LLM: add more models to the arc perf test (#9297)

* LLM: add more models to the arc perf test

* remove some old models

* install some dependencies

.github/workflows/llm_performance_tests.yml

@@ -98,12 +98,16 @@ jobs:
 
       - name: Install dependencies
         shell: bash
+        # pip install transformers_stream_generator for model internlm-chat-7b-8k
+        # pip install tiktoken for model Qwen-7B-Chat-10-12
         run: |
           python -m pip install --upgrade pip
           python -m pip install --upgrade wheel
           python -m pip install --upgrade omegaconf
           python -m pip install --upgrade pandas
           python -m pip install --upgrade einops
+          python -m pip install --upgrade transformers_stream_generator
+          python -m pip install --upgrade tiktoken
 
       - name: Download llm binary
         uses: ./.github/actions/llm/download-llm-binary

python/llm/test/benchmark/arc-perf-test.yaml

@@ -1,9 +1,16 @@
 repo_id:
-  - 'THUDM/chatglm2-6b'
   - 'meta-llama/Llama-2-7b-chat-hf'
   - 'meta-llama/Llama-2-13b-chat-hf'
+  - 'THUDM/chatglm2-6b'
   - 'tiiuae/falcon-7b-instruct-with-patch'
   - 'mosaicml/mpt-7b-chat'
+  - 'redpajama/gptneox-7b-redpajama-bf16'
+  - 'databricks/dolly-v1-6b'
+  - 'databricks/dolly-v2-7b'
+  - 'databricks/dolly-v2-12b'
+  - 'internlm/internlm-chat-7b-8k'
+  - 'Qwen/Qwen-7B-Chat-10-12'
+  - 'BAAI/AquilaChat-7B'
 local_model_hub: '/mnt/disk1/models'
 warm_up: 1
 num_trials: 3