| { | |
| "d_model": 384, | |
| "n_heads": 8, | |
| "n_layers": 6, | |
| "d_ff": 1536, | |
| "batch_size": 20, | |
| "max_steps": 30000, | |
| "n_kv_heads": 4, | |
| "sliding_window": 4096, | |
| "attention_bias": false, | |
| "rms_norm_eps": 1e-06, | |
| "gradient_accumulation_steps": 4, | |
| "muon_lr": 0.01, | |
| "max_seq_len": 512, | |
| "num_documents": 2000, | |
| "max_tokens": 500000, | |
| "eval_every": 500, | |
| "eval_steps": 100, | |
| "weight_decay": 0.1, | |
| "dropout": 0.1, | |
| "grad_clip": 1.0, | |
| "use_amp": true, | |
| "vocab_size": 63982, | |
| "model_type": "minimal_llm", | |
| "architectures": [ | |
| "MinimalLLM" | |
| ], | |
| "tokenizer_class": "TokenizersBackend", | |
| "pad_token_id": 63982, | |
| "eos_token_id": 63982, | |
| "bos_token_id": 63982, | |
| "unk_token_id": 1, | |
| "torch_dtype": "float32", | |
| "transformers_version": "5.13.1", | |
| "special_tokens_map": { | |
| "bos_token": "<|endoftext|>", | |
| "eos_token": "<|endoftext|>", | |
| "unk_token": "<unk>", | |
| "pad_token": "<|endoftext|>" | |
| } | |
| } |