| { | |
| "always_save_checkpoint": true, | |
| "architectures": [ | |
| "CustomGPTModel" | |
| ], | |
| "backend": "nccl", | |
| "batch_size": 12, | |
| "beta1": 0.9, | |
| "beta2": 0.95, | |
| "bias": false, | |
| "block_size": 1024, | |
| "compile": true, | |
| "dataset": "openwebtext", | |
| "decay_lr": true, | |
| "device": "cuda", | |
| "dropout": 0.0, | |
| "dtype": "bfloat16", | |
| "eval_interval": 1, | |
| "eval_iters": 1, | |
| "eval_only": false, | |
| "grad_clip": 1.0, | |
| "gradient_accumulation_steps": 40, | |
| "init_from": "scratch", | |
| "learning_rate": 0.0006, | |
| "log_interval": 1, | |
| "lr_decay_iters": 5, | |
| "max_iters": 5, | |
| "min_lr": 6e-05, | |
| "model_type": "custom_gpt", | |
| "n_embd": 768, | |
| "n_head": 12, | |
| "n_layer": 12, | |
| "out_dir": "out", | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.42.3", | |
| "vocab_size": 50304, | |
| "wandb_log": false, | |
| "wandb_project": "owt", | |
| "wandb_run_name": "gpt2", | |
| "warmup_iters": 1, | |
| "weight_decay": 0.1 | |
| } | |