{ "auto_map": { "AutoConfig": "configuration_llava_onevision2.LlavaOnevision2Config", "AutoModelForCausalLM": "modeling_llava_onevision2.LlavaOnevision2ForConditionalGeneration" }, "dtype": "bfloat16", "image_token_id": 151655, "model_type": "llava_onevision2", "text_config": { "_name_or_path": "/ov2/pretrain_models/Qwen3-8B/", "architectures": [ "LlavaOnevision2ForConditionalGeneration" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 151643, "dtype": "bfloat16", "eos_token_id": 151645, "head_dim": 128, "hidden_act": "silu", "hidden_size": 4096, "initializer_range": 0.02, "intermediate_size": 12288, "layer_types": [ "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention" ], "max_position_embeddings": 40960, "max_window_layers": 36, "model_type": "qwen3", "num_attention_heads": 32, "num_hidden_layers": 36, "num_key_value_heads": 8, "rms_norm_eps": 1e-06, "rope_scaling": null, "rope_theta": 8000000, "sliding_window": null, "use_cache": true, "use_sliding_window": false, "vocab_size": 151936 }, "transformers_version": "4.57.3", "video_token_id": 151656, "vision_config": { "attention_dropout": 0.0, "frame_windows_size": 4, "hidden_act": "gelu", "hidden_size": 1024, "image_size": 448, "initializer_range": 0.02, "intermediate_size": 4096, "layer_norm_eps": 1e-06, "layer_norm_type": "layer_norm", "max_position_embeddings": 8192, "model_type": "llava_onevision2", "num_attention_heads": 16, "num_channels": 3, "num_hidden_layers": 24, "out_hidden_size": 4096, "patch_position_encoding_type": "absolute", "patch_size": 14, "rope_theta": 10000.0, "spatial_merge_size": 2, "temporal_patch_size": 1, "text_hidden_size": 4096, "tokens_per_second": 1, "use_head": false, "use_patch_position_encoding": false }, "vision_end_token_id": 151653, "vision_start_token_id": 151652 }