| { | |
| "architectures": [ | |
| "Grok1ModelForCausalLM" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_grok1.Grok1Config", | |
| "AutoModel": "modeling_grok1.Grok1Model", | |
| "AutoModelForCausalLM": "modeling_grok1.Grok1ModelForCausalLM" | |
| }, | |
| "vocab_size": 131072, | |
| "hidden_size": 6144, | |
| "intermediate_size": 32768, | |
| "num_hidden_layers": 64, | |
| "num_attention_heads": 48, | |
| "num_key_value_heads": 8, | |
| "attn_output_multiplier": 0.08838834764831845, | |
| "embedding_multiplier_scale": 78.38367176906169, | |
| "output_multiplier_scale": 0.5773502691896257, | |
| "max_attn_value": 30.0, | |
| "max_position_embeddings": 8192, | |
| "rms_norm_eps": 1e-5, | |
| "use_cache": true, | |
| "pad_token_id": 0, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "tie_word_embeddings": true, | |
| "num_experts_per_tok": 2, | |
| "num_experts": 8, | |
| "output_router_logits": false, | |
| "router_aux_loss_coef": 0.001, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.35.0" | |
| } |