Update entrypoint.sh.template

entrypoint.sh.template

@@ -4,8 +4,10 @@
 mongod &
 
 # Start the text-generation-inference process
-pip3 install -U transformers flash-attn && pip install -U transformers flash-attn
-text-generation-launcher --model-id ${MODEL_NAME} --quantize gptq --num-shard 1 --port 8080 --trust-remote-code &
+pip3 install -U transformers && pip install -U transformers
+mkdir /data/model_gptq
+text-generation-server quantize ${MODEL_NAME} /data/model_gptq
+text-generation-launcher --model-id /data/model_gptq --quantize gptq --num-shard 1 --port 8080 &
 
 # Wait for text-generation-inference to start
 curl --retry 60 --retry-delay 10 --retry-connrefused http://127.0.0.1:8080/health