Function read_tensor_bytes

server/scripts/quantize_gemma_dflash_q8.py:70–74

(path: Path, header_size: int, info: dict)

        header_size = struct.unpack("<Q", f.read(8))[0]
        header_json = f.read(header_size).decode("utf-8")
        return header_size, json.loads(header_json)


def read_tensor_bytes(path: Path, header_size: int, info: dict) -> bytes:
    start, end = info["data_offsets"]
    with open(path, "rb") as f:
        f.seek(8 + header_size + start)
        return f.read(end - start)