{ "format": 1, "language": "fr", "source": { "repo": "jonatasgrosman/wav2vec2-large-xlsr-53-french", "revision": "7c79e105a6525d38e1e69f640b974b4a679723cc", "license": "apache-2.0", "base_model": "facebook/wav2vec2-large-xlsr-53", "datasets": [ "common_voice" ], "files": { "pytorch_model.bin": { "sha256": "2fc4760bab7bf7d0d6c06e9b7f829a27a3e66eefee3733f53c88dbedf6607f2e", "size": 1262175703 }, "README.md": { "sha256": "caa75d02b02b3fd2cc01dbabe278710ce497535d517e05ac8ce98a564e1434ea", "size": 5543 }, "config.json": { "sha256": "ee87cfb5a4ad72d24b79eb26ab4b7c26ecc2adc1c0e0e6bc6c91e551d884ef43", "size": 1531 }, "preprocessor_config.json": { "sha256": "ca5999a45e98bb76ea87a461ba28a23ad32a5bb9f733b8e0f6546ff38b6c612d", "size": 262 }, "special_tokens_map.json": { "sha256": "bb7068de1150661a10b55f9e4b12a0e77af8bf91f5e45e1b58afaf1d0e17f675", "size": 85 }, "vocab.json": { "sha256": "e83a3c849e480757fabd9fe927b870581e926407bcc7e1b6dcc8bda1e78a8d78", "size": 560 } } }, "model": { "file": "model.int8.onnx", "precision": "int8-dynamic", "input": "input_values", "output": "logits", "sample_rate": 16000, "frame_samples": 320, "min_samples": 400, "normalize": "zero-mean-unit-variance", "blank_id": 0, "vocab_size": 59, "word_delimiter": "|", "text_case": "preserve", "unit": "char" }, "text": { "spaces_separate_words": true, "unknown_characters": "wildcard" }, "verify": { "tools": { "onnxruntime": [ "1.30.0", "1.23.2" ] }, "logit_maxabs_fp32_vs_torch": 2.86102294921875e-05, "frame_argmax_agreement_int8": 1.0, "fp32_size": 1262431039, "int8_size": 356311445, "frames": 49, "runtime": { "1.30.0": { "provider": "CPUExecutionProvider", "runs": { "0.5": { "samples": 8000, "shape": [ 1, 24, 59 ], "seconds": 0.069967 }, "60.0": { "samples": 960000, "shape": [ 1, 2999, 59 ], "seconds": 14.596255 } } }, "1.23.2": { "provider": "CPUExecutionProvider", "runs": { "0.5": { "samples": 8000, "shape": [ 1, 24, 59 ], "seconds": 0.039183 }, "60.0": { "samples": 960000, "shape": [ 1, 2999, 59 ], "seconds": 8.811797 } } } }, "word_start_delta_ms_int8_vs_fp32": { "median": 0.0, "p95": 0.0, "max": 0.0, "spans": 2, "test_audio": "espeak-ng generated sentence" }, "sanity": { "monotonic_sentence_spans": 2, "unknown_wildcard_spans": 2 }, "cost": { "audio_seconds": 30, "alignment_seconds": 3.854004, "peak_rss_bytes": 2853023744 }, "synthetic_sentence": { "voice": "fr-fr", "caption": "bonjour monde", "samples": 14702, "note": "pipeline/monotonicity check only; not an accuracy claim" } }, "files": { "model.int8.onnx": { "sha256": "efd4d759980e6351c0e6610644b733a1bc36f30840b31de29274274fb888cec7", "size": 356311445 } }, "elapsed_seconds": 41.258 }