Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Test transformers 41 #12191

Open
wants to merge 2 commits into
base: test_transformers_41
Choose a base branch
from
Open

Test transformers 41 #12191

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
aea1459
new main
MeouSker77 Aug 8, 2024
88a7daf
update
MeouSker77 Aug 8, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
206 changes: 8 additions & 198 deletions .github/workflows/llm_performance_tests.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -145,14 +145,16 @@ jobs:
python -m pip install --upgrade expecttest
bash python/llm/test/run-llm-install-tests.sh

- name: Test on xpu(transformers==4.36.2)
- name: Test on xpu(transformers==4.41.2 config)
shell: bash
run: |
date_for_test_version=$(date -d yesterday +%Y-%m-%d)
sed -i "s/date.today()/\"$date_for_test_version\"/g" python/llm/dev/benchmark/all-in-one/run.py
source /opt/intel/oneapi/setvars.sh
export USE_XETLA=OFF
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
# upgrade transformers
python -m pip install transformers==4.41.2
cp python/llm/test/benchmark/arc-perf-test.yaml python/llm/dev/benchmark/all-in-one/config.yaml
cd python/llm/dev/benchmark/all-in-one
mkdir test_batch1
Expand Down Expand Up @@ -182,14 +184,14 @@ jobs:
python run.py
mv *.csv test_batch4

- name: Test on xpu(transformers==4.37.0)
- name: Test on xpu(transformers==4.41.2 437)
shell: bash
run: |
source /opt/intel/oneapi/setvars.sh
export USE_XETLA=OFF
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
# upgrade transformers for model Qwen/Qwen1.5-7B-Chat
python -m pip install transformers==4.37.0
python -m pip install transformers==4.41.2
# batch_size 1
cp python/llm/test/benchmark/arc-perf-transformers-437.yaml python/llm/dev/benchmark/all-in-one/config.yaml
cd python/llm/dev/benchmark/all-in-one
Expand All @@ -214,14 +216,14 @@ jobs:
python run.py
mv *.csv test_batch4

- name: Test on xpu(transformers==4.40.0)
- name: Test on xpu(transformers==4.41.2 440)
shell: bash
run: |
source /opt/intel/oneapi/setvars.sh
export USE_XETLA=OFF
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
# upgrade transformers for model Qwen/Qwen1.5-MoE-A2.7B-Chat
python -m pip install transformers==4.40.0
python -m pip install transformers==4.41.2
python -m pip install trl
# batch_size 1
cp python/llm/test/benchmark/arc-perf-transformers-440.yaml python/llm/dev/benchmark/all-in-one/config.yaml
Expand Down Expand Up @@ -924,203 +926,11 @@ 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/4096-512/1024-128/g' python/llm/dev/benchmark/all-in-one/run.py
sed -i 's/2048-256/1024-128/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

Expand Down
2 changes: 1 addition & 1 deletion python/llm/src/ipex_llm/utils/benchmark_util_4_29.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2452,7 +2452,7 @@ def greedy_search(
last_token_time.append(end - st)

# stop if we exceed the maximum length
if stopping_criteria(input_ids, scores):
if stopping_criteria(input_ids, scores)[0]:
this_peer_finished = True

if this_peer_finished and not synced_gpus:
Expand Down
Loading