{ "format": 1, "language": "ar", "source": { "repo": "jonatasgrosman/wav2vec2-large-xlsr-53-arabic", "revision": "af46c2d8531b8dcbb5e23b952f739b372c2e5d2d", "license": "apache-2.0", "base_model": "facebook/wav2vec2-large-xlsr-53", "datasets": [ "common_voice", "arabic_speech_corpus" ], "files": { "README.md": { "sha256": "6545ef1c22f8053e9f257a3368a19642c434451f5a611f9c662976a1e0dd4586", "size": 8306 }, "config.json": { "sha256": "9283014cd08d1fbcb13f7965c4dda8ef54a988e40945ff2616e15afcd0f1fac7", "size": 1565 }, "preprocessor_config.json": { "sha256": "c403ce09975b90dff0dd8302c42d422e9de1f166cd7772df23490069893cb0cf", "size": 158 }, "pytorch_model.bin": { "sha256": "a0b26f6d9d3edfde1784aef863c192a8cc1e438a23b45910ab648531ebe1857b", "size": 1262142936 }, "special_tokens_map.json": { "sha256": "bb7068de1150661a10b55f9e4b12a0e77af8bf91f5e45e1b58afaf1d0e17f675", "size": 85 }, "vocab.json": { "sha256": "96a92ff9bd106f825c878f502fe341f9f0488ace92d3c76dc40a6b30296efedd", "size": 507 } } }, "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": 51, "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": 0.00032806396484375, "frame_argmax_agreement_fp32_vs_torch": 1.0, "frame_argmax_agreement_int8": 1.0, "fp32_size": 1262398239, "int8_size": 356303343, "frames": 49, "runtime": { "1.30.0": { "provider": "CPUExecutionProvider", "runs": { "0.5": { "samples": 8000, "shape": [ 1, 24, 51 ], "seconds": 0.065743 }, "60.0": { "samples": 960000, "shape": [ 1, 2999, 51 ], "seconds": 9.251672 } } }, "1.23.2": { "provider": "CPUExecutionProvider", "runs": { "0.5": { "samples": 8000, "shape": [ 1, 24, 51 ], "seconds": 0.047933 }, "60.0": { "samples": 960000, "shape": [ 1, 2999, 51 ], "seconds": 8.617029 } } } }, "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": 4.811386, "peak_rss_bytes": 2886012928 }, "synthetic_sentence": { "voice": "ar", "caption": "\u0645\u0631\u062d\u0628\u0627 \u0628\u0627\u0644\u0639\u0627\u0644\u0645", "samples": 26254, "note": "pipeline/monotonicity check only; not an accuracy claim" } }, "files": { "model.int8.onnx": { "sha256": "b9829586625dca46c28eaf6654c7e4d6956738eb3b128c77fcb3ef9df31e5b75", "size": 356303343 } }, "elapsed_seconds": 39.699 }