add 3k and 4k input of nightly perf test on iGPU (#11701)

* Add 3k&4k input in workflow for iGPU (#11685)

* add 3k&4k input in workflow

* comment for test

* comment models for accelarate test

* remove OOM models

* modify typo

* change test model (#11696)

* reverse test models (#11700)

.github/workflows/llm_performance_tests.yml

@@ -924,11 +924,203 @@ jobs:
 
           call conda deactivate
 
+      # 3072-384 int4+fp16
+      - name: Prepare igpu perf test (3072-384 int4+fp16)
+        shell: bash
+        run: |
+          sed -i 's/2048-256/3072-384/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i 's/{today}_test3/{today}_test1/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i "s/path to your local model hub/$MODEL_HUB_DIR/g" python/llm/test/benchmark/igpu-perf/3072-384_int4_fp16.yaml
+      
+      - name: Test on igpu (3072-384 int4+fp16)
+        shell: cmd
+        run: |
+          call conda activate igpu-perf
+          pip install transformers==4.36.2
+
+          set SYCL_CACHE_PERSISTENT=1
+          set BIGDL_LLM_XMX_DISABLED=1
+          REM for llava
+          set TRANSFORMERS_OFFLINE=1
+
+          cd python\llm\dev\benchmark\all-in-one
+          move ..\..\..\test\benchmark\igpu-perf\3072-384_int4_fp16.yaml config.yaml
+          set PYTHONIOENCODING=utf-8
+          python run.py >> %CSV_SAVE_PATH%\3072-384_int4_fp16\log\%LOG_FILE% 2>&1
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          python ..\..\..\test\benchmark\igpu-perf\check_csv_results.py --yaml-file config.yaml --suffix test1
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+
+          call conda deactivate
+
+      - name: Prepare igpu perf test for transformers 4.37 (3072-384 int4+fp16)
+        shell: bash
+        run: |
+          sed -i 's/{today}_test1/{today}_test2/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i "s/path to your local model hub/$MODEL_HUB_DIR/g" python/llm/test/benchmark/igpu-perf/3072-384_int4_fp16_437.yaml
+
+      - name: Test on igpu for transformers 4.37 (3072-384 int4+fp16)
+        shell: cmd
+        run: |
+          call conda activate igpu-perf
+          pip install transformers==4.37.0
+
+          set SYCL_CACHE_PERSISTENT=1
+          set BIGDL_LLM_XMX_DISABLED=1
+
+          cd python\llm\dev\benchmark\all-in-one
+          move ..\..\..\test\benchmark\igpu-perf\3072-384_int4_fp16_437.yaml config.yaml
+          set PYTHONIOENCODING=utf-8
+          python run.py >> %CSV_SAVE_PATH%\3072-384_int4_fp16\log\%LOG_FILE% 2>&1
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          python ..\..\..\test\benchmark\igpu-perf\check_csv_results.py --yaml-file config.yaml --suffix test2
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+
+          call conda deactivate
+
+      - name: Prepare igpu perf test for transformers 4.38 (3072-384 int4+fp16)
+        shell: bash
+        run: |
+          sed -i 's/{today}_test2/{today}_test3/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i "s/path to your local model hub/$MODEL_HUB_DIR/g" python/llm/test/benchmark/igpu-perf/3072-384_int4_fp16_438.yaml
+
+      - name: Test on igpu for transformers 4.38 (3072-384 int4+fp16)
+        shell: cmd
+        run: |
+          call conda activate igpu-perf
+          pip install transformers==4.38.2
+
+          set SYCL_CACHE_PERSISTENT=1
+          set BIGDL_LLM_XMX_DISABLED=1
+
+          cd python\llm\dev\benchmark\all-in-one
+          move ..\..\..\test\benchmark\igpu-perf\3072-384_int4_fp16_438.yaml config.yaml
+          set PYTHONIOENCODING=utf-8
+          python run.py >> %CSV_SAVE_PATH%\3072-384_int4_fp16\log\%LOG_FILE% 2>&1
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          python ..\..\..\test\benchmark\igpu-perf\check_csv_results.py --yaml-file config.yaml --suffix test3
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+
+          call conda deactivate
+
+      - name: Concat csv and generate html (3072-384 int4+fp16)
+        shell: cmd
+        run: |
+          call conda activate html-gen
+
+          cd python\llm\dev\benchmark\all-in-one
+          python ..\..\..\test\benchmark\concat_csv.py
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          del /q *test*.csv
+          move *.csv %CSV_SAVE_PATH%\3072-384_int4_fp16\
+          cd ..\..\..\test\benchmark
+          python csv_to_html.py -f %CSV_SAVE_PATH%\3072-384_int4_fp16\
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          move %CSV_SAVE_PATH%\3072-384_int4_fp16\*.html %CSV_SAVE_PATH%
+
+          call conda deactivate
+
+      # 4096-512 int4+fp16
+      - name: Prepare igpu perf test (4096-512 int4+fp16)
+        shell: bash
+        run: |
+          sed -i 's/3072-384/4096-512/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i 's/{today}_test3/{today}_test1/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i "s/path to your local model hub/$MODEL_HUB_DIR/g" python/llm/test/benchmark/igpu-perf/4096-512_int4_fp16.yaml
+      
+      - name: Test on igpu (4096-512 int4+fp16)
+        shell: cmd
+        run: |
+          call conda activate igpu-perf
+          pip install transformers==4.36.2
+
+          set SYCL_CACHE_PERSISTENT=1
+          set BIGDL_LLM_XMX_DISABLED=1
+          REM for llava
+          set TRANSFORMERS_OFFLINE=1
+
+          cd python\llm\dev\benchmark\all-in-one
+          move ..\..\..\test\benchmark\igpu-perf\4096-512_int4_fp16.yaml config.yaml
+          set PYTHONIOENCODING=utf-8
+          python run.py >> %CSV_SAVE_PATH%\4096-512_int4_fp16\log\%LOG_FILE% 2>&1
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          python ..\..\..\test\benchmark\igpu-perf\check_csv_results.py --yaml-file config.yaml --suffix test1
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+
+          call conda deactivate
+
+      - name: Prepare igpu perf test for transformers 4.37 (4096-512 int4+fp16)
+        shell: bash
+        run: |
+          sed -i 's/{today}_test1/{today}_test2/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i "s/path to your local model hub/$MODEL_HUB_DIR/g" python/llm/test/benchmark/igpu-perf/4096-512_int4_fp16_437.yaml
+
+      - name: Test on igpu for transformers 4.37 (4096-512 int4+fp16)
+        shell: cmd
+        run: |
+          call conda activate igpu-perf
+          pip install transformers==4.37.0
+
+          set SYCL_CACHE_PERSISTENT=1
+          set BIGDL_LLM_XMX_DISABLED=1
+
+          cd python\llm\dev\benchmark\all-in-one
+          move ..\..\..\test\benchmark\igpu-perf\4096-512_int4_fp16_437.yaml config.yaml
+          set PYTHONIOENCODING=utf-8
+          python run.py >> %CSV_SAVE_PATH%\4096-512_int4_fp16\log\%LOG_FILE% 2>&1
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          python ..\..\..\test\benchmark\igpu-perf\check_csv_results.py --yaml-file config.yaml --suffix test2
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+
+          call conda deactivate
+
+      - name: Prepare igpu perf test for transformers 4.38 (4096-512 int4+fp16)
+        shell: bash
+        run: |
+          sed -i 's/{today}_test2/{today}_test3/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i "s/path to your local model hub/$MODEL_HUB_DIR/g" python/llm/test/benchmark/igpu-perf/4096-512_int4_fp16_438.yaml
+
+      - name: Test on igpu for transformers 4.38 (4096-512 int4+fp16)
+        shell: cmd
+        run: |
+          call conda activate igpu-perf
+          pip install transformers==4.38.2
+
+          set SYCL_CACHE_PERSISTENT=1
+          set BIGDL_LLM_XMX_DISABLED=1
+
+          cd python\llm\dev\benchmark\all-in-one
+          move ..\..\..\test\benchmark\igpu-perf\4096-512_int4_fp16_438.yaml config.yaml
+          set PYTHONIOENCODING=utf-8
+          python run.py >> %CSV_SAVE_PATH%\4096-512_int4_fp16\log\%LOG_FILE% 2>&1
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          python ..\..\..\test\benchmark\igpu-perf\check_csv_results.py --yaml-file config.yaml --suffix test3
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+
+          call conda deactivate
+
+      - name: Concat csv and generate html (4096-512 int4+fp16)
+        shell: cmd
+        run: |
+          call conda activate html-gen
+
+          cd python\llm\dev\benchmark\all-in-one
+          python ..\..\..\test\benchmark\concat_csv.py
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          del /q *test*.csv
+          move *.csv %CSV_SAVE_PATH%\4096-512_int4_fp16\
+          cd ..\..\..\test\benchmark
+          python csv_to_html.py -f %CSV_SAVE_PATH%\4096-512_int4_fp16\
+          if %ERRORLEVEL% neq 0 (exit /b 1)
+          move %CSV_SAVE_PATH%\4096-512_int4_fp16\*.html %CSV_SAVE_PATH%
+
+          call conda deactivate
+
       # load_low_bit 1024-128 int4+fp16 
       - name: Prepare igpu perf test (load_low_bit 1024-128 int4+fp16)
         shell: bash
         run: |
-          sed -i 's/2048-256/1024-128/g' python/llm/dev/benchmark/all-in-one/run.py
+          sed -i 's/4096-512/1024-128/g' python/llm/dev/benchmark/all-in-one/run.py
           sed -i 's/{today}_test3/{today}_test1/g' python/llm/dev/benchmark/all-in-one/run.py
           sed -i "s/path to your local model hub/$MODEL_HUB_DIR/g" python/llm/test/benchmark/igpu-perf/1024-128_int4_fp16_loadlowbit.yaml
 

python/llm/test/benchmark/igpu-perf/3072-384_int4_fp16.yaml

@@ -0,0 +1,24 @@
+repo_id:
+  - 'THUDM/chatglm3-6b'
+  - 'THUDM/glm-4-9b-chat'
+  - 'baichuan-inc/Baichuan2-7B-Chat'
+  - 'meta-llama/Llama-2-7b-chat-hf'
+  - 'meta-llama/Llama-2-13b-chat-hf'
+  - 'meta-llama/Meta-Llama-3-8B-Instruct'
+  - 'mistralai/Mistral-7B-Instruct-v0.2'
+  - 'deepseek-ai/deepseek-coder-7b-instruct-v1.5'
+  - '01-ai/Yi-6B-Chat'
+  - 'Qwen/Qwen-VL-Chat'
+  - 'openbmb/MiniCPM-1B-sft-bf16'
+  - 'openbmb/MiniCPM-2B-sft-bf16'
+local_model_hub: 'path to your local model hub'
+warm_up: 1
+num_trials: 3
+num_beams: 1 # default to greedy search
+low_bit: 'sym_int4' # default to use 'sym_int4' (i.e. symmetric int4)
+batch_size: 1 # default to 1
+in_out_pairs:
+  - '3072-384'
+test_api:
+  - "transformer_int4_fp16_gpu_win" # on Intel GPU for Windows, use fp16 for non-linear layer
+cpu_embedding: True # whether put embedding to CPU (only avaiable now for gpu win related test_api)

python/llm/test/benchmark/igpu-perf/3072-384_int4_fp16_437.yaml

@@ -0,0 +1,17 @@
+repo_id:
+  - 'Qwen/Qwen1.5-7B-Chat'
+  - 'Qwen/Qwen2-7B-Instruct'
+  - 'microsoft/Phi-3-mini-4k-instruct'
+  - 'microsoft/Phi-3-mini-128k-instruct'
+  - 'microsoft/phi-3-vision-128k-instruct'
+local_model_hub: 'path to your local model hub'
+warm_up: 1
+num_trials: 3
+num_beams: 1 # default to greedy search
+low_bit: 'sym_int4' # default to use 'sym_int4' (i.e. symmetric int4)
+batch_size: 1 # default to 1
+in_out_pairs:
+  - '3072-384'
+test_api:
+  - "transformer_int4_fp16_gpu_win" # on Intel GPU for Windows (catch GPU peak memory)
+cpu_embedding: True # whether put embedding to CPU (only avaiable now for gpu win related test_api)

python/llm/test/benchmark/igpu-perf/3072-384_int4_fp16_438.yaml

@@ -0,0 +1,14 @@
+repo_id:
+  - 'stabilityai/stablelm-zephyr-3b'
+  #- 'google/gemma-7b-it'
+local_model_hub: 'path to your local model hub'
+warm_up: 1
+num_trials: 3
+num_beams: 1 # default to greedy search
+low_bit: 'sym_int4' # default to use 'sym_int4' (i.e. symmetric int4)
+batch_size: 1 # default to 1
+in_out_pairs:
+  - '3072-384'
+test_api:
+  - "transformer_int4_fp16_gpu_win" # on Intel GPU for Windows (catch GPU peak memory)
+cpu_embedding: True # whether put embedding to CPU (only avaiable now for gpu win related test_api)

python/llm/test/benchmark/igpu-perf/4096-512_int4_fp16.yaml

@@ -0,0 +1,23 @@
+repo_id:
+  - 'THUDM/chatglm3-6b'
+  - 'THUDM/glm-4-9b-chat'
+  - 'baichuan-inc/Baichuan2-7B-Chat'
+  - 'meta-llama/Llama-2-7b-chat-hf'
+  - 'meta-llama/Llama-2-13b-chat-hf'
+  - 'meta-llama/Meta-Llama-3-8B-Instruct'
+  - 'mistralai/Mistral-7B-Instruct-v0.2'
+  - 'deepseek-ai/deepseek-coder-7b-instruct-v1.5'
+  - '01-ai/Yi-6B-Chat'
+  - 'openbmb/MiniCPM-1B-sft-bf16'
+  - 'openbmb/MiniCPM-2B-sft-bf16'
+local_model_hub: 'path to your local model hub'
+warm_up: 1
+num_trials: 3
+num_beams: 1 # default to greedy search
+low_bit: 'sym_int4' # default to use 'sym_int4' (i.e. symmetric int4)
+batch_size: 1 # default to 1
+in_out_pairs:
+  - '4096-512'
+test_api:
+  - "transformer_int4_fp16_gpu_win" # on Intel GPU for Windows, use fp16 for non-linear layer
+cpu_embedding: True # whether put embedding to CPU (only avaiable now for gpu win related test_api)

python/llm/test/benchmark/igpu-perf/4096-512_int4_fp16_437.yaml

@@ -0,0 +1,17 @@
+repo_id:
+  - 'Qwen/Qwen1.5-7B-Chat'
+  - 'Qwen/Qwen2-7B-Instruct'
+  - 'microsoft/Phi-3-mini-4k-instruct'
+  - 'microsoft/Phi-3-mini-128k-instruct'
+  - 'microsoft/phi-3-vision-128k-instruct'
+local_model_hub: 'path to your local model hub'
+warm_up: 1
+num_trials: 3
+num_beams: 1 # default to greedy search
+low_bit: 'sym_int4' # default to use 'sym_int4' (i.e. symmetric int4)
+batch_size: 1 # default to 1
+in_out_pairs:
+  - '4096-512'
+test_api:
+  - "transformer_int4_fp16_gpu_win" # on Intel GPU for Windows (catch GPU peak memory)
+cpu_embedding: True # whether put embedding to CPU (only avaiable now for gpu win related test_api)

python/llm/test/benchmark/igpu-perf/4096-512_int4_fp16_438.yaml

@@ -0,0 +1,14 @@
+repo_id:
+  - 'stabilityai/stablelm-zephyr-3b'
+  #- 'google/gemma-7b-it'
+local_model_hub: 'path to your local model hub'
+warm_up: 1
+num_trials: 3
+num_beams: 1 # default to greedy search
+low_bit: 'sym_int4' # default to use 'sym_int4' (i.e. symmetric int4)
+batch_size: 1 # default to 1
+in_out_pairs:
+  - '4096-512'
+test_api:
+  - "transformer_int4_fp16_gpu_win" # on Intel GPU for Windows (catch GPU peak memory)
+cpu_embedding: True # whether put embedding to CPU (only avaiable now for gpu win related test_api)