ultraVAD / config.json
patricklifixie's picture
Copy from fixie-ai/turntaking-multilingual-llama8b-2a (rev: main)
3cc71d4 verified
{
"architectures": [
"UltravoxModel"
],
"audio_config": {
"_name_or_path": "openai/whisper-large-v3-turbo",
"activation_dropout": 0.0,
"activation_function": "gelu",
"apply_spec_augment": false,
"architectures": [
"WhisperForConditionalGeneration"
],
"attention_dropout": 0.0,
"bos_token_id": 50257,
"classifier_proj_size": 256,
"d_model": 1280,
"decoder_attention_heads": 20,
"decoder_ffn_dim": 5120,
"decoder_layerdrop": 0.0,
"decoder_layers": 4,
"decoder_start_token_id": 50258,
"dropout": 0.0,
"encoder_attention_heads": 20,
"encoder_ffn_dim": 5120,
"encoder_layerdrop": 0.0,
"encoder_layers": 32,
"eos_token_id": 50257,
"init_std": 0.02,
"mask_feature_length": 10,
"mask_feature_min_masks": 0,
"mask_feature_prob": 0.0,
"mask_time_length": 10,
"mask_time_min_masks": 2,
"mask_time_prob": 0.05,
"max_source_positions": 1500,
"max_target_positions": 448,
"median_filter_width": 7,
"model_type": "whisper",
"num_hidden_layers": 32,
"num_mel_bins": 128,
"pad_token_id": 50257,
"scale_embedding": false,
"torch_dtype": "float16",
"use_cache": true,
"use_weighted_layer_sum": false,
"vocab_size": 51866
},
"audio_latency_block_size": null,
"audio_model_id": null,
"audio_token_index": 128256,
"auto_map": {
"AutoConfig": "ultravox_config.UltravoxConfig",
"AutoModel": "ultravox_model.UltravoxModel",
"AutoProcessor": "fixie-ai/ultravox-v0_5-llama-3_1-8b--ultravox_processing.UltravoxProcessor"
},
"custom_pipelines": {
"ultravox-pipeline": {
"impl": "ultravox_pipeline.UltravoxPipeline",
"pt": [
"AutoModel"
],
"tf": [],
"type": "multimodal"
}
},
"hidden_size": 4096,
"ignore_index": -100,
"initializer_range": 0.02,
"llm_only_training": false,
"model_type": "ultravox",
"norm_init": 0.4,
"pad_token_id": 128009,
"projector_act": "swiglu",
"projector_ln_mid": true,
"stack_factor": 8,
"text_model_id": "fixie-ai/turntaking-pretraining-it-multilingual-3c",
"torch_dtype": "bfloat16",
"transformers_version": "4.51.3",
"vocab_size": 128256
}