Chengjie Li

BUG: import error or execute error (NULL pointer access) for the latest prebuilt version `v0.2.81`

8

For the latest version `v0.2.81`: 1. If I install it via prebuilt channel: ``` pip install -U llama-cpp-python --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu ``` Import error happens: ``` from llama_cpp import Llama ```...

使用如下命令安装cuda版 CMAKE_ARGS="-DGGML_CUDA=ON" pip install -U chatglm-cpp Linux云服务机器，资源有限，AWS g4dn.xlarge 安装直接导致机器卡死，ssh都响应不了 v0.4.0有此现象，v0.3.x没有 请问有没有能限制资源使用的cmake args？

1. Remove some builtin old models - Baichuan baichuan-chat - Starcoder starcoderplus starchat - glaive-coder - wizardlm-v1.0 - vicuna-v1.3 vicuna-v1.5 - OpenBuddy - orca - falcon - Chatglm chatglm-2 -...

Parameters between requests must have same image size and max_sequence_length and ~~sampler name~~. - ~~TODO: error handling~~ - ~~TODO: more log (processbar?)~~ - TODO: doc - TODO: abort request