🔥 Remove our exllama code because we use auto-gptq vendored kernels (IBM#59)

We recently found that AutoGPTQ vendors its own versions of exllama and
exllamav2 kernels in
[augotgptq_extension](https://github.com/AutoGPTQ/AutoGPTQ/tree/main/autogptq_extension)
that are installed with the library. Since we install AutoGPTQ after we
installed our own builds of the exllama kernels, the AutoGPTQ ones
overwrite our copies. So it turns out that we don't need to vendor and
compile our own exllama kernels.

Signed-off-by: Travis Johnson <[email protected]>

Author: tjohnson31415

Dockerfile

@@ -253,24 +253,6 @@ COPY server/custom_kernels/ /usr/src/.
 RUN cd /usr/src && python setup.py build_ext && python setup.py install
 
 
-## Build transformers exllama kernels ##########################################
-FROM python-builder as exllama-kernels-builder
-
-WORKDIR /usr/src
-
-COPY server/exllama_kernels/ .
-RUN python setup.py build
-
-
-## Build transformers exllamav2 kernels ########################################
-FROM python-builder as exllamav2-kernels-builder
-
-WORKDIR /usr/src
-
-COPY server/exllamav2_kernels/ .
-RUN python setup.py build
-
-
 ## Flash attention v2 cached build image #######################################
 FROM base as flash-att-v2-cache
 
@@ -301,12 +283,6 @@ ENV PATH=/opt/tgis/bin:$PATH
 RUN --mount=type=bind,from=flash-att-v2-cache,src=/usr/src/flash-attention-v2,target=/usr/src/flash-attention-v2 \
     pip install /usr/src/flash-attention-v2/*.whl --no-cache-dir
 
-# Copy build artifacts from exllama kernels builder
-COPY --from=exllama-kernels-builder /usr/src/build/lib.linux-x86_64-cpython-* ${SITE_PACKAGES}
-
-# Copy build artifacts from exllamav2 kernels builder
-COPY --from=exllamav2-kernels-builder /usr/src/build/lib.linux-x86_64-cpython-* ${SITE_PACKAGES}
-
 # Copy over the auto-gptq wheel and install it
 RUN --mount=type=bind,from=auto-gptq-cache,src=/usr/src/auto-gptq-wheel,target=/usr/src/auto-gptq-wheel \
     pip install /usr/src/auto-gptq-wheel/*.whl --no-cache-dir