VjiUddlZddlZddlmZddlmZddlmZmZddl m Z ddl m Z dZdeeeejejee fd ed e fd Zejd d Zejeed <eded edfdZd efdZdS)N)contextmanager)Path) GeneratorUnion)Tensor)_corez This should never happen. Please report an issue following the steps in https://github.com/pytorch/torchcodec/issues/new?assignees=&labels=&projects=&template=bug-report.yml. source seek_modereturnct|trtj||St|tr"tjt||St|t jst|t jrtj||St|trtj ||St|trtj ||St|t j rtdt|dr%t|drtj||Stdt!|d)NzYsource is for reading text, likely from open(..., 'r'). Try with 'rb' for binary reading?readseekzUnknown source type: z. Supported types are str, Path, bytes, Tensor and file-like objects with read(self, size: int) -> bytes and seek(self, offset: int, whence: int) -> int methods.) isinstancestrcorecreate_from_filerio RawIOBaseBufferedReadercreate_from_file_likebytescreate_from_bytesrcreate_from_tensor TextIOBase TypeErrorhasattrtype)r r s \/root/voice-cloning/.venv/lib/python3.11/site-packages/torchcodec/decoders/_decoder_utils.pycreate_decoderrsj &#=$VY777 FD ! !=$S[[)<<< FBL ) )=Z@Q-R-R=)&)<<< FE " " =%fi888 FF # # =&vy999 FBM * * = g     =WVV%<%<= )&)<<<  ?V  ? ? ?   _CUDA_BACKENDffmpeg)defaultbackend)NNNc#K|}|dvrtd|dt|} dVt|dS#t|wxYw)aContext Manager to set the CUDA backend for :class:`~torchcodec.decoders.VideoDecoder`. This context manager allows you to specify which CUDA backend implementation to use when creating :class:`~torchcodec.decoders.VideoDecoder` instances with CUDA devices. .. note:: **We recommend trying the "beta" backend instead of the default "ffmpeg" backend!** The beta backend is faster, and will eventually become the default in future versions. It may have rough edges that we'll polish over time, but it's already quite stable and ready for adoption. Let us know what you think! Only the creation of the decoder needs to be inside the context manager, the decoding methods can be called outside of it. You still need to pass ``device="cuda"`` when creating the :class:`~torchcodec.decoders.VideoDecoder` instance. If a CUDA device isn't specified, this context manager will have no effect. See example below. This is thread-safe and async-safe. Args: backend (str): The CUDA backend to use. Can be "ffmpeg" (default) or "beta". We recommend trying "beta" as it's faster! Example: >>> with set_cuda_backend("beta"): ... decoder = VideoDecoder("video.mp4", device="cuda") ... ... # Only the decoder creation needs to be part of the context manager. ... # Decoder will now the beta CUDA implementation: ... decoder.get_frame_at(0) )r"betazInvalid CUDA backend (z,). Supported values are 'ffmpeg' and 'beta'.N)lower ValueErrorr!setreset)r$previous_states rset_cuda_backendr,?sFmmooG((( ZW Z Z Z   #&&w//N, N+++++ N++++s A))Bc4tS)N)r!getr r_get_cuda_backendr0os     r ) contextvarsr contextlibrpathlibrtypingrrtorchr torchcodecrrERROR_REPORTING_INSTRUCTIONSrrrrr ContextVarr!__annotations__r,r0r/r rr:sc %%%%%%########$$$$$$   #tR\2+