Compile bug: Converting the Model to Llama.cpp GGUF

[ErdemYavuz55]

Git commit

https://github.com/ggerganov/llama.cpp/releases/tag/b4390

Operating systems

Windows

GGML backends

CPU

Problem description & steps to reproduce

https://www.datacamp.com/tutorial/llama3-fine-tuning-locally

I am trying this code in Kaggle Notebook. In this tutorial, I tried to "3. Converting the Model to Llama.cpp GGUF". But I had some issues. Could you help me about these issues ?

%cd /kaggle/working
!git clone --depth=1 https://github.com/ggerganov/llama.cpp.git
%cd /kaggle/working/llama.cpp
!sed -i 's|MK_LDFLAGS += -lcuda|MK_LDFLAGS += -L/usr/local/nvidia/lib64 -lcuda|' Makefile
!LLAMA_CUDA=1 conda run -n base make -j > /dev/null

/kaggle/working
Cloning into 'llama.cpp'...
remote: Enumerating objects: 1217, done.
remote: Counting objects: 100% (1217/1217), done.
remote: Compressing objects: 100% (944/944), done.
remote: Total 1217 (delta 260), reused 765 (delta 221), pack-reused 0 (from 0)
Receiving objects: 100% (1217/1217), 19.22 MiB | 19.39 MiB/s, done.
Resolving deltas: 100% (260/260), done.
/kaggle/working/llama.cpp
Error unknown MAMBA_EXE: "/opt/conda/bin/conda", filename must be mamba or micromamba

CondaError: Run 'conda init' before 'conda activate'

Makefile:2: *** The Makefile build is deprecated. Use the CMake build instead. For more details, see https://github.com/ggerganov/llama.cpp/blob/master/docs/build.md. Stop.

First Bad Commit

No response

Relevant log output

/kaggle/working
Cloning into 'llama.cpp'...
remote: Enumerating objects: 1217, done.
remote: Counting objects: 100% (1217/1217), done.
remote: Compressing objects: 100% (944/944), done.
remote: Total 1217 (delta 260), reused 765 (delta 221), pack-reused 0 (from 0)
Receiving objects: 100% (1217/1217), 19.22 MiB | 19.39 MiB/s, done.
Resolving deltas: 100% (260/260), done.
/kaggle/working/llama.cpp
Error unknown MAMBA_EXE: "/opt/conda/bin/conda", filename must be mamba or micromamba

CondaError: Run 'conda init' before 'conda activate'

Makefile:2: *** The Makefile build is deprecated. Use the CMake build instead. For more details, see https://github.com/ggerganov/llama.cpp/blob/master/docs/build.md.  Stop.