When using the completion feature in llama.cpp, I consistently encountered a 400 error. Later, while manually debugging through the plugin, I discovered that the actual error description was being ignored, with only an error code provided. This makes configuration extremely difficult (not everyone has the time or expertise to debug and identify the issue).
When using the completion feature in llama.cpp, I consistently encountered a 400 error. Later, while manually debugging through the plugin, I discovered that the actual error description was being ignored, with only an error code provided. This makes configuration extremely difficult (not everyone has the time or expertise to debug and identify the issue).