Fixed endless generations in test_callback_one_string test (openvinot…

…oolkit#1189) See https://productionresultssa7.blob.core.windows.net/actions-results/1d9c7ed7-f60f-463a-8dba-2898f6f39e0a/workflow-job-run-2ad6080a-b02b-5d2d-feb5-675d9367578e/logs/job/job-logs.txt?rsct=text%2Fplain&se=2024-11-11T12%3A35%3A03Z&sig=cE1uxQm2H0TswczNt4jESUwM1%2BzN2uQBHxsm7Tqw0AI%3D&ske=2024-11-11T22%3A38%3A18Z&skoid=ca7593d4-ee42-46cd-af88-8b886a2f84eb&sks=b&skt=2024-11-11T10%3A38%3A18Z&sktid=398a6654-997b-47e9-b12b-9515b896b4de&skv=2024-08-04&sp=r&spr=https&sr=b&st=2024-11-11T12%3A24%3A58Z&sv=2024-08-04 A lot of words are printed as max_new_tokens is quite huge number, while exception is expected as streamer cannot work with batch. To fix hanging of tests in this PR openvinotoolkit#882
ilya-lavrenov · Nov 20, 2024 · 9509c9f · 9509c9f
1 parent 189ec37
commit 9509c9f
Show file tree

Hide file tree

Showing 5 changed files with 20 additions and 17 deletions.
diff --git a/.github/workflows/linux.yml b/.github/workflows/linux.yml
@@ -259,19 +259,19 @@ jobs:
         run: |
           source ${OV_INSTALL_DIR}/setupvars.sh
           python -m pip install ./thirdparty/openvino_tokenizers/[transformers] -r ./tests/python_tests/requirements.txt --find-links ${OV_INSTALL_DIR}/wheels --upgrade-strategy eager
-          python -m pytest ./tests/python_tests/test_chat_generate_api.py::test_set_chat_template
+          python -m pytest -v ./tests/python_tests/test_chat_generate_api.py::test_set_chat_template
         env:
           PYTHONPATH: "./build/:$PYTHONPATH"
 
       - name: Test bindings (wheel)
         run: |
           source ${OV_INSTALL_DIR}/setupvars.sh
           python -m pip install . --verbose --find-links ${OV_INSTALL_DIR}/wheels
-          python -m pytest ./tests/python_tests --ignore ./tests/python_tests/test_whisper_generate_api.py --ignore ./tests/python_tests/test_vlm_api.py -k "not test_set_chat_template"
+          python -m pytest -v ./tests/python_tests --ignore ./tests/python_tests/test_whisper_generate_api.py --ignore ./tests/python_tests/test_vlm_api.py -k "not test_set_chat_template"
 
       - run: >
             source ${OV_INSTALL_DIR}/setupvars.sh
-            && python -m pytest ./tests/python_tests/test_vlm_api.py
+            && python -m pytest -v ./tests/python_tests/test_vlm_api.py
 
   genai_python_lib_whisper:
     name: OpenVINO genai extension whisper tests (cmake + wheel)
@@ -350,15 +350,15 @@ jobs:
         run: |
           source ${OV_INSTALL_DIR}/setupvars.sh
           python -m pip install ./thirdparty/openvino_tokenizers/[transformers] -r ./tests/python_tests/requirements.txt --find-links ${OV_INSTALL_DIR}/wheels --upgrade-strategy eager
-          python -m pytest ./tests/python_tests/test_whisper_generate_api.py -k test_smoke
+          python -m pytest -v ./tests/python_tests/test_whisper_generate_api.py -k test_smoke
         env:
           PYTHONPATH: "./build/:$PYTHONPATH"
 
       - name: Test bindings (wheel)
         run: |
           source ${OV_INSTALL_DIR}/setupvars.sh
           python -m pip install . --verbose --find-links ${OV_INSTALL_DIR}/wheels
-          python -m pytest ./tests/python_tests/test_whisper_generate_api.py -k "not test_smoke"
+          python -m pytest -v ./tests/python_tests/test_whisper_generate_api.py -k "not test_smoke"
 
   genai_package:
     name: OpenVINO genai extension (install to OpenVINO package)

diff --git a/.github/workflows/llm_bench-python.yml b/.github/workflows/llm_bench-python.yml
@@ -99,7 +99,7 @@ jobs:
         pip install git+https://github.com/huggingface/optimum.git
         GIT_CLONE_PROTECTION_ACTIVE=false pip install ${{ env.WWB_PATH }}
         python -m pip install -U --pre openvino openvino-tokenizers openvino-genai --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly --force-reinstall
-        python -m pytest tools/who_what_benchmark/tests
+        python -m pytest -v tools/who_what_benchmark/tests
   stateful:
     runs-on: ubuntu-20.04
     steps:
@@ -121,4 +121,4 @@ jobs:
           GIT_CLONE_PROTECTION_ACTIVE=false pip install tools/who_what_benchmark/
           pip install pytest
           python -m pip install -U --pre openvino openvino-tokenizers openvino-genai --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly --force-reinstall
-          python -m pytest tools/who_what_benchmark/tests
+          python -m pytest -v tools/who_what_benchmark/tests
diff --git a/.github/workflows/mac.yml b/.github/workflows/mac.yml
@@ -225,7 +225,7 @@ jobs:
         run: |
           source ${OV_INSTALL_DIR}/setupvars.sh
           python -m pip install ./thirdparty/openvino_tokenizers/[transformers] -r ./tests/python_tests/requirements.txt --find-links ${OV_INSTALL_DIR}/wheels --upgrade-strategy eager
-          python -m pytest ./tests/python_tests/test_chat_generate_api.py::test_set_chat_template
+          python -m pytest -v ./tests/python_tests/test_chat_generate_api.py::test_set_chat_template
         env:
           PYTHONPATH: "./build/:$PYTHONPATH"
 
@@ -234,7 +234,7 @@ jobs:
           source ${OV_INSTALL_DIR}/setupvars.sh
           python -m pip install . --verbose --find-links ${OV_INSTALL_DIR}/wheels
           python -c "from openvino_genai import LLMPipeline"
-          python -m pytest ./tests/python_tests/ --ignore ./tests/python_tests/test_whisper_generate_api.py --ignore ./tests/python_tests/test_vlm_api.py -k "not test_set_chat_template"
+          python -m pytest -v ./tests/python_tests/ --ignore ./tests/python_tests/test_whisper_generate_api.py --ignore ./tests/python_tests/test_vlm_api.py -k "not test_set_chat_template"
 
   genai_python_lib_whisper:
     name: OpenVINO genai extension whisper tests (cmake + wheel)
@@ -289,7 +289,7 @@ jobs:
         run: |
           source ${OV_INSTALL_DIR}/setupvars.sh
           python -m pip install ./thirdparty/openvino_tokenizers/[transformers] -r ./tests/python_tests/requirements.txt --find-links ${OV_INSTALL_DIR}/wheels --upgrade-strategy eager
-          python -m pytest ./tests/python_tests/test_whisper_generate_api.py -k test_smoke
+          python -m pytest -v ./tests/python_tests/test_whisper_generate_api.py -k test_smoke
         env:
           PYTHONPATH: "./build/:$PYTHONPATH"
 
@@ -298,7 +298,7 @@ jobs:
           source ${OV_INSTALL_DIR}/setupvars.sh
           python -m pip install . --verbose --find-links ${OV_INSTALL_DIR}/wheels
           python -c "from openvino_genai import LLMPipeline"
-          python -m pytest ./tests/python_tests/test_whisper_generate_api.py -k "not test_smoke"
+          python -m pytest -v ./tests/python_tests/test_whisper_generate_api.py -k "not test_smoke"
 
   genai_package:
     name: OpenVINO genai extension (install to OpenVINO package)

diff --git a/.github/workflows/windows.yml b/.github/workflows/windows.yml
@@ -236,15 +236,15 @@ jobs:
         run: |
           . "${{ env.OV_INSTALL_DIR }}/setupvars.ps1"
           python -m pip install ./thirdparty/openvino_tokenizers/[transformers] -r ./tests/python_tests/requirements.txt --find-links ${env:OV_INSTALL_DIR}/wheels --upgrade-strategy eager
-          python -m pytest ./tests/python_tests/test_chat_generate_api.py::test_set_chat_template
+          python -m pytest -v ./tests/python_tests/test_chat_generate_api.py::test_set_chat_template
         env:
           PYTHONPATH: "./build/" # cmd evaluates variables in a different way. Setting PYTHONPATH before setupvars.bat instead of doing that after solves that.
 
       - name: Test bindings (wheel)
         run: |
           . "${{ env.OV_INSTALL_DIR }}/setupvars.ps1"
           python -m pip install . --verbose
-          python -m pytest ./tests/python_tests/ --ignore ./tests/python_tests/test_whisper_generate_api.py --ignore ./tests/python_tests/test_vlm_api.py -k "not test_set_chat_template"
+          python -m pytest -v ./tests/python_tests/ --ignore ./tests/python_tests/test_whisper_generate_api.py --ignore ./tests/python_tests/test_vlm_api.py -k "not test_set_chat_template"
 
   genai_python_lib_whisper:
     name: OpenVINO genai extension whisper tests (cmake + wheel)
@@ -300,15 +300,15 @@ jobs:
         run: |
           . "${{ env.OV_INSTALL_DIR }}/setupvars.ps1"
           python -m pip install ./thirdparty/openvino_tokenizers/[transformers] -r ./tests/python_tests/requirements.txt --find-links ${env:OV_INSTALL_DIR}/wheels --upgrade-strategy eager
-          python -m pytest ./tests/python_tests/test_whisper_generate_api.py -k test_smoke
+          python -m pytest -v ./tests/python_tests/test_whisper_generate_api.py -k test_smoke
         env:
           PYTHONPATH: "./build/" # cmd evaluates variables in a different way. Setting PYTHONPATH before setupvars.bat instead of doing that after solves that.
 
       - name: Test bindings (wheel)
         run: |
           . "${{ env.OV_INSTALL_DIR }}/setupvars.ps1"
           python -m pip install . --verbose
-          python -m pytest ./tests/python_tests/test_whisper_generate_api.py -k "not test_smoke"
+          python -m pytest -v ./tests/python_tests/test_whisper_generate_api.py -k "not test_smoke"
 
   genai_python_lib_vlm:
     name: OpenVINO genai VLM tests (cmake + wheel)
@@ -364,7 +364,7 @@ jobs:
         run: |
           . "${{ env.OV_INSTALL_DIR }}/setupvars.ps1"
           python -m pip install ./thirdparty/openvino_tokenizers/[transformers] -r ./tests/python_tests/requirements.txt --find-links ${env:OV_INSTALL_DIR}/wheels --upgrade-strategy eager
-          python -m pytest ./tests/python_tests/test_vlm_api.py
+          python -m pytest -v ./tests/python_tests/test_vlm_api.py
         env:
           PYTHONPATH: "./build/" # cmd evaluates variables in a different way. Setting PYTHONPATH before setupvars.bat instead of doing that after solves that.
 

diff --git a/src/cpp/src/continuous_batching_impl.cpp b/src/cpp/src/continuous_batching_impl.cpp
@@ -262,6 +262,9 @@ ContinuousBatchingPipeline::ContinuousBatchingImpl::generate(const std::vector<o
         }
     }, streamer);
 
+    OPENVINO_ASSERT(streamer_ptr == nullptr || input_ids.size() == 1 && (sampling_params[0].is_greedy_decoding() || sampling_params[0].is_multinomial()),
+        "Currently streaming is possible only with batch size=1 and only for greedy or multinomial decoding");
+
     std::vector<GenerationHandle> generations;
     for (size_t request_id = 0; request_id < input_ids.size(); ++request_id) {
         OPENVINO_ASSERT(1 == input_ids[request_id].get_shape().at(0), "Use multiple tensors to pass a batch.");
@@ -283,7 +286,7 @@ ContinuousBatchingPipeline::ContinuousBatchingImpl::generate(const std::vector<o
         m_requests.clear();
     };
 
-    bool continue_generation = true, step_throws_exception = false;
+    bool continue_generation = true;
     while (has_non_finished_requests() && continue_generation) {
         try {
             step();