| { | |
| "model_type": "trm", | |
| "architectures": [ | |
| "TinyRecursiveModel" | |
| ], | |
| "torch_dtype": "float32", | |
| "vocab_size": 50257, | |
| "dim": 256, | |
| "n_heads": 8, | |
| "n_layers": 2, | |
| "mlp_ratio": 4, | |
| "max_seq_len": 512, | |
| "n_latent_recursions": 4, | |
| "n_improvement_cycles": 2, | |
| "batch_size": 4, | |
| "epochs": 1, | |
| "lr": 0.0001, | |
| "warmup_steps": 1000, | |
| "n_supervision_steps": 3, | |
| "max_train_samples": null, | |
| "max_val_samples": null | |
| } |