| { | |
| "model_type": "Gemma3", | |
| "architecture": "Gemma3-270M", | |
| "training_dataset": "TinyStories", | |
| "vocab_size": 256000, | |
| "context_length": 8192, | |
| "emb_dim": 2048, | |
| "n_heads": 8, | |
| "n_layers": 18, | |
| "hidden_dim": 16384, | |
| "head_dim": 256, | |
| "max_iters": 150000, | |
| "tokenizer": "gpt2" | |
| } |