{
  "format": 1,
  "language": "hi",
  "source": {
    "repo": "ai4bharat/indicwav2vec-hindi",
    "revision": "e746fa3e8de3f6629b9ded29d9c1c565c566f3db",
    "license": "apache-2.0",
    "base_model": "AI4Bharat IndicWav2Vec (exact upstream base revision not stated in public Hub metadata)",
    "datasets": [],
    "files": {
      "README.md": {
        "sha256": "0f44b9db7abbfc4904aac472d01b23307295ef05d548dae4ee663bac145d0d6b",
        "size": 1576
      },
      "config.json": {
        "sha256": "3d55edadee031808fede41c5a0d231e273ede8674a39f13ff53b1095a0bf6789",
        "size": 2100
      },
      "preprocessor_config.json": {
        "sha256": "8f9f560daa144e525174a9ce159ac4d168e010006d6c3e46cad02cae72a6c5ad",
        "size": 260
      },
      "pytorch_model.bin": {
        "sha256": "bd9a59ed209130ef8b59514947c5174d6bc7dabb7828ee34c1873e8d002cdba3",
        "size": 1262181719
      },
      "special_tokens_map.json": {
        "sha256": "bb7068de1150661a10b55f9e4b12a0e77af8bf91f5e45e1b58afaf1d0e17f675",
        "size": 85
      },
      "tokenizer_config.json": {
        "sha256": "4db5b134af1cd571f4c0ddf1dd81d546853f24549fce87b98655510d0df394f6",
        "size": 257
      },
      "vocab.json": {
        "sha256": "0c61b05b008701ae1662361276bb65e93433a052b523113a8811b3da6e88b767",
        "size": 741
      }
    }
  },
  "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": 68,
    "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": 6.67572021484375e-05,
    "frame_argmax_agreement_fp32_vs_torch": 1.0,
    "frame_argmax_agreement_int8": 1.0,
    "fp32_size": 1262467939,
    "int8_size": 356334100,
    "frames": 49,
    "runtime": {
      "1.30.0": {
        "provider": "CPUExecutionProvider",
        "runs": {
          "0.5": {
            "samples": 8000,
            "shape": [
              1,
              24,
              68
            ],
            "seconds": 0.055193
          },
          "60.0": {
            "samples": 960000,
            "shape": [
              1,
              2999,
              68
            ],
            "seconds": 8.266201
          }
        }
      },
      "1.23.2": {
        "provider": "CPUExecutionProvider",
        "runs": {
          "0.5": {
            "samples": 8000,
            "shape": [
              1,
              24,
              68
            ],
            "seconds": 0.05425
          },
          "60.0": {
            "samples": 960000,
            "shape": [
              1,
              2999,
              68
            ],
            "seconds": 8.864183
          }
        }
      }
    },
    "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": 4.991424,
      "peak_rss_bytes": 2877698048
    },
    "synthetic_sentence": {
      "voice": "hi",
      "caption": "\u0928\u092e\u0938\u094d\u0924\u0947 \u0926\u0941\u0928\u093f\u092f\u093e",
      "samples": 17870,
      "note": "pipeline/monotonicity check only; not an accuracy claim"
    }
  },
  "files": {
    "model.int8.onnx": {
      "sha256": "2499aa855f4a6d76ba583909f53e677fce3cb54f82104291d2afe80b41eea169",
      "size": 356334100
    }
  },
  "elapsed_seconds": 37.575
}