niobures/ECAPA-TDNN

By niobures · ↓ 0 · ♥ 0 · Updated 2025-09-25

What's in this repo

This page lists the files in niobures/ECAPA-TDNN, including voxceleb_ECAPA512_LM.onnx, voxceleb_ECAPA1024_LM.onnx, voxceleb_ECAPA1024.onnx and more.

4 ONNX models (.onnx files)
141 files, including 20 .py, 18 .png, 15 .txt, 13 .md, 11 .json, 10 .zip
Last updated 2025-09-25

ONNX files

File	Size	Download
models/wespeaker-ecapa-tdnn512-LM/voxceleb_ECAPA512_LM.onnx	—	Download
models/wespeaker-voxceleb-ecapa-tdnn1024-LM/voxceleb_ECAPA1024_LM.onnx	—	Download
models/wespeaker-voxceleb-ecapa-tdnn1024/voxceleb_ECAPA1024.onnx	—	Download
models/wespeaker-voxceleb-ecapa-tdnn512/voxceleb_ECAPA512.onnx	—	Download

How to use

These ONNX files can be loaded with ONNX Runtime. To inspect a model graph, upload the file to the Netron viewer. Download links point to the original files on Hugging Face.

No README provided by the author.

Model info

coreml onnx safetensors region:us

↓ 0 ♥ 0

Created: 2025-09-25

Updated: 2025-09-25

View on Hugging Face

Files (141)

.gitattributes

ECAPA-TDNN based online discussion activity-level evaluation.pdf

ECAPA-TDNN. Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verification.pdf

Validation of an ECAPA-TDNN system for Forensic Automatic Speaker Recognition under case work conditions.pdf

code/AudioExperiementation (compare to SpeakerNet).zip

code/ECAPA-TDNN (lawlict).zip

code/ECAPA-TDNN-CNCeleb.zip

code/ECAPA-TDNN.zip

code/Speaker-Verification (NeMo).zip

code/WeSpeaker/releases/wespeaker-1.2.0.zip

code/WeSpeaker/wespeaker.zip

code/speaker-verification-ResnetSE-and-ECAPA-TDNN.zip

code/voice-gender-classifier.zip

models/Hyper-ECAPA-TDNN/.gitattributes

models/Hyper-ECAPA-TDNN/README.md

models/Hyper-ECAPA-TDNN/model_0002.model

models/Hyper-ECAPA-TDNN/source.txt

models/VPR_zhvoice_EcapaTdnn/.gitattributes

models/VPR_zhvoice_EcapaTdnn/README.md

models/VPR_zhvoice_EcapaTdnn/config.yml

models/VPR_zhvoice_EcapaTdnn/model.pt

models/VPR_zhvoice_EcapaTdnn/model.state

models/VPR_zhvoice_EcapaTdnn/optimizer.pt

models/VPR_zhvoice_EcapaTdnn/source.txt

models/cnceleb_ecapa_tdnn/.gitattributes

models/cnceleb_ecapa_tdnn/README.md

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/16epoch.pth

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/RESULTS.md

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/config.yaml

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/backward_time.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/clip.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/eer.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/forward_time.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/gpu_max_cached_mem_GB.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/grad_norm.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/iter_time.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/loss.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/loss_scale.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/mindcf.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/n_trials.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/nontrg_mean.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/nontrg_std.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/optim0_lr0.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/optim_step_time.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/train_time.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/trg_mean.png

models/cnceleb_ecapa_tdnn/exp/spk_train_ecapa_tdnn_raw_sp/images/trg_std.png

models/cnceleb_ecapa_tdnn/meta.yaml

models/cnceleb_ecapa_tdnn/source.txt

models/ecapa-tdnn-mlpackage/.gitattributes

models/ecapa-tdnn-mlpackage/ECAPA_TDNN_EndToEnd.mlpackage.zip

models/ecapa-tdnn-mlpackage/ECAPA_TDNN_EndToEnd.mlpackage/Data/com.apple.CoreML/model.mlmodel

models/ecapa-tdnn-mlpackage/ECAPA_TDNN_EndToEnd.mlpackage/Data/com.apple.CoreML/weights/weight.bin

models/ecapa-tdnn-mlpackage/ECAPA_TDNN_EndToEnd.mlpackage/Manifest.json

models/ecapa-tdnn-mlpackage/README.md

models/ecapa-tdnn-mlpackage/source.txt

models/ecapa-tdnn-vox2/.gitattributes

models/ecapa-tdnn-vox2/README.md

models/ecapa-tdnn-vox2/classifier.ckpt

models/ecapa-tdnn-vox2/embedding_model.ckpt

models/ecapa-tdnn-vox2/hyperparams.yaml

models/ecapa-tdnn-vox2/label_encoder.txt

models/ecapa-tdnn-vox2/source.txt

models/ecapa-tdnn-voxceleb1-c512-aam/.gitattributes

models/ecapa-tdnn-voxceleb1-c512-aam/README.md

models/ecapa-tdnn-voxceleb1-c512-aam/all_results.json

models/ecapa-tdnn-voxceleb1-c512-aam/angular_loss.py

models/ecapa-tdnn-voxceleb1-c512-aam/audio_processing.py

models/ecapa-tdnn-voxceleb1-c512-aam/config.json

models/ecapa-tdnn-voxceleb1-c512-aam/configuration_ecapa_tdnn.py

models/ecapa-tdnn-voxceleb1-c512-aam/conv_asr.py

models/ecapa-tdnn-voxceleb1-c512-aam/eval_results.json

models/ecapa-tdnn-voxceleb1-c512-aam/feature_extraction_ecapa_tdnn.py

models/ecapa-tdnn-voxceleb1-c512-aam/features.py

models/ecapa-tdnn-voxceleb1-c512-aam/log_history.json

models/ecapa-tdnn-voxceleb1-c512-aam/model.safetensors

models/ecapa-tdnn-voxceleb1-c512-aam/modeling_ecapa_tdnn.py

models/ecapa-tdnn-voxceleb1-c512-aam/module.py

models/ecapa-tdnn-voxceleb1-c512-aam/preprocessor_config.json

models/ecapa-tdnn-voxceleb1-c512-aam/source.txt

models/ecapa-tdnn-voxceleb1-c512-aam/spectrogram_augment.py

models/ecapa-tdnn-voxceleb1-c512-aam/tdnn_attention.py

models/ecapa-tdnn-voxceleb1-c512-aam/train_results.json

models/ecapa-tdnn-voxceleb1-c512-aam/trainer_state.json

models/ecapa-tdnn-voxceleb1-c512-aam/training_args.bin

models/ecapa-tdnn-voxceleb2-c512-aam/.gitattributes

models/ecapa-tdnn-voxceleb2-c512-aam/README.md

models/ecapa-tdnn-voxceleb2-c512-aam/angular_loss.py

models/ecapa-tdnn-voxceleb2-c512-aam/audio_processing.py

models/ecapa-tdnn-voxceleb2-c512-aam/config.json

models/ecapa-tdnn-voxceleb2-c512-aam/configuration_ecapa_tdnn.py

models/ecapa-tdnn-voxceleb2-c512-aam/conv_asr.py

models/ecapa-tdnn-voxceleb2-c512-aam/feature_extraction_ecapa_tdnn.py

models/ecapa-tdnn-voxceleb2-c512-aam/features.py

models/ecapa-tdnn-voxceleb2-c512-aam/model.safetensors

models/ecapa-tdnn-voxceleb2-c512-aam/modeling_ecapa_tdnn.py

models/ecapa-tdnn-voxceleb2-c512-aam/module.py

models/ecapa-tdnn-voxceleb2-c512-aam/preprocessor_config.json

models/ecapa-tdnn-voxceleb2-c512-aam/source.txt

models/ecapa-tdnn-voxceleb2-c512-aam/spectrogram_augment.py

models/ecapa-tdnn-voxceleb2-c512-aam/tdnn_attention.py

models/ecapa-tdnn-voxceleb2-c512-aam/training_args.bin

models/nemo-ecapa-tdnn/.gitattributes

models/nemo-ecapa-tdnn/model_config.yaml

models/nemo-ecapa-tdnn/model_weights.ckpt

models/nemo-ecapa-tdnn/source.txt

models/spkrec-ecapa-voxceleb/.gitattributes

models/spkrec-ecapa-voxceleb/README.md

models/spkrec-ecapa-voxceleb/classifier.ckpt

models/spkrec-ecapa-voxceleb/config.json

models/spkrec-ecapa-voxceleb/embedding_model.ckpt

models/spkrec-ecapa-voxceleb/example1.wav

models/spkrec-ecapa-voxceleb/example2.flac

models/spkrec-ecapa-voxceleb/hyperparams.yaml

models/spkrec-ecapa-voxceleb/label_encoder.txt

models/spkrec-ecapa-voxceleb/mean_var_norm_emb.ckpt

models/spkrec-ecapa-voxceleb/source.txt

models/wespeaker-ecapa-tdnn512-LM/.gitattributes

models/wespeaker-ecapa-tdnn512-LM/README.md

models/wespeaker-ecapa-tdnn512-LM/avg_model.pt

models/wespeaker-ecapa-tdnn512-LM/config.yaml

models/wespeaker-ecapa-tdnn512-LM/source.txt

models/wespeaker-ecapa-tdnn512-LM/voxceleb_ECAPA512_LM.onnx ONNX

models/wespeaker-voxceleb-ecapa-tdnn1024-LM/.gitattributes

models/wespeaker-voxceleb-ecapa-tdnn1024-LM/README.md

models/wespeaker-voxceleb-ecapa-tdnn1024-LM/avg_model.pt

models/wespeaker-voxceleb-ecapa-tdnn1024-LM/config.yaml

models/wespeaker-voxceleb-ecapa-tdnn1024-LM/source.txt

models/wespeaker-voxceleb-ecapa-tdnn1024-LM/voxceleb_ECAPA1024_LM.onnx ONNX

models/wespeaker-voxceleb-ecapa-tdnn1024/.gitattributes

models/wespeaker-voxceleb-ecapa-tdnn1024/README.md

models/wespeaker-voxceleb-ecapa-tdnn1024/avg_model.pt

models/wespeaker-voxceleb-ecapa-tdnn1024/config.yaml

models/wespeaker-voxceleb-ecapa-tdnn1024/source.txt

models/wespeaker-voxceleb-ecapa-tdnn1024/voxceleb_ECAPA1024.onnx ONNX

models/wespeaker-voxceleb-ecapa-tdnn512/.gitattributes

models/wespeaker-voxceleb-ecapa-tdnn512/README.md

models/wespeaker-voxceleb-ecapa-tdnn512/avg_model.pt

models/wespeaker-voxceleb-ecapa-tdnn512/config.yaml

models/wespeaker-voxceleb-ecapa-tdnn512/source.txt

models/wespeaker-voxceleb-ecapa-tdnn512/voxceleb_ECAPA512.onnx ONNX

Related models

More models

No task type declared.