1.1.3 - Optional Timings
Sometimes the Llama.cpp server will return values within timings as optional. This would cause the model to fail to decode. This has been fixed by just making them optional, so decoding still succeeds.
Sometimes the Llama.cpp server will return values within timings as optional. This would cause the model to fail to decode. This has been fixed by just making them optional, so decoding still succeeds.