Merge pull request #16 from luiyen/llama2

llama2-7b-chat-hf

Author: luiyen

.github/workflows/test-action.yml

@@ -53,7 +53,7 @@ jobs:
           githubRepository: ${{ github.repository }}
           githubPullRequestNumber: ${{ github.event.pull_request.number }}
           gitCommitHash: ${{ github.event.pull_request.head.sha }}
-          repoId: "microsoft/codereviewer"
+          repoId: "meta-llama/Llama-2-7b-chat-hf"
           temperature: "0.2"
           maxNewTokens: "250"
           topK: "50"

README.md

@@ -69,7 +69,7 @@ jobs:
           githubRepository: ${{ github.repository }}
           githubPullRequestNumber: ${{ github.event.pull_request.number }}
           gitCommitHash: ${{ github.event.pull_request.head.sha }}
-          repoId: "microsoft/codereviewer"
+          repoId: "meta-llama/Llama-2-7b-chat-hf"
           temperature: "0.2"
           maxNewTokens: "250"
           topK: "50"

action.yml

@@ -23,7 +23,7 @@ inputs:
   repoId:
     description: "LLM model"
     required: true
-    default: "microsoft/codereviewer"
+    default: "meta-llama/Llama-2-7b-chat-hf"
   maxNewTokens:
     description: "The amount of new tokens to be generated, this does not include the input length it is a estimate of the size of generated text you want. Each new tokens slows down the request, so look for balance between response times and length of text generated."
     required: false

entrypoint.py

@@ -86,13 +86,14 @@ def get_review(
     )
     for chunked_diff in chunked_diff_list:
         question=chunked_diff
-        template = """Provide a concise summary of the bug found in the code, describing its characteristics,
-                      location, and potential effects on the overall functionality and performance of the application.
-                      Present the potential issues and errors first, following by the most important findings, in your summary
-                      Important: Include block of code / diff in the summary also the line number.
-                      ```
-                      {question}
-                      ```
+        template = """Provide a concise summary of the bug found in the code, describing its characteristics, 
+        location, and potential effects on the overall functionality and performance of the application.
+        Present the potential issues and errors first, following by the most important findings, in your summary
+        Important: Include block of code / diff in the summary also the line number.
+
+        Diff:
+
+        {question}
         """
 
         prompt = PromptTemplate(template=template, input_variables=["question"])
@@ -106,12 +107,12 @@ def get_review(
 
     question="\n".join(chunked_reviews)
     template = """Summarize the following file changed in a pull request submitted by a developer on GitHub,
-                  focusing on major modifications, additions, deletions, and any significant updates within the files.
-                  Do not include the file name in the summary and list the summary with bullet points.
-                  Important: Include block of code / diff in the summary also the line number.
-                  ```
-                  {question}
-                  ```
+    focusing on major modifications, additions, deletions, and any significant updates within the files.
+    Do not include the file name in the summary and list the summary with bullet points.
+    Important: Include block of code / diff in the summary also the line number.
+    
+    Diff:
+    {question}
     """
     prompt = PromptTemplate(template=template, input_variables=["question"])
     llm_chain = LLMChain(prompt=prompt, llm=llm)