From 40313a73648465085d388a3275486b1d7e9fc8f4 Mon Sep 17 00:00:00 2001
From: Matt Curfman <matt@tecnovelty.com>
Date: Fri, 16 Aug 2024 22:41:15 -0700
Subject: [PATCH] Revert to ipex-llm version of ollama for gpu acceleration

---
 Dockerfile | 16 +++++++++-------
 README.md  |  7 +++++--
 _init.sh   |  1 +
 _run.sh    |  2 ++
 4 files changed, 17 insertions(+), 9 deletions(-)
 create mode 100644 _init.sh
 create mode 100644 _run.sh

diff --git a/Dockerfile b/Dockerfile
index 898177e..0cf23a6 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -10,6 +10,8 @@ RUN apt update && \
     gnupg \
     wget \
     curl \
+    python3 \
+    python3-pip \
     ocl-icd-libopencl1
 
 # Intel GPU compute user-space drivers
@@ -34,21 +36,21 @@ RUN wget -qO - https://apt.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-P
    tee /etc/apt/sources.list.d/oneAPI.list && \
   apt update && \
   apt install --no-install-recommends -q -y \
-  intel-oneapi-runtime-libs
+  intel-basekit
 
 # Required oneAPI environment variables
 ENV USE_XETLA=OFF
 ENV SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
 ENV SYCL_CACHE_PERSISTENT=1
 
-# Ollama
-RUN curl -fsSL https://ollama.com/install.sh | sh 
+COPY _init.sh /usr/share/lib/init_workspace.sh
+COPY _run.sh /usr/share/lib/run_workspace.sh
+
+# Ollama via ipex-llm 
+RUN pip3 install --pre --upgrade ipex-llm[cpp] 
 
 ENV OLLAMA_NUM_GPU=999
 ENV OLLAMA_HOST=0.0.0.0:11434
 
-# https://github.com/ollama/ollama/issues/1590
-ENV OLLAMA_INTEL_GPU=1
-
-ENTRYPOINT ["/usr/local/bin/ollama", "serve"]
+ENTRYPOINT ["/bin/bash", "/usr/share/lib/run_workspace.sh"]
 
diff --git a/README.md b/README.md
index 1677412..ab4bd9d 100644
--- a/README.md
+++ b/README.md
@@ -30,6 +30,9 @@ $ docker-compose -f docker-compose-wsl2.yml up
 
 Then launch your web browser to http://localhost:3000 to launch the web ui.  Create a local OpenWeb UI credential, then click the settings icon in the top right of the screen, then select 'Models', then click 'Show', then download a model like 'llama3.1:8b-instruct-q8_0' for Intel ARC A770 16GB VRAM
 
+# Known issues
+* Little effort has been made to prune the packages pulled into the Ollama docker image for Intel GPU
+
 # References
-* https://github.com/ollama/ollama/issues/1590
-* https://github.com/ollama/ollama/pull/3278
+* https://dgpu-docs.intel.com/driver/client/overview.html
+* https://ipex-llm.readthedocs.io/en/latest/doc/LLM/Quickstart/ollama_quickstart.html
diff --git a/_init.sh b/_init.sh
new file mode 100644
index 0000000..7408d56
--- /dev/null
+++ b/_init.sh
@@ -0,0 +1 @@
+source /opt/intel/oneapi/setvars.sh
diff --git a/_run.sh b/_run.sh
new file mode 100644
index 0000000..b766a5c
--- /dev/null
+++ b/_run.sh
@@ -0,0 +1,2 @@
+source /usr/share/lib/init_workspace.sh
+/usr/local/lib/python3.10/dist-packages/bigdl/cpp/libs/ollama serve