ggerganov / llama.cpp

LLM inference in C/C++
MIT License
66.45k stars 9.56k forks source link

AttributeError: /home/developer/mambaforge/envs/CodeLlama/lib/libllama.so: undefined symbol: llama_n_ctx_train #3126

Closed phalexo closed 6 months ago

phalexo commented 1 year ago

Ubuntu 20.04 LLAMA_CUBLAS=on pip install llama-cpp-python --force-reinstall --upgrade --no-cache-dir (CodeLlama) developer@ai:~$ python --version Python 3.10.12

from llama_cpp import Llama

results in

Traceback (most recent call last): File "", line 1, in File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/site-packages/llama_cpp/init.py", line 1, in from .llama_cpp import * File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/site-packages/llama_cpp/llama_cpp.py", line 532, in _lib.llama_n_ctx_train.argtypes = [llama_context_p] File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/ctypes/init.py", line 387, in getattr func = self.getitem(name) File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/ctypes/init.py", line 392, in getitem func = self._FuncPtr((name_or_ordinal, self)) AttributeError: /home/developer/mambaforge/envs/CodeLlama/lib/libllama.so: undefined symbol: llama_n_ctx_train

github-actions[bot] commented 6 months ago

This issue was closed because it has been inactive for 14 days since being marked as stale.