| { | |
| "_name_or_path": "hf/", | |
| "architectures": [ | |
| "GrokForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "attn_output_multiplier": 0.08838834764831845, | |
| "auto_map": { | |
| "AutoConfig": "configuration_grok.GrokConfig", | |
| "AutoModelForCausalLM": "modeling_grok.GrokForCausalLM" | |
| }, | |
| "bos_token_id": 1, | |
| "embedding_multiplier_scale": 78.38367176906169, | |
| "eos_token_id": 2, | |
| "hidden_act": "gelu_new", | |
| "hidden_size": 6144, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 32768, | |
| "max_position_embeddings": 8192, | |
| "model_type": "grok", | |
| "num_attention_heads": 48, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 64, | |
| "num_key_value_heads": 8, | |
| "num_local_experts": 8, | |
| "output_multiplier_scale": 0.5773502691896257, | |
| "output_router_logits": false, | |
| "pad_token_id": 0, | |
| "rms_norm_eps": 1e-05, | |
| "rope_theta": 10000.0, | |
| "router_aux_loss_coef": 0.02, | |
| "sliding_window": null, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.38.2", | |
| "use_cache": true, | |
| "vocab_size": 131072 | |
| } | |