| { |
| "_name_or_path": "base_LLM_200M/checkpoint-755000", |
| "abili": false, |
| "architectures": [ |
| "MptForCausalLM" |
| ], |
| "attn_config": { |
| "model_type": "" |
| }, |
| "d_model": 768, |
| "emb_pdrop": 0.0, |
| "embedding_fraction": 1.0, |
| "expansion_ratio": 4, |
| "init_device": "cpu", |
| "initializer_range": 0.02, |
| "layer_norm_epsilon": 1e-05, |
| "learned_pos_emb": true, |
| "logit_scale": null, |
| "max_seq_len": 8192, |
| "model_type": "mpt", |
| "n_heads": 12, |
| "n_layers": 12, |
| "no_bias": true, |
| "norm_type": "low_precision_layernorm", |
| "pad_token_id": 128001, |
| "resid_pdrop": 0.0, |
| "rope": true, |
| "rope_theta": 100000, |
| "torch_dtype": "float32", |
| "transformers_version": "4.40.1", |
| "use_cache": false, |
| "verbose": 0, |
| "vocab_size": 128256 |
| } |
|
|
