{ "architectures": [ "Idefics2ForConditionalGeneration" ], "dtype": "bfloat16", "image_token_id": 32001, "model_type": "idefics2", "perceiver_config": { "attention_dropout": 0.0, "hidden_act": "silu", "hidden_size": 16, "initializer_range": 0.02, "model_type": "idefics2_perceiver", "num_key_value_heads": 4, "resampler_depth": 3, "resampler_head_dim": 96, "resampler_n_heads": 16, "resampler_n_latents": 64, "rms_norm_eps": 1e-06 }, "text_config": { "attention_dropout": 0.0, "head_dim": null, "hidden_act": "silu", "hidden_size": 16, "initializer_range": 0.02, "intermediate_size": 14336, "layer_types": null, "max_position_embeddings": 32768, "model_type": "mistral", "num_attention_heads": 4, "num_hidden_layers": 2, "num_key_value_heads": 2, "pad_token_id": 0, "rms_norm_eps": 1e-05, "rope_theta": 10000.0, "sliding_window": 4096, "use_cache": true, "vocab_size": 32003 }, "tie_word_embeddings": false, "transformers_version": "4.57.0.dev0", "use_cache": true, "vision_config": { "attention_dropout": 0.0, "embed_dim": 64, "hidden_act": "gelu_pytorch_tanh", "hidden_size": 16, "image_size": 980, "initializer_range": 0.02, "intermediate_size": 4304, "layer_norm_eps": 1e-06, "model_type": "idefics2_vision", "num_attention_heads": 4, "num_channels": 3, "num_hidden_layers": 2, "num_key_value_heads": 2, "patch_size": 14 } }