ | |
| { | |
| "epoch": 1, | |
| "train_loss": 7.748603820800781e-07, | |
| "train_loss_last": 7.748603820800781e-07, | |
| "train_rows": 2, | |
| "train_steps": 1, | |
| "train_step_mean_ms": 292.7449999988312, | |
| "train_grad_norm_mean": 0.000255584716796875, | |
| "val_loss": 0.00033167004585266113, | |
| "val_rows": 2, | |
| "val_steps": 1 | |
| } | |
| ] | |
| } |