| { | |
| "A_initializer_range": [ | |
| 1, | |
| 16 | |
| ], | |
| "_name_or_path": "OuteAI/Lite-Oute-2-Mamba2Attn-250M-Instruct", | |
| "architectures": [ | |
| "Mamba2ForCausalLM" | |
| ], | |
| "attention_conv_kernel": 4, | |
| "attention_head_dim": 128, | |
| "attention_layers_idx": [ | |
| 6, | |
| 12, | |
| 18, | |
| 24 | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_mamba2attn.Mamba2Config", | |
| "AutoModelForCausalLM": "modeling_mamba2attn.Mamba2ForCausalLM" | |
| }, | |
| "bos_token_id": 1, | |
| "chunk_size": 256, | |
| "classifier_dropout": 0.1, | |
| "conv_initializer_range": null, | |
| "emb_initializer_range": 0.02, | |
| "eos_token_id": 2, | |
| "expand": 2, | |
| "hidden_act": "silu", | |
| "hidden_size": 1024, | |
| "intermediate_size": 2048, | |
| "layer_norm_epsilon": 1e-05, | |
| "mamba2_conv_kernel": 4, | |
| "mamba2_head_dim": 64, | |
| "mamba2_num_heads": 32, | |
| "max_position_embeddings": 8192, | |
| "mlp_intermediate_size": 0, | |
| "mlp_padding_size": 128, | |
| "model_type": "mamba2", | |
| "num_attention_heads": 16, | |
| "num_hidden_layers": 32, | |
| "num_key_value_heads": 16, | |
| "pad_token_id": 0, | |
| "rescale_prenorm_residual": false, | |
| "residual_in_fp32": true, | |
| "rope_emb_dim": 64, | |
| "rope_scaling": null, | |
| "rope_theta": 10000.0, | |
| "state_size": 128, | |
| "tie_embedding_weights": true, | |
| "time_step_floor": 0.0001, | |
| "time_step_limit": [ | |
| 0.0, | |
| Infinity | |
| ], | |
| "time_step_max": 0.1, | |
| "time_step_min": 0.001, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.44.2", | |
| "use_attention_out_bias": false, | |
| "use_attention_qkv_bias": false, | |
| "use_cache": true, | |
| "use_conv_bias": true, | |
| "use_mamba2_bias": false, | |
| "use_mlp_bias": false, | |
| "vocab_size": 32768 | |
| } | |