[ci] remove flash attn installation in ete test workflow (#3908)

* update

* update

* update

Author: zhulinJulia24

.github/workflows/benchmark.yml

@@ -103,9 +103,6 @@ jobs:
           name: my-artifact-${{ github.run_id }}-py310
       - name: Install lmdeploy - dependency
         run: |
-          # manually install flash attn
-          # the install packeage from. https://github.com/Dao-AILab/flash-attention/releases
-          python3 -m pip install /root/packages/flash_attn-*.whl
           python3 -m pip install -r /nvme/qa_test_models/offline_pkg/requirements.txt
       - name: Install lmdeploy
         if: ${{github.event_name == 'schedule' || !inputs.offline_mode}}
@@ -119,7 +116,6 @@ jobs:
           python3 -m pip install -r requirements/test.txt
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           mkdir ${{env.REPORT_DIR}}/allure-results/.pytest_cache -p

.github/workflows/daily_ete_test.yml

@@ -152,18 +152,15 @@ jobs:
           cp -r ${{env.TEST_CODE_PATH}}/. .
       - name: Install lmdeploy - dependency
         run: |
-          # manually install flash attn
-          # the install packeage from. https://github.com/Dao-AILab/flash-attention/releases
-          python3 -m pip install /root/packages/flash_attn-*.whl
           python3 -m pip install -r ${{env.OFFLINE_REQUIREMENTS}}
       - name: Install lmdeploy
         run: |
           python3 -m pip uninstall lmdeploy -y && python3 -m pip install lmdeploy-*.whl --no-deps
           python3 -m pip install -r requirements/test.txt
+          python3 -m pip install transformers==4.53.1
           pip install ${{env.DEEPSEEK_VL}} --no-deps
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           rm -rf allure-results
@@ -237,9 +234,6 @@ jobs:
           cp -r ${{env.TEST_CODE_PATH}}/. .
       - name: Install lmdeploy - dependency
         run: |
-          # manually install flash attn
-          # the install packeage from. https://github.com/Dao-AILab/flash-attention/releases
-          python3 -m pip install /root/packages/flash_attn-*.whl
           python3 -m pip install -r ${{env.OFFLINE_REQUIREMENTS}}
       - name: Install lmdeploy
         run: |
@@ -249,7 +243,6 @@ jobs:
           pip install ${{env.DEEPSEEK_VL}} --no-deps
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           cp -r /root/lora .
@@ -343,9 +336,6 @@ jobs:
           cp -r ${{env.TEST_CODE_PATH}}/. .
       - name: Install lmdeploy - dependency
         run: |
-          # manually install flash attn
-          # the install packeage from. https://github.com/Dao-AILab/flash-attention/releases
-          python3 -m pip install /root/packages/flash_attn-*.whl
           python3 -m pip install -r ${{env.OFFLINE_REQUIREMENTS}}
       - name: Install lmdeploy
         run: |
@@ -354,7 +344,6 @@ jobs:
           pip install ${{env.DEEPSEEK_VL}} --no-deps
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           rm -rf allure-results
@@ -430,9 +419,6 @@ jobs:
           cp -r ${{env.TEST_CODE_PATH}}/. .
       - name: Install lmdeploy - dependency
         run: |
-          # manually install flash attn
-          # the install packeage from. https://github.com/Dao-AILab/flash-attention/releases
-          python3 -m pip install /root/packages/flash_attn-*.whl
           python3 -m pip install -r ${{env.OFFLINE_REQUIREMENTS}}
       - name: Install lmdeploy
         run: |
@@ -441,7 +427,6 @@ jobs:
           pip install ${{env.DEEPSEEK_VL}} --no-deps
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           rm -rf allure-results
@@ -490,9 +475,6 @@ jobs:
           cp -r ${{env.TEST_CODE_PATH}}/. .
       - name: Install lmdeploy - dependency
         run: |
-          # manually install flash attn
-          # the install packeage from. https://github.com/Dao-AILab/flash-attention/releases
-          python3 -m pip install /root/packages/flash_attn-*.whl
           python3 -m pip install -r ${{env.OFFLINE_REQUIREMENTS}}
       - name: Install lmdeploy
         run: |
@@ -501,7 +483,6 @@ jobs:
           pip install ${{env.DEEPSEEK_VL}} --no-deps
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           rm -rf allure-results
@@ -552,9 +533,6 @@ jobs:
           cp -r ${{env.TEST_CODE_PATH}}/. .
       - name: Install lmdeploy - dependency
         run: |
-          # manually install flash attn
-          # the install packeage from. https://github.com/Dao-AILab/flash-attention/releases
-          python3 -m pip install /root/packages/flash_attn-*.whl
           python3 -m pip install sentence_transformers==2.2.2 --no-deps
           python3 -m pip install -r ${{env.OFFLINE_REQUIREMENTS}}
       - name: Install lmdeploy
@@ -571,7 +549,6 @@ jobs:
           echo "OPENCOMPASS_DIR=$(pwd)" >> $GITHUB_ENV
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           rm -rf allure-results

.github/workflows/daily_ete_test_3090.yml

@@ -153,10 +153,10 @@ jobs:
       - name: Install lmdeploy
         run: |
           python3 -m pip install lmdeploy-*.whl --no-deps
+          python3 -m pip install transformers==4.53.1
           python3 -m pip install -r requirements/test.txt
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           rm -rf allure-results
@@ -229,7 +229,6 @@ jobs:
           python3 -m pip install -r requirements/test.txt
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           rm -rf allure-results
@@ -294,7 +293,6 @@ jobs:
           python3 -m pip install -r requirements/test.txt
       - name: Check env
         run: |
-          pip uninstall -y nvidia-nccl-cu11
           python3 -m pip list
           lmdeploy check_env
           rm -rf allure-results

.github/workflows/evaluate.yml

@@ -17,7 +17,7 @@ on:
         required: true
         description: 'Tested TurboMind models list. eg. [internlm_chat_7b,internlm_chat_7b_w8a16]'
         type: string
-        default: '[turbomind_internlm2_5_7b_chat, pytorch_internlm2_5_7b_chat, turbomind_internlm2_5_7b_chat_batch1, turbomind_internlm2_5_7b_chat_batch1_4bits, turbomind_internlm3_8b_instruct, pytorch_internlm3_8b_instruct, turbomind_internlm2_5_20b_chat, pytorch_internlm2_5_20b_chat, turbomind_qwen1_5_7b_chat, pytorch_qwen1_5_7b_chat, turbomind_llama3_8b_instruct, pytorch_llama3_8b_instruct, turbomind_llama3_1_8b_instruct, pytorch_llama3_1_8b_instruct, turbomind_qwen2_7b_instruct, pytorch_qwen2_7b_instruct, turbomind_qwen2_5_7b_instruct, pytorch_qwen2_5_7b_instruct, turbomind_qwen2_5_32b_instruct, pytorch_qwen2_5_32b_instruct, turbomind_llama2_7b_chat, pytorch_qwen1_5_moe_2_7b_chat, pytorch_gemma_2_9b_it, pytorch_gemma_2_27b_it, turbomind_internlm2_5_7b_chat_4bits, turbomind_internlm2_5_7b_chat_kvint4, turbomind_internlm2_5_7b_chat_kvint8, pytorch_internlm2_5_7b_chat_w8a8, turbomind_internlm3_8b_instruct_4bits, turbomind_internlm3_8b_instruct_kvint4, turbomind_internlm3_8b_instruct_kvint8, pytorch_internlm3_8b_instruct_w8a8, turbomind_llama3_8b_instruct_4bits, turbomind_llama3_8b_instruct_kvint4, turbomind_llama3_1_8b_instruct_4bits, turbomind_llama3_1_8b_instruct_kvint4, turbomind_llama3_1_8b_instruct_kvint8,turbomind_llama3_8b_instruct_kvint8, pytorch_llama3_1_8b_instruct_w8a8, turbomind_qwen2_7b_instruct_4bits, turbomind_qwen2_7b_instruct_kvint8, pytorch_qwen2_7b_instruct_w8a8, turbomind_qwen2_5_7b_instruct_4bits, turbomind_qwen2_5_7b_instruct_kvint8, pytorch_qwen2_5_7b_instruct_w8a8, turbomind_qwen2_5_32b_instruct_4bits, turbomind_qwen2_5_32b_instruct_kvint8, pytorch_qwen2_5_32b_instruct_w8a8,turbomind_llama2_7b_chat_4bits, turbomind_llama2_7b_chat_kvint4, turbomind_llama2_7b_chat_kvint8]'
+        default: '[turbomind_internlm2_5_7b_chat, pytorch_internlm2_5_7b_chat, turbomind_internlm2_5_7b_chat_batch1, turbomind_internlm2_5_7b_chat_batch1_4bits, turbomind_internlm3_8b_instruct, pytorch_internlm3_8b_instruct, turbomind_internlm2_5_20b_chat, pytorch_internlm2_5_20b_chat, turbomind_qwen1_5_7b_chat, pytorch_qwen1_5_7b_chat, turbomind_llama3_8b_instruct, pytorch_llama3_8b_instruct, turbomind_llama3_1_8b_instruct, pytorch_llama3_1_8b_instruct, turbomind_qwen2_7b_instruct, pytorch_qwen2_7b_instruct, turbomind_qwen2_5_7b_instruct, pytorch_qwen2_5_7b_instruct, turbomind_qwen2_5_32b_instruct, pytorch_qwen2_5_32b_instruct, turbomind_llama2_7b_chat, pytorch_qwen1_5_moe_2_7b_chat, pytorch_gemma_2_9b_it, pytorch_gemma_2_27b_it, turbomind_internlm2_5_7b_chat_4bits, turbomind_internlm2_5_7b_chat_kvint4, turbomind_internlm2_5_7b_chat_kvint8, pytorch_internlm2_5_7b_chat_w8a8, turbomind_internlm3_8b_instruct_4bits, turbomind_internlm3_8b_instruct_kvint4, turbomind_internlm3_8b_instruct_kvint8, pytorch_internlm3_8b_instruct_w8a8, turbomind_llama3_8b_instruct_4bits, turbomind_llama3_8b_instruct_kvint4, turbomind_llama3_1_8b_instruct_4bits, turbomind_llama3_1_8b_instruct_kvint4, turbomind_llama3_1_8b_instruct_kvint8,turbomind_llama3_8b_instruct_kvint8, pytorch_llama3_1_8b_instruct_w8a8, turbomind_qwen2_7b_instruct_kvint8, turbomind_qwen2_5_7b_instruct_4bits, turbomind_qwen2_5_7b_instruct_kvint8, pytorch_qwen2_5_7b_instruct_w8a8, turbomind_qwen2_5_32b_instruct_4bits, turbomind_qwen2_5_32b_instruct_kvint8,turbomind_llama2_7b_chat_4bits, turbomind_llama2_7b_chat_kvint4, turbomind_llama2_7b_chat_kvint8]'
       chat_datasets:
         required: true
         description: 'Tested datasets list. eg. [*bbh_datasets,*ceval_datasets,*cmmlu_datasets,*GaokaoBench_datasets,*gpqa_datasets,*gsm8k_datasets,*hellaswag_datasets,*humaneval_datasets,*ifeval_datasets,*math_datasets,*sanitized_mbpp_datasets,*mmlu_datasets,*nq_datasets,*race_datasets,*TheoremQA_datasets,*triviaqa_datasets,*winogrande_datasets,*crowspairs_datasets]'
@@ -132,8 +132,6 @@ jobs:
           name: my-artifact-${{ github.run_id }}-py310
       - name: Install lmdeploy - dependency
         run: |
-          # manually install flash attn
-          # the install packeage from. https://github.com/Dao-AILab/flash-attention/releases
           python3 -m pip install -r /root/models/offline_pkg/requirements.txt
       - name: Install lmdeploy
         if: ${{github.event_name == 'schedule' || !inputs.offline_mode}}