add inference test for llama-2-7b-chat-hf and mpt-7b-chat with hpu (#201

)

Signed-off-by: jiafu zhang <jiafu.zhang@intel.com>
Co-authored-by: chen, suyue <suyue.chen@intel.com>

.github/workflows/chatbot-inference-llama-2-7b-chat-hf-hpu.yml

@@ -0,0 +1,44 @@
+name: Chatbot inference on llama-2-7b-chat-hf with hpu
+
+on:
+  workflow_call:
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}-inf-lla-7b-hpu
+  cancel-in-progress: true
+
+jobs:
+  inference:
+    name: inference test
+    runs-on: guadi2-4
+    steps:
+      - name: Clean Up Working Directory
+        run: sudo rm -rf ~/itrex-actions-runner/_work/intel-extension-for-transformers/intel-extension-for-transformers/workflows/chatbot
+
+      - name: Checkout
+        uses: actions/checkout@v2
+
+      - name: Load environment variables
+        run: cat ~/itrex-actions-runner/.env >> $GITHUB_ENV
+
+      - name: Build Docker Image
+        run: docker build ./ --target hpu -f workflows/chatbot/inference/docker/Dockerfile -t chatbotinfer-hpu:latest && yes | docker container prune && yes | docker image prune
+
+      - name: Start Docker Container
+        run: |
+          cid=$(docker ps -q --filter "name=chatbotinfer-hpu")
+          if [[ ! -z "$cid" ]]; then docker stop $cid && docker rm $cid; fi
+          docker run -tid -v /mnt/DP_disk1/huggingface/cache/:/root/.cache/huggingface/hub -v .:/root/chatbot --name="chatbotinfer-hpu" --hostname="chatbotinfer-hpu-container" chatbotinfer-hpu:latest
+
+      - name: Run Inference Test
+        run: |
+          docker exec "chatbotinfer-hpu" bash -c "cd /root/chatbot && source activate && conda activate chatbot-demo; python workflows/chatbot/inference/generate.py --base_model_path \"meta-llama/Llama-2-7b-chat-hf\" --hf_access_token \"${{ env.HF_ACCESS_TOKEN }}\" --habana --use_hpu_graphs --instructions \"Transform the following sentence into one that shows contrast. The tree is rotten.\" "
+
+      - name: Stop Container
+        if: success() || failure()
+        run: |
+          cid=$(docker ps -q --filter "name=chatbotinfer-hpu")
+          if [[ ! -z "$cid" ]]; then docker stop $cid && docker rm $cid; fi
+
+      - name: Test Summary
+        run: echo "Inference completed successfully"

.github/workflows/chatbot-inference-mpt-7b-chat-hpu.yml

@@ -0,0 +1,41 @@
+name: Chatbot inference on mosaicml/mpt-7b-chat with hpu
+
+on:
+  workflow_call:
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}-inf-mpt-7b-hpu
+  cancel-in-progress: true
+
+jobs:
+  inference:
+    name: inference test
+    runs-on: guadi2-4
+    steps:
+      - name: Clean Up Working Directory
+        run: sudo rm -rf ~/itrex-actions-runner/_work/intel-extension-for-transformers/intel-extension-for-transformers/workflows/chatbot
+
+      - name: Checkout
+        uses: actions/checkout@v2
+
+      - name: Build Docker Image
+        run: docker build ./ --target hpu -f workflows/chatbot/inference/docker/Dockerfile -t chatbotinfer-hpu:latest && yes | docker container prune && yes | docker image prune
+
+      - name: Start Docker Container
+        run: |
+          cid=$(docker ps -q --filter "name=chatbotinfer-hpu")
+          if [[ ! -z "$cid" ]]; then docker stop $cid && docker rm $cid; fi
+          docker run -tid -v /mnt/DP_disk1/huggingface/cache/:/root/.cache/huggingface/hub -v .:/root/chatbot --name="chatbotinfer-hpu" --hostname="chatbotinfer-hpu-container" chatbotinfer-hpu:latest
+
+      - name: Run Inference Test
+        run: |
+          docker exec "chatbotinfer-hpu" bash -c "cd /root/chatbot && source activate && conda activate chatbot-demo; python workflows/chatbot/inference/generate.py --base_model_path \"mosaicml/mpt-7b-chat\" --habana --use_hpu_graphs --instructions \"Transform the following sentence into one that shows contrast. The tree is rotten.\" "
+
+      - name: Stop Container
+        if: success() || failure()
+        run: |
+          cid=$(docker ps -q --filter "name=chatbotinfer-hpu")
+          if [[ ! -z "$cid" ]]; then docker stop $cid && docker rm $cid; fi
+
+      - name: Test Summary
+        run: echo "Inference completed successfully"

.github/workflows/chatbot-test.yml

@@ -10,6 +10,8 @@ on:
       - '.github/workflows/chatbot-inference-llama-2-7b-chat-hf.yml'
       - '.github/workflows/chatbot-inference-mpt-7b-chat.yml'
       - '.github/workflows/chatbot-finetune-mpt-7b-chat.yml'
+      - '.github/workflows/chatbot-inference-llama-2-7b-chat-hf-hpu.yml'
+      - '.github/workflows/chatbot-inference-mpt-7b-chat-hpu.yml'
       - '.github/workflows/chatbot-finetune-mpt-7b-chat-hpu.yml'
       - '.github/workflows/script/chatbot/**'
       - '.github/workflows/sample_data/**'
@@ -36,6 +38,12 @@ jobs:
   call-finetune-mpt-7b-chat:
     uses: ./.github/workflows/chatbot-finetune-mpt-7b-chat.yml
 
+  call-inference-llama-2-7b-chat-hf-hpu:
+    uses: ./.github/workflows/chatbot-inference-llama-2-7b-chat-hf-hpu.yml
+
+  call-inference-mpt-7b-chat-hpu:
+    uses: ./.github/workflows/chatbot-inference-mpt-7b-chat-hpu.yml
+
   call-finetune-mpt-7b-chat-hpu:
     uses: ./.github/workflows/chatbot-finetune-mpt-7b-chat-hpu.yml