| { | |
| "architectures": [ | |
| "ArcticForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 1, | |
| "enable_expert_tensor_parallelism": false, | |
| "eos_token_id": 2, | |
| "hidden_act": "silu", | |
| "hidden_size": 7168, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 4864, | |
| "max_position_embeddings": 4096, | |
| "model_type": "yak", | |
| "moe_eval_capacity_factor": 1, | |
| "moe_expert_parallel_size": 8, | |
| "moe_layer_frequency": 1, | |
| "moe_min_capacity": 0, | |
| "moe_token_dropping": true, | |
| "moe_train_capacity_factor": 1, | |
| "num_attention_heads": 56, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 35, | |
| "num_key_value_heads": 56, | |
| "num_local_experts": 128, | |
| "parallel_attn_mlp_res": true, | |
| "quantization": null, | |
| "rms_norm_eps": 1e-05, | |
| "rope_theta": 1000000.0, | |
| "router_aux_loss_coef": 0.001, | |
| "sliding_window": null, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.40.0.dev0", | |
| "use_cache": false, | |
| "vocab_size": 32000 | |
| } |