clika_runtime.io functions
encode_audio
encode_audio(*args, **kwargs)
encode_audio(samples: clika_runtime._core.Tensor, sample_rate: int) -> bytes
Encode a waveform as a 16-bit PCM WAV payload (bytes); the same accepted shapes and dtypes as save_audio, and save_audio writes exactly these bytes.
load_audio
load_audio(*args, **kwargs)
load_audio(path: str, target_sample_rate: int = 0, target_channels: int = 0) -> tuple
Decode an audio file (WAV / MP3 / FLAC / OGG / ...) to a host Float32 tensor: returns (samples, sample_rate, channels, frames); samples is [frames] for mono or [frames, channels] interleaved. A target_sample_rate / target_channels of 0 keeps the file's native value; a non-zero value resamples / remixes to it.
load_gguf
load_gguf(*args, **kwargs)
load_gguf(path: str, device: clika_runtime._core.Device = Device(CPU:-1)) -> tuple
Load a .gguf checkpoint: returns (tensors as dict[str, Tensor], the file's metadata key/value pairs as JSON text, e.g. general.architecture).
load_npy
load_npy(*args, **kwargs)
load_npy(path: str, device: clika_runtime._core.Device = Device(CPU:-1)) -> clika_runtime._core.Tensor
Load a .npy file as one tensor.
load_safetensors
load_safetensors(*args, **kwargs)
load_safetensors(path: str, device: clika_runtime._core.Device = Device(CPU:-1)) -> dict
Load a .safetensors file as dict[str, Tensor] on device.
save_audio
save_audio(*args, **kwargs)
save_audio(samples: clika_runtime._core.Tensor, sample_rate: int, path: str) -> None
Write a waveform as a 16-bit PCM WAV file at path (created / truncated). samples is [frames] (mono) or [frames, channels] interleaved: a float tensor holds values in [-1, 1] (values outside clip); an Int16 tensor is written as is. sample_rate is in Hz.
save_npy
save_npy(*args, **kwargs)
save_npy(tensor: clika_runtime._core.Tensor, path: str) -> None
Write one tensor to a .npy file.
save_safetensors
save_safetensors(*args, **kwargs)
save_safetensors(tensors: dict, path: str) -> None
Write a dict[str, Tensor] to a .safetensors file.