Merge branch 'master' into cb-by-default

.github/workflows/bandit.yml

@@ -13,5 +13,5 @@ jobs:
       - uses: actions/setup-python@v4
         with:
           python-version: 3.11
-      - run: python -m pip install bandit
+      - run: python -m pip install bandit==1.8.0
       - run: python -m bandit --recursive --configfile bandit.yml .

.github/workflows/causal_lm_cpp.yml

@@ -53,12 +53,12 @@ jobs:
           wget https://huggingface.co/smangrul/tinyllama_lora_sql/resolve/main/adapter_model.safetensors?download=true -O adapter_model.safetensors
       - run: >
           . ./ov/setupvars.sh
-          && timeout 35s ./build/samples/cpp/multinomial_causal_lm/multinomial_causal_lm ./open_llama_3b_v2/ a
+          && timeout 35s ./build/samples/cpp/text_generation/multinomial_causal_lm ./open_llama_3b_v2/ a
         env:
           PYTHONPATH: "./build"
       - run: >
           . ./ov/setupvars.sh
-          && timeout 35s ./samples/python/multinomial_causal_lm/multinomial_causal_lm.py ./open_llama_3b_v2/ b
+          && timeout 35s ./samples/python/text_generation/multinomial_causal_lm.py ./open_llama_3b_v2/ b
         env:
           PYTHONPATH: "./build"
       - run: >
@@ -78,8 +78,8 @@ jobs:
       matrix:
         executable:
           [
-            ./build/samples/cpp/beam_search_causal_lm/beam_search_causal_lm,
-            python ./samples/python/beam_search_causal_lm/beam_search_causal_lm.py,
+            ./build/samples/cpp/text_generation/beam_search_causal_lm,
+            python ./samples/python/text_generation/beam_search_causal_lm.py,
           ]
     runs-on: ubuntu-20.04
     defaults:
@@ -338,8 +338,8 @@ jobs:
           optimum-cli export openvino --trust-remote-code --weight-format fp16 --model Qwen/Qwen1.5-7B-Chat Qwen1.5-7B-Chat
       - run: >
           . ./ov/setupvars.sh
-          && timeout 50s ./build/samples/cpp/beam_search_causal_lm/beam_search_causal_lm ./Qwen1.5-7B-Chat/ "你好！"
-          | diff <(timeout 50s ./samples/python/beam_search_causal_lm/beam_search_causal_lm.py ./Qwen1.5-7B-Chat/ "你好！") -
+          && timeout 50s ./build/samples/cpp/text_generation/beam_search_causal_lm ./Qwen1.5-7B-Chat/ "你好！"
+          | diff <(timeout 50s ./samples/python/text_generation/beam_search_causal_lm.py ./Qwen1.5-7B-Chat/ "你好！") -
         env:
           PYTHONPATH: "./build"
 
@@ -373,8 +373,8 @@ jobs:
           optimum-cli export openvino --trust-remote-code --weight-format fp16 --model microsoft/phi-2 phi-2
       - run: >
           . ./ov/setupvars.sh
-          && timeout 50s ./build/samples/cpp/beam_search_causal_lm/beam_search_causal_lm ./phi-2/ 69
-          | diff <(timeout 50s ./samples/python/beam_search_causal_lm/beam_search_causal_lm.py ./phi-2/ 69) -
+          && timeout 50s ./build/samples/cpp/text_generation/beam_search_causal_lm ./phi-2/ 69
+          | diff <(timeout 50s ./samples/python/text_generation/beam_search_causal_lm.py ./phi-2/ 69) -
         env:
           PYTHONPATH: "./build"
 
@@ -408,8 +408,8 @@ jobs:
           optimum-cli export openvino --trust-remote-code --weight-format fp16 --model argilla/notus-7b-v1 notus-7b-v1
       - run: >
           . ./ov/setupvars.sh
-          && timeout 50s ./build/samples/cpp/beam_search_causal_lm/beam_search_causal_lm ./notus-7b-v1/ 69
-          | diff <(timeout 50s ./samples/python/beam_search_causal_lm/beam_search_causal_lm.py ./notus-7b-v1/ 69) -
+          && timeout 50s ./build/samples/cpp/text_generation/beam_search_causal_lm ./notus-7b-v1/ 69
+          | diff <(timeout 50s ./samples/python/text_generation/beam_search_causal_lm.py ./notus-7b-v1/ 69) -
         env:
           PYTHONPATH: "./build"
 
@@ -445,9 +445,9 @@ jobs:
       - name: run and compare
         run: |
           source ./ov/setupvars.sh
-          ./build/samples/cpp/speculative_decoding_lm/speculative_decoding_lm ./dolly-v2-7b/ ./dolly-v2-3b/ "Alan Turing was a" > predictions_speculative.txt
+          ./build/samples/cpp/text_generation/speculative_decoding_lm ./dolly-v2-7b/ ./dolly-v2-3b/ "Alan Turing was a" > predictions_speculative.txt
           ./build/samples/cpp/text_generation/greedy_causal_lm ./dolly-v2-7b/ "Alan Turing was a" > predictions_greedy.txt
-          python ./samples/python/speculative_decoding_lm/speculative_decoding_lm.py ./dolly-v2-7b/ ./dolly-v2-3b/ "Alan Turing was a" > predictions_py.txt
+          python ./samples/python/text_generation/speculative_decoding_lm.py ./dolly-v2-7b/ ./dolly-v2-3b/ "Alan Turing was a" > predictions_py.txt
           python -c "
           with open('predictions_greedy.txt', 'r') as f:
               predicted_greedy = f.readline()
@@ -502,9 +502,9 @@ jobs:
           Question: Can you please add 2 and 3
           A:' > ./prompt.txt
 
-          ./build/samples/cpp/prompt_lookup_decoding_lm/prompt_lookup_decoding_lm ./TinyLlama-1.1B-Chat-v1.0/ "$(<prompt.txt)" > predictions_prompt_lookup.txt
+          ./build/samples/cpp/text_generation/prompt_lookup_decoding_lm ./TinyLlama-1.1B-Chat-v1.0/ "$(<prompt.txt)" > predictions_prompt_lookup.txt
           ./build/samples/cpp/text_generation/greedy_causal_lm ./TinyLlama-1.1B-Chat-v1.0/ "$(<prompt.txt)" > predictions_greedy.txt
-          python ./samples/python/prompt_lookup_decoding_lm/prompt_lookup_decoding_lm.py ./TinyLlama-1.1B-Chat-v1.0/ "$(<prompt.txt)" > predictions_py.txt
+          python ./samples/python/text_generation/prompt_lookup_decoding_lm.py ./TinyLlama-1.1B-Chat-v1.0/ "$(<prompt.txt)" > predictions_py.txt
           python -c "
           with open('predictions_greedy.txt', 'r') as f:
               predicted_greedy = f.readline()
@@ -664,7 +664,7 @@ jobs:
         run: |
           source ./ov/setupvars.sh
           printf 'What is 2 + 2?\nWhat is the previous answer?\nAdd 1 to it.\nSubtract 5 from it.\nWhy is the sun yellow?\nWhat was my first question?\n' > ./input.txt
-          timeout 30s ./build/samples/cpp/chat_sample/chat_sample ./TinyLlama-1.1B-Chat-v1.0/ < input.txt > ./pred.txt
+          timeout 30s ./build/samples/cpp/text_generation/chat_sample ./TinyLlama-1.1B-Chat-v1.0/ < input.txt > ./pred.txt
           python -c "
           from transformers import AutoTokenizer, AutoModelForCausalLM
           model_id = 'TinyLlama/TinyLlama-1.1B-Chat-v1.0'
@@ -693,7 +693,7 @@ jobs:
           "
           diff pred.txt ref.txt
           echo "Chat sample cpp" passed
-          timeout 30s ./samples/python/chat_sample/chat_sample.py ./TinyLlama-1.1B-Chat-v1.0/ < input.txt > ./pred2.txt
+          timeout 30s ./samples/python/text_generation/chat_sample.py ./TinyLlama-1.1B-Chat-v1.0/ < input.txt > ./pred2.txt
           diff pred2.txt ref.txt
           echo "Chat sample python" passed
 

.github/workflows/linux.yml

@@ -374,7 +374,7 @@ jobs:
       - name: Test multinomial_causal_lm.py
         if: ${{ 'Release' == matrix.build-type }} # Python bindings can be built in Release only
         timeout-minutes: 1
-        run: ${{ env.INSTALL_DIR }}/samples/python/multinomial_causal_lm/multinomial_causal_lm.py ./TinyLlama-1.1B-Chat-v1.0/ 0
+        run: ${{ env.INSTALL_DIR }}/samples/python/text_generation/multinomial_causal_lm.py ./TinyLlama-1.1B-Chat-v1.0/ 0
         working-directory: ${{ env.MODELS_DIR }}
 
       - name: Test whisper_speech_recognition.py

.github/workflows/mac.yml

@@ -395,7 +395,7 @@ jobs:
         if: ${{ 'Release' == matrix.build-type }} # Python bindings can be built in Release only
         run: |
           source ${OV_INSTALL_DIR}/setupvars.sh
-          ${OV_INSTALL_DIR}/samples/python/multinomial_causal_lm/multinomial_causal_lm.py ./TinyLlama-1.1B-Chat-v1.0/ 0
+          ${OV_INSTALL_DIR}/samples/python/text_generation/multinomial_causal_lm.py ./TinyLlama-1.1B-Chat-v1.0/ 0
         timeout-minutes: 1
 
       - name: Test python samples (whisper_speech_recognition)

.github/workflows/windows.yml

@@ -311,10 +311,9 @@ jobs:
           python -m pip install . --verbose --find-links ${env:OV_INSTALL_DIR}/wheels
           python -m pip install ./tools/who_what_benchmark --find-links ${env:OV_INSTALL_DIR}/wheels
           
-          # will install transformers 4.46.3 version
           # transformers 4.46.3 will enable return_timestamps tests
           # this check enabled for windows only. Ticket: 160205.
-          python -m pip install git+https://github.com/huggingface/optimum-intel.git@753f84db6e0966580eb9eaa74a808213be730631
+          python -m pip install transformers==4.46.3
           
           python -m pytest -v ./tests/python_tests/test_whisper_pipeline.py -k "not test_smoke"
 
@@ -470,7 +469,7 @@ jobs:
         if: ${{ 'Release' == matrix.build-type }} # Python bindings can be built in Release only
         run: |
           . "${{ env.OV_INSTALL_DIR }}/setupvars.ps1"
-          python ${{ env.OV_INSTALL_DIR }}\samples\python\multinomial_causal_lm\multinomial_causal_lm.py TinyLlama-1.1B-Chat-v1.0 0
+          python ${{ env.OV_INSTALL_DIR }}\samples\python\text_generation\multinomial_causal_lm.py TinyLlama-1.1B-Chat-v1.0 0
 
       - name: Test python samples (whisper_speech_recognition)
         if: ${{ 'Release' == matrix.build-type }} # Python bindings can be built in Release only

SUPPORTED_MODELS.md

@@ -323,6 +323,10 @@ In addition to image generation models, `InpaintingPipeline` supports specialize
           <li><a href="https://huggingface.co/OpenGVLab/InternVL2-2B"><code>OpenGVLab/InternVL2-2B</code></a></li>
           <li><a href="https://huggingface.co/OpenGVLab/InternVL2-4B"><code>OpenGVLab/InternVL2-4B</code></a></li>
           <li><a href="https://huggingface.co/OpenGVLab/InternVL2-8B"><code>OpenGVLab/InternVL2-8B</code></a></li>
+          <li><a href="https://huggingface.co/OpenGVLab/InternVL2_5-1B"><code>OpenGVLab/InternVL2_5-1B</code></a></li>
+          <li><a href="https://huggingface.co/OpenGVLab/InternVL2_5-2B"><code>OpenGVLab/InternVL2_5-2B</code></a></li>
+          <li><a href="https://huggingface.co/OpenGVLab/InternVL2_5-4B"><code>OpenGVLab/InternVL2_5-4B</code></a></li>
+          <li><a href="https://huggingface.co/OpenGVLab/InternVL2_5-8B"><code>OpenGVLab/InternVL2_5-8B</code></a></li>
         </ul>
       </td>
     </tr>

samples/CMakeLists.txt

@@ -2,14 +2,7 @@
 # SPDX-License-Identifier: Apache-2.0
 #
 
-add_subdirectory(cpp/beam_search_causal_lm)
-add_subdirectory(cpp/benchmark_genai)
-add_subdirectory(cpp/chat_sample)
 add_subdirectory(cpp/text_generation)
-add_subdirectory(cpp/lora_greedy_causal_lm)
-add_subdirectory(cpp/multinomial_causal_lm)
-add_subdirectory(cpp/prompt_lookup_decoding_lm)
-add_subdirectory(cpp/speculative_decoding_lm)
 add_subdirectory(cpp/image_generation)
 add_subdirectory(cpp/visual_language_chat)
 add_subdirectory(cpp/whisper_speech_recognition)
@@ -22,27 +15,15 @@ install(FILES
     COMPONENT cpp_samples_genai)
 
 install(DIRECTORY
-            cpp/beam_search_causal_lm
-            cpp/benchmark_genai
-            cpp/chat_sample
             cpp/text_generation
             cpp/image_generation
-            cpp/lora_greedy_causal_lm
-            cpp/multinomial_causal_lm
-            # Don't install prompt_lookup_decoding_lm because it doesn't use openvino_genai library and is not verified yet.
-            cpp/speculative_decoding_lm
             cpp/visual_language_chat
             cpp/whisper_speech_recognition
         DESTINATION samples/cpp COMPONENT cpp_samples_genai)
 
 install(DIRECTORY
-            python/beam_search_causal_lm
-            python/benchmark_genai
-            python/chat_sample
             python/text_generation
             python/image_generation
-            python/multinomial_causal_lm
-            python/speculative_decoding_lm
             python/visual_language_chat
             python/whisper_speech_recognition
         DESTINATION samples/python COMPONENT cpp_samples_genai