{
  "format": 1,
  "language": "es",
  "source": {
    "repo": "jonatasgrosman/wav2vec2-large-xlsr-53-spanish",
    "revision": "96d7e9b4e4a78af515a3c6d3cee7c0826045d276",
    "license": "apache-2.0",
    "base_model": "facebook/wav2vec2-large-xlsr-53",
    "datasets": [
      "common_voice"
    ],
    "files": {
      "pytorch_model.bin": {
        "sha256": "41c110e55d2eac8c79486ad87dbe8f9527ed034fe087a6adf03c891eeba914c1",
        "size": 1262101911
      },
      "README.md": {
        "sha256": "887069eccb1eb58f6c167dbbf6662727c341956a5f5120d5ab7a4207b5fd4aa9",
        "size": 5629
      },
      "config.json": {
        "sha256": "39ffe954848d5161270eef4378a17a538bffeefc310482f409ec099f7fceaa5b",
        "size": 1531
      },
      "preprocessor_config.json": {
        "sha256": "ca5999a45e98bb76ea87a461ba28a23ad32a5bb9f733b8e0f6546ff38b6c612d",
        "size": 262
      },
      "special_tokens_map.json": {
        "sha256": "bb7068de1150661a10b55f9e4b12a0e77af8bf91f5e45e1b58afaf1d0e17f675",
        "size": 85
      },
      "vocab.json": {
        "sha256": "7aea8bcd8ba1c176f12cc96c1e8f45247548f4de3218106f7352e3d833cfafdc",
        "size": 380
      }
    }
  },
  "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": 41,
    "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": 0.00015497207641601562,
    "frame_argmax_agreement_int8": 1.0,
    "fp32_size": 1262357239,
    "int8_size": 356293089,
    "frames": 49,
    "runtime": {
      "1.30.0": {
        "provider": "CPUExecutionProvider",
        "runs": {
          "0.5": {
            "samples": 8000,
            "shape": [
              1,
              24,
              41
            ],
            "seconds": 0.065142
          },
          "60.0": {
            "samples": 960000,
            "shape": [
              1,
              2999,
              41
            ],
            "seconds": 18.251437
          }
        }
      },
      "1.23.2": {
        "provider": "CPUExecutionProvider",
        "runs": {
          "0.5": {
            "samples": 8000,
            "shape": [
              1,
              24,
              41
            ],
            "seconds": 0.03831
          },
          "60.0": {
            "samples": 960000,
            "shape": [
              1,
              2999,
              41
            ],
            "seconds": 7.42069
          }
        }
      }
    },
    "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": 5.11725,
      "peak_rss_bytes": 2851131392
    },
    "synthetic_sentence": {
      "voice": "es",
      "caption": "hola mundo",
      "samples": 13911,
      "note": "pipeline/monotonicity check only; not an accuracy claim"
    }
  },
  "files": {
    "model.int8.onnx": {
      "sha256": "6531e09eb1646c7789e90f7d606c23e533e8570c3e6983ae99c0c4a5dfaee775",
      "size": 356293089
    }
  },
  "elapsed_seconds": 37.542
}