| { | |
| "activation_function": "swiglu", | |
| "add_bias": false, | |
| "add_qkv_bias": false, | |
| "attention_head_type": "mha", | |
| "attention_multiplier": 0.015625, | |
| "attention_softmax_in_fp32": true, | |
| "attn_pdrop": 0, | |
| "bos_token_id": 0, | |
| "embd_pdrop": 0, | |
| "eos_token_id": 0, | |
| "init_method": "mup", | |
| "initializer_range": 0.1, | |
| "layer_norm_epsilon": 1e-05, | |
| "m_emb": 12, | |
| "m_residual": 0.22, | |
| "m_width": 9, | |
| "model_type": "stickbreaking", | |
| "multi_query": false, | |
| "n_embd": 2304, | |
| "n_head": 36, | |
| "n_inner": 9216, | |
| "n_layer": 40, | |
| "n_positions": 4096, | |
| "normalization_function": "rmsnorm", | |
| "num_key_value_heads": 36, | |
| "pad_token_id": 50256, | |
| "position_embedding_type": "nope", | |
| "resid_pdrop": 0, | |
| "sb_remainder": false, | |
| "scale_attn_weights": true, | |
| "transformers_version": "4.45.0", | |
| "upcast_logits_for_loss": true, | |
| "use_cache": true, | |
| "vocab_size": 50304 | |
| } | |