Instructions to use espnet/geolid_vl107only_shared_frozen with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- ESPnet
How to use espnet/geolid_vl107only_shared_frozen with ESPnet:
unknown model type (must be text-to-speech or automatic-speech-recognition)
- Notebooks
- Google Colab
- Kaggle
Download evaluation.json from espnet/geolid_vl107only_shared_frozen: direct link, hf CLI and curl.
- Browser
- Download file 4.79 kB
-
https://huggingface.co/espnet/geolid_vl107only_shared_frozen/resolve/main/evaluation.json
- Command line
-
hf download hf://espnet/geolid_vl107only_shared_frozen/evaluation.json
-
curl -L -o evaluation.json https://huggingface.co/espnet/geolid_vl107only_shared_frozen/resolve/main/evaluation.json
4.79 kB
| { | |
| "dataset": "VoxLingua107", | |
| "split": "dev", | |
| "repo_id": "espnet/geolid_vl107only_shared_frozen", | |
| "source_hf_revision": "2f5db8ebef03999239736e04ee7235d280028d07", | |
| "implementation_revision": "301372e4da311d2a3d9574a5a43b3037ae6e67c1", | |
| "model": "ESPnetLIDModel", | |
| "model_sha256": "ae4f1040fbd461765654a9537073584578e2d2090ee5683bd06c2ad4e86f4bd0", | |
| "config_sha256": "cb14eaab395d9609478d0f6ddda80d3a69155e3556dc79a0c3a845b0534e296b", | |
| "manifest_sha256": "6ad3c8fc6c74fab2720fe186525fc3bd20c4ab7b1be35d227c5c7f5456ba6c54", | |
| "utterances": 1609, | |
| "languages": 33, | |
| "metric": "utterance-level accuracy (%)", | |
| "correct": 1517, | |
| "accuracy_percent": 94.28216283405843, | |
| "paper_accuracy_percent": 94.3, | |
| "entrypoint": "espnet2.bin.lid_inference", | |
| "device": "cuda", | |
| "dtype": "float32", | |
| "batch_size": 1, | |
| "num_workers": 8, | |
| "use_preprocessor": false, | |
| "fix_duration": false, | |
| "input_order": "wav.scp sorted by utterance ID; standard ESPnet streaming iterator", | |
| "environment": { | |
| "python": "3.12.3", | |
| "torch": "2.11.0+cu128", | |
| "torchaudio": "2.11.0+cu128", | |
| "transformers": "5.14.1", | |
| "espnet-s3prl": "2026.5.14", | |
| "soundfile": "0.14.0" | |
| }, | |
| "finished_at": "2026-10-23T01:03:28.797864+00:00", | |
| "seconds": 150.1240851636976, | |
| "predictions_file": "evaluation_predictions.jsonl", | |
| "predictions_sha256": "b93df29ac4c7c3fbafdeb5ce93a6a3aac3b56c6032d9d357bc49efb997d25cb3", | |
| "comparison": { | |
| "baseline_implementation_revision": "ca777d59eb49519e3827e40cc34997f2123d6682", | |
| "baseline_correct": 1517, | |
| "matched_predictions": 1609, | |
| "changed_predictions": 0, | |
| "baseline_predictions_sha256": "b93df29ac4c7c3fbafdeb5ce93a6a3aac3b56c6032d9d357bc49efb997d25cb3" | |
| }, | |
| "per_language": { | |
| "ara": { | |
| "total": 100, | |
| "correct": 92 | |
| }, | |
| "aze": { | |
| "total": 34, | |
| "correct": 33 | |
| }, | |
| "cmn": { | |
| "total": 23, | |
| "correct": 23 | |
| }, | |
| "dan": { | |
| "total": 99, | |
| "correct": 94 | |
| }, | |
| "deu": { | |
| "total": 82, | |
| "correct": 79 | |
| }, | |
| "ell": { | |
| "total": 5, | |
| "correct": 4 | |
| }, | |
| "eng": { | |
| "total": 80, | |
| "correct": 76 | |
| }, | |
| "est": { | |
| "total": 85, | |
| "correct": 79 | |
| }, | |
| "fas": { | |
| "total": 100, | |
| "correct": 91 | |
| }, | |
| "fin": { | |
| "total": 93, | |
| "correct": 91 | |
| }, | |
| "fra": { | |
| "total": 100, | |
| "correct": 97 | |
| }, | |
| "hrv": { | |
| "total": 4, | |
| "correct": 1 | |
| }, | |
| "hun": { | |
| "total": 4, | |
| "correct": 4 | |
| }, | |
| "hye": { | |
| "total": 100, | |
| "correct": 94 | |
| }, | |
| "isl": { | |
| "total": 26, | |
| "correct": 26 | |
| }, | |
| "ita": { | |
| "total": 49, | |
| "correct": 49 | |
| }, | |
| "jpn": { | |
| "total": 42, | |
| "correct": 42 | |
| }, | |
| "lav": { | |
| "total": 95, | |
| "correct": 95 | |
| }, | |
| "lit": { | |
| "total": 13, | |
| "correct": 11 | |
| }, | |
| "mkd": { | |
| "total": 4, | |
| "correct": 4 | |
| }, | |
| "nld": { | |
| "total": 100, | |
| "correct": 96 | |
| }, | |
| "nno": { | |
| "total": 3, | |
| "correct": 3 | |
| }, | |
| "nor": { | |
| "total": 24, | |
| "correct": 13 | |
| }, | |
| "pol": { | |
| "total": 2, | |
| "correct": 2 | |
| }, | |
| "por": { | |
| "total": 1, | |
| "correct": 1 | |
| }, | |
| "rus": { | |
| "total": 29, | |
| "correct": 28 | |
| }, | |
| "slv": { | |
| "total": 9, | |
| "correct": 8 | |
| }, | |
| "spa": { | |
| "total": 55, | |
| "correct": 54 | |
| }, | |
| "srp": { | |
| "total": 7, | |
| "correct": 4 | |
| }, | |
| "swe": { | |
| "total": 100, | |
| "correct": 98 | |
| }, | |
| "tur": { | |
| "total": 72, | |
| "correct": 66 | |
| }, | |
| "ukr": { | |
| "total": 17, | |
| "correct": 17 | |
| }, | |
| "urd": { | |
| "total": 52, | |
| "correct": 42 | |
| } | |
| }, | |
| "batch_dependence": { | |
| "s3prl_frame_alignment": "S3PRL reports ceil(samples/320) frames and repeats the final frame only at the batch boundary. Shorter items can instead expose padded-query frames from Wav2Vec2.", | |
| "ecapa_tdnn": "Temporal convolutions and unmasked SE global means depend on the padded sequence length before the masked final pooling.", | |
| "mvn_and_final_pooling_mask_padding": true, | |
| "reported_measurement_batch_size": 1 | |
| }, | |
| "resolved_config_sha256": "db997d8fd357165b6310db3b9229e6bcbdc7454c6eaf9b032c14385f46623d56", | |
| "languages_sha256": "1dbd458b4cb9148d9ab5cc83abc3fcf9c4835f02c7f2afb94589d2c7190194bc", | |
| "log_file": "evaluation.log", | |
| "log_sha256": "828c7572e528cc389cac9675cdcaec55f8b85a88e53b10c3a8edb118b9245304", | |
| "prediction_differences_file": "evaluation_prediction_differences.json", | |
| "previous_evaluation_file": "evaluation_previous.json", | |
| "source_config_url": "https://huggingface.co/espnet/geolid_vl107only_shared_frozen/blob/2f5db8ebef03999239736e04ee7235d280028d07/config.yaml", | |
| "loader": "Unmodified standard ESPnet CLI loader; strict loading was not forced" | |
| } | |