{ "architectures": [ "GraniteSpeechPlusForConditionalGeneration" ], "audio_token_index": 0, "downsample_rate": 5, "dtype": "float32", "encoder_config": { "cat_hidden_layers": [ 0 ], "context_size": 200, "conv_expansion_factor": 2, "conv_kernel_size": 15, "dim_head": 16, "dropout": 0.1, "feedforward_mult": 4, "hidden_dim": 32, "input_dim": 80, "max_pos_emb": 512, "model_type": "granite_speech_plus_encoder", "num_heads": 2, "num_layers": 2, "output_dim": 42 }, "has_lora_adapter": true, "initializer_range": 0.02, "model_type": "granite_speech_plus", "pad_token_id": 100256, "projector_config": { "attention_probs_dropout_prob": 0.1, "cross_attention_frequency": 2, "encoder_hidden_size": 64, "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 32, "initializer_range": 0.02, "intermediate_size": 256, "layer_norm_eps": 1e-12, "max_position_embeddings": 512, "model_type": "blip_2_qformer", "num_attention_heads": 4, "num_hidden_layers": 2, "pad_token_id": 0, "use_qformer_text_input": false, "vocab_size": 30522 }, "text_config": { "attention_bias": false, "attention_dropout": 0.0, "attention_multiplier": 1.0, "bos_token_id": 100257, "embedding_multiplier": 1.0, "eos_token_id": 100257, "hidden_act": "gelu", "hidden_size": 32, "initializer_range": 0.02, "intermediate_size": 32, "logits_scaling": 1.0, "max_position_embeddings": 512, "mlp_bias": false, "model_type": "granite", "num_attention_heads": 2, "num_hidden_layers": 2, "num_key_value_heads": 2, "pad_token_id": 100256, "residual_multiplier": 1.0, "rms_norm_eps": 1e-06, "rope_parameters": { "rope_theta": 10000.0, "rope_type": "default" }, "tie_word_embeddings": false, "use_cache": true, "vocab_size": 100353 }, "tie_word_embeddings": true, "transformers_version": "5.16.0.dev0", "window_size": 15 }