stokiz/higgs-audio-v3-tts-4b-models-onnx

Higgs Audio v3 TTS model payload

This dataset was prepared by HiggsAudioModelsUploader.

• —Source repository: onnx-community/higgs-audio-v3-tts-4b
• —Revision: main
• —Layout: onnx
• —Required backend: higgs-audio-v3-onnx
• —Default model file: cpu_fp32/audio_embed.onnx
• —Contains Transformers trustremotecode files: False
• —Files: 50

Expected Kaggle runner environment:

HIGGS_AUDIO_MODEL_DATASET_REF=/higgs-audio-v3-tts-4b-models-onnx
HIGGS_AUDIO_BACKEND=onnx
HIGGS_AUDIO_DEVICE=cuda
HIGGS_AUDIO_ONNX_VARIANT=cuda_fp32
HIGGS_AUDIO_ONNX_QUANTIZATION=fp32
HIGGS_AUDIO_ONNX_EXECUTION_PROVIDER=cuda
HIGGS_AUDIO_ONNX_MODEL_FILE=

Reference-speaker audio is intentionally kept in a separate dataset selected by HIGGS_AUDIO_REFERENCE_DATASET_REF. Task rows use ReferenceAudio as a file name/path and ReferenceText as the transcript for that reference audio.

Payload files

• —assets/model_architecture.png (asset, 63078 bytes)
• —cpu_fp16/audio_embed.onnx (onnx-model, 42025695 bytes)
• —cpu_fp16/audio_encoder.onnx (onnx-model, 654423379 bytes)
• —cpu_fp16/audio_heads.onnx (onnx-model, 42026801 bytes)
• —cpu_fp16/audio_tokenizer.onnx (onnx-model, 86516167 bytes)
• —cpu_fp16/llm_decoder.onnx (onnx-model, 930593 bytes)
• —cpu_fp16/llm_decoder.onnx.data (onnx-external-data, 7275413504 bytes)
• —cpu_fp16/text_embed.onnx (onnx-model, 777912621 bytes)
• —cpu_fp32/audio_embed.onnx (onnx-model, 84050504 bytes)
• —cpu_fp32/audio_encoder.onnx (onnx-model, 654423379 bytes)
• —cpu_fp32/audio_heads.onnx (onnx-model, 84051322 bytes)
• —cpu_fp32/audio_tokenizer.onnx (onnx-model, 86516167 bytes)
• —cpu_fp32/llm_decoder.onnx (onnx-model, 317522 bytes)
• —cpu_fp32/llm_decoder.onnx.data (onnx-external-data, 14550827008 bytes)
• —cpu_fp32/text_embed.onnx (onnx-model, 1555824893 bytes)
• —cpu_int4/audio_embed.onnx (onnx-model, 13461914 bytes)
• —cpu_int4/audio_encoder.onnx (onnx-model, 654423379 bytes)
• —cpu_int4/audio_heads.onnx (onnx-model, 13462729 bytes)
• —cpu_int4/audio_tokenizer.onnx (onnx-model, 86516167 bytes)
• —cpu_int4/llm_decoder.onnx (onnx-model, 384495 bytes)
• —cpu_int4/llm_decoder.onnx.data (onnx-external-data, 2291924992 bytes)
• —cpu_int4/text_embed.onnx (onnx-model, 777912621 bytes)
• —cuda_fp16/audio_embed.onnx (onnx-model, 42025695 bytes)
• —cuda_fp16/audio_encoder.onnx (onnx-model, 654423379 bytes)
• —cuda_fp16/audio_heads.onnx (onnx-model, 42026801 bytes)
• —cuda_fp16/audio_tokenizer.onnx (onnx-model, 86516167 bytes)
• —cuda_fp16/llm_decoder.onnx (onnx-model, 317346 bytes)
• —cuda_fp16/llm_decoder.onnx.data (onnx-external-data, 7275413504 bytes)
• —cuda_fp16/text_embed.onnx (onnx-model, 777912621 bytes)
• —cuda_fp32/audio_embed.onnx (onnx-model, 84050504 bytes)
• —cuda_fp32/audio_encoder.onnx (onnx-model, 654423379 bytes)
• —cuda_fp32/audio_heads.onnx (onnx-model, 84051322 bytes)
• —cuda_fp32/audio_tokenizer.onnx (onnx-model, 86516167 bytes)
• —cuda_fp32/llm_decoder.onnx (onnx-model, 303211 bytes)
• —cuda_fp32/llm_decoder.onnx.data (onnx-external-data, 14550827008 bytes)
• —cuda_fp32/text_embed.onnx (onnx-model, 1555824893 bytes)
• —cuda_int4/audio_embed.onnx (onnx-model, 13461914 bytes)
• —cuda_int4/audio_encoder.onnx (onnx-model, 654423379 bytes)
• —cuda_int4/audio_heads.onnx (onnx-model, 13462729 bytes)
• —cuda_int4/audio_tokenizer.onnx (onnx-model, 86516167 bytes)
• —cuda_int4/llm_decoder.onnx (onnx-model, 380118 bytes)
• —cuda_int4/llm_decoder.onnx.data (onnx-external-data, 2054291456 bytes)
• —cuda_int4/text_embed.onnx (onnx-model, 777912621 bytes)
• —LICENSE (support, 25685 bytes)
• —openvino_int4/audio_embed.onnx (onnx-model, 13461914 bytes)
• —openvino_int4/audio_heads.onnx (onnx-model, 13462729 bytes)
• —openvino_int4/audio_tokenizer.onnx (onnx-model, 86516167 bytes)
• —openvino_int4/llm_decoder.onnx (onnx-model, 384495 bytes)
• —openvino_int4/llm_decoder.onnx.data (onnx-external-data, 2291924992 bytes)
• —README.md (support, 42844 bytes)