inclusionAI/LLaDA2.2-mini
29912
1{2 "architectures": [3 "LLaDA2MoeModelLM"4 ],5 "attention_dropout": 0.0,6 "auto_map": {7 "AutoConfig": "configuration_llada2_moe.LLaDA2MoeConfig",8 "AutoModel": "modeling_llada2_moe.LLaDA2MoeModel",9 "AutoModelForCausalLM": "modeling_llada2_moe.LLaDA2MoeModelLM"10 },11 "num_hidden_layers": 20,12 "hidden_size": 2048,13 "intermediate_size": 5120,14 "first_k_dense_replace": 1,15 "hidden_act": "silu",16 "max_position_embeddings": 131072,17 "model_type": "llada2_moe",18 "moe_intermediate_size": 512,19 "norm_topk_prob": true,20 "num_experts_per_tok": 8,21 "expert_capacity": 48,22 "block_size": 32,23 "delete_token_id": 156930,24 "split_token_id": 156931,25 "norm_head": false,26 "num_attention_heads": 16,27 "num_experts": 256,28 "num_key_value_heads": 4,29 "rope_theta": 3000000,30 "rope_scaling": null,31 "tie_word_embeddings": false,32 "torch_dtype": "bfloat16",33 "transformers_version": "5.2.0",34 "use_bias": false,35 "use_rmsnorm": true,36 "rms_norm_eps": 1e-06,37 "head_dim": 128,38 "num_shared_experts": 1,39 "use_cache": false,40 "use_qk_norm": true,41 "use_qkv_bias": false,42 "embedding_dropout": 0.0,43 "norm_softmax": false,44 "output_dropout": 0.0,45 "vocab_size": 157184,46 "rotary_dim": 64,47 "using_split_qkv_in_self_attention": false,48 "router_dtype": "fp32",49 "moe_router_enable_expert_bias": true,50 "routed_scaling_factor": 2.5,51 "n_group": 8,52 "topk_group": 4,53 "score_function": "sigmoid",54 "initializer_range": 0.02,55 "max_window_layers": 28,56 "output_router_logits": false,57 "pad_token_id": 156892,58 "partial_rotary_factor": 0.5,59 "use_sliding_window": false60}61 