{ "architectures": [ "Dots3NoteForConditionalGeneration" ], "audio_config": { "adapter_input_size": 8, "adapter_output_size": 8, "attention_bias": true, "attention_dropout": 0.0, "conv_output_size": 2, "downsample_hidden_size": 2, "dropout": 0.0, "dtype": "bfloat16", "feature_size": 8, "head_dim": 8, "hidden_act": "silu", "hidden_size": 8, "hop_length": 160, "intermediate_size": 16, "max_position_embeddings": 128, "model_type": "dots3_note_audio_encoder", "num_attention_heads": 1, "num_hidden_layers": 1, "num_key_value_heads": 1, "rope_parameters": { "partial_rotary_factor": 0.5, "rope_theta": 10000.0, "rope_type": "default" }, "samples_per_token": 1280, "sampling_rate": 16000 }, "audio_token_id": 77, "dtype": "bfloat16", "image_token_id": 17, "model_type": "dots3_note", "text_config": { "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 0, "dtype": "bfloat16", "eos_token_id": 25, "head_dim": 4, "hidden_act": "silu", "hidden_size": 8, "index_head_dim": 8, "index_n_heads": 2, "index_topk": 4, "initializer_range": 0.02, "intermediate_size": 16, "kv_lora_rank": 2, "layer_types": [ "indexed_attention", "indexed_attention", "sliding_attention" ], "max_position_embeddings": 512, "mlp_bias": false, "mlp_layer_types": [ "dense", "sparse", "sparse" ], "model_type": "dots3_note_text", "moe_intermediate_size": 4, "n_group": 1, "n_routed_experts": 3, "n_shared_experts": 1, "norm_topk_prob": true, "num_attention_heads": 2, "num_experts_per_tok": 2, "num_hidden_layers": 3, "num_key_value_heads": 2, "output_router_logits": false, "pad_token_id": 16, "per_layer_config": { "2": { "head_dim": 8, "kv_lora_rank": 4, "num_attention_heads": 1, "num_key_value_heads": 1, "qk_nope_head_dim": 6 } }, "q_lora_rank": 4, "qk_nope_head_dim": 2, "qk_rope_head_dim": 2, "rms_norm_eps": 1e-05, "rope_parameters": { "indexed_attention": { "partial_rotary_factor": 0.5, "rope_theta": 80000000.0, "rope_type": "default" }, "sliding_attention": { "partial_rotary_factor": 0.25, "rope_theta": 50000.0, "rope_type": "default" } }, "routed_scaling_factor": 1.0, "shared_experts_intermediate_size": 4, "sliding_window": 4, "tie_word_embeddings": false, "topk_group": 1, "use_cache": true, "v_head_dim": 4, "vocab_size": 341 }, "tie_word_embeddings": false, "transformers_version": "5.19.0.dev0", "video_token_id": 37, "vision_config": { "adapter_in_dim": 8, "adapter_merge_size": 2, "adapter_out_dim": 8, "attention_bias": false, "attention_dropout": 0.0, "dtype": "bfloat16", "embed_dim": 8, "hidden_act": "silu", "hidden_size": 8, "initializer_range": 0.02, "intermediate_size": 16, "mlp_bias": false, "mlp_layer_types": [ "dense", "sparse", "sparse" ], "model_type": "dots3_note_vision_encoder", "moe_intermediate_size": 4, "num_attention_heads": 2, "num_channels": 3, "num_experts_per_tok": 2, "num_hidden_layers": 3, "num_local_experts": 3, "patch_size": 2, "per_layer_config": { "2": { "num_local_experts": 4 } }, "rms_norm_eps": 1e-05, "rope_parameters": { "rope_theta": 10000.0, "rope_type": "axial" }, "router_scaling_factor": 1.0, "spatial_merge_size": 2 } }