{ "architectures": [ "Mistral3ForConditionalGeneration" ], "bos_token_id": 2, "dtype": "float32", "eos_token_id": 3, "image_seq_length": 4, "image_token_index": 1, "model_dtype": "mistral3", "model_type": "mistral3", "multimodal_projector_bias": false, "pad_token_id": 4, "projector_hidden_act": "gelu", "spatial_merge_size": 2, "text_config": { "attention_dropout": 0.0, "bos_token_id": 2, "eos_token_id": 3, "head_dim": null, "hidden_act": "silu", "hidden_size": 16, "initializer_range": 0.02, "intermediate_size": 37, "max_position_embeddings": 512, "model_type": "mistral", "num_attention_heads": 4, "num_hidden_layers": 1, "num_key_value_heads": 2, "pad_token_id": 4, "rms_norm_eps": 1e-05, "rope_theta": 1000000000.0, "sliding_window": null, "use_cache": true, "vocab_size": 32000 }, "transformers_version": "4.57.1", "vision_config": { "attention_dropout": 0.0, "head_dim": 4, "hidden_act": "gelu", "hidden_size": 16, "image_size": 30, "initializer_range": 0.02, "intermediate_size": 37, "model_type": "pixtral", "num_attention_heads": 4, "num_channels": 3, "num_hidden_layers": 1, "patch_size": 6, "rope_theta": 10000.0 }, "vision_feature_layer": -1 }