{ "model_type": "fusion-embedding-connector", "d_resampler": 384, "n_query": 64, "d_llm": 2048, "mrl_dims": [ 2048, 1536, 1024, 512, 256, 128, 64 ], "trained_params": 16397952, "protocol": { "text_query_template": "<|im_start|>system\n{instruction}<|im_end|>\n<|im_start|>user\n{text}<|im_end|>\n<|im_start|>assistant\n", "image_doc_template": "<|im_start|>system\nRepresent the user's input.<|im_end|>\n<|im_start|>user\n<|vision_start|><|image_pad|><|vision_end|><|im_end|>\n<|im_start|>assistant\n", "default_query_instruction": "Retrieve images or text relevant to the user's query.", "pooling": "last non-pad token", "cross_modal_readout": "per-modality mean-centering recommended (see inference.center)" }, "benchmarks": { "audiocaps_883_minrank5": { "eval_shard": "audiocaps_test816", "n_clips": 883, "n_captions": 4411, "dim": 1024, "n_total_clips": 883, "split_filtered": false, "protocol": "min-rank-over-refs", "a2t_R@1": 0.33182331919670105, "a2t_R@5": 0.6024914979934692, "a2t_R@10": 0.740656852722168, "a2t_mAP@10": 0.23749861121177673, "t2a_R@1": 0.2799818515777588, "t2a_R@5": 0.6023577451705933, "t2a_R@10": 0.7463160157203674, "t2a_mAP@10": 0.4181894063949585 }, "in_domain_eval": { "a2t_R@1": 0.4699999988079071, "a2t_R@10": 0.8849999904632568, "a2t_mAP@10": 0.6164186596870422 } }, "d_audio": 3584, "resampler_depth": 6, "resampler_heads": 8, "resampler_ffn_mult": 4, "resampler_dropout": 0.0, "adapter_act": "silu", "mrl_default": 1024, "audio_pad_id": 151654, "eos_id": 151645, "pad_id": 151643, "audio_pad_token": "<|audio_pad|>", "base_model": "Qwen/Qwen3-VL-Embedding-2B", "audio_model": "Qwen/Qwen2.5-Omni-7B", "max_text_tokens": 512, "n_decoder_layers": 28, "architectures": [ "FusionEmbeddingModel" ], "auto_map": { "AutoConfig": "configuration_fusion_embedding.FusionEmbeddingConfig", "AutoModel": "modeling_fusion_embedding.FusionEmbeddingModel" }, "adapter_rank": 0 }