CI

openvinotoolkit · Oct 13, 2024 · 5c31966 · 5c31966
1 parent cc458e2
commit 5c31966
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 2 deletions.
diff --git a/.github/workflows/causal_lm_cpp.yml b/.github/workflows/causal_lm_cpp.yml
@@ -752,6 +752,10 @@ jobs:
         run: |
           source ./ov/setupvars.sh
           timeout 50s ./build/samples/cpp/continuous_batching_accuracy/continuous_batching_accuracy -m ./TinyLlama-1.1B-Chat-v1.0/ -n 5
+      - name: Run speculative_decoding
+        run: |
+          source ./ov/setupvars.sh
+          timeout 50s ./build/samples/cpp/continuous_batching_speculative_decoding/continuous_batching_speculative_decoding -m ./TinyLlama-1.1B-Chat-v1.0/ -a ./TinyLlama-1.1B-Chat-v1.0/ -n 5
       - name: Run throughput_benchmark
         run: |
           wget -q https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/resolve/main/ShareGPT_V3_unfiltered_cleaned_split.json
@@ -804,6 +808,11 @@ jobs:
           set PATH=.\build\openvino_genai\;%PATH%
           call .\ov\setupvars.bat
           .\build\samples\cpp\continuous_batching_accuracy\Release\continuous_batching_accuracy.exe -m .\TinyLlama-1.1B-Chat-v1.0\ -n 5
+      - name: Run speculative_decoding
+        run: |
+          set PATH=.\build\openvino_genai\;%PATH%
+          call .\ov\setupvars.bat
+          .\build\samples\cpp\continuous_speculative_decoding\Release\continuous_speculative_decoding.exe -m .\TinyLlama-1.1B-Chat-v1.0\ -a .\TinyLlama-1.1B-Chat-v1.0\ -n 5
       - name: Run throughput_benchmark
         run: |
           curl -o .\ShareGPT_V3_unfiltered_cleaned_split.json -s -L "https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/resolve/main/ShareGPT_V3_unfiltered_cleaned_split.json"
@@ -847,6 +856,10 @@ jobs:
         run: |
           source ./ov/setupvars.sh
           timeout 120s ./build/samples/cpp/continuous_batching_accuracy/continuous_batching_accuracy -m ./TinyLlama-1.1B-Chat-v1.0/ -n 5
+      - name: Run speculative_decoding
+        run: |
+          source ./ov/setupvars.sh
+          timeout 120s ./build/samples/cpp/continuous_speculative_decoding/continuous_speculative_decoding -m ./TinyLlama-1.1B-Chat-v1.0/ -a ./TinyLlama-1.1B-Chat-v1.0/ -n 5
       - name: Run throughput_benchmark
         run: |
           wget -q https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/resolve/main/ShareGPT_V3_unfiltered_cleaned_split.json

diff --git a/samples/cpp/continuous_batching_accuracy/continuous_batching_accuracy.cpp b/samples/cpp/continuous_batching_accuracy/continuous_batching_accuracy.cpp
@@ -61,8 +61,8 @@ int main(int argc, char* argv[]) try {
     };
 
     std::vector<ov::genai::GenerationConfig> sampling_params_examples {
-        // ov::genai::beam_search(),
-        // ov::genai::greedy(),
+        ov::genai::beam_search(),
+        ov::genai::greedy(),
         ov::genai::multinomial(),
     };