inclusionAI/Ling-lite
812.5k
1{2 "architectures": [3 "BailingMoeForCausalLM"4 ],5 "attention_dropout": 0.0,6 "auto_map": {7 "AutoConfig": "configuration_bailing_moe.BailingMoeConfig",8 "AutoModel": "modeling_bailing_moe.BailingMoeModel",9 "AutoModelForCausalLM": "modeling_bailing_moe.BailingMoeForCausalLM",10 "AutoModelForTokenClassification": "modeling_bailing_moe.BailingMoeForTokenClassification"11 },12 "eos_token_id": 126081,13 "pad_token_id": 126081,14 "first_k_dense_replace": 0,15 "hidden_act": "silu",16 "hidden_size": 2048,17 "initializer_range": 0.006,18 "intermediate_size": 5632,19 "max_position_embeddings": 32768,20 "model_type": "bailing_moe",21 "moe_intermediate_size": 1408,22 "num_experts": 64,23 "num_shared_experts": 2,24 "norm_topk_prob": true,25 "num_attention_heads": 16,26 "num_experts_per_tok": 6,27 "num_hidden_layers": 28,28 "num_key_value_heads": 4,29 "pretraining_tp": 1,30 "rms_norm_eps": 1e-06,31 "rope_scaling": null,32 "rope_theta": 600000,33 "tie_word_embeddings": false,34 "torch_dtype": "bfloat16",35 "transformers_version": "4.36.0",36 "use_cache": true,37 "use_bias": false,38 "use_qkv_bias": false,39 "vocab_size": 126464,40 "embedding_dropout": 0.0,41 "norm_head": true,42 "norm_softmax": false,43 "output_dropout": 0.0,44 "output_router_logits": false45}46 