bupa1018 commited on
Commit
0ae54ee
·
1 Parent(s): 330be71

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +30 -31
app.py CHANGED
@@ -52,6 +52,31 @@ login(HF_TOKEN)
52
  api = HfApi()
53
 
54
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
55
  def rag_workflow(query):
56
  """
57
  RAGChain class to perform the complete RAG workflow.
@@ -98,32 +123,14 @@ def rag_workflow(query):
98
 
99
  return response
100
 
101
-
102
- def initialize():
103
- global vectorstore, chunks, llm
104
-
105
 
106
- download_gitlab_repo_to_hfspace(GITLAB_API_URL, GITLAB_PROJECT_ID, GITLAB_PROJECT_VERSION)
107
-
108
- code_texts, code_references = extract_repo_files(DATA_DIR, ['kadi_apy'], [])
109
- #doc_texts, doc_references = extract_files_and_filepath_from_dir(DATA_DIR, ['docs/source/'], [])
110
- doc_texts, doc_references = extract_repo_files(DATA_DIR, [], [])
111
-
112
- print("LEEEEEEEEEEEENGTH of code_texts: ", len(code_texts))
113
- print("LEEEEEEEEEEEENGTH of doc_files: ", len(doc_texts))
114
-
115
- code_chunks = chunk_pythoncode_and_add_metadata(code_texts, code_references)
116
- doc_chunks = chunk_text_and_add_metadata(doc_texts, doc_references, CHUNK_SIZE, CHUNK_OVERLAP)
117
-
118
- print(f"Total number of code_chunks: {len(code_chunks)}")
119
- print(f"Total number of doc_chunks: {len(doc_chunks)}")
120
 
121
- vectorstore = setup_vectorstore(doc_chunks + code_chunks, EMBEDDING_MODEL_NAME, VECTORSTORE_DIRECTORY)
122
- llm = get_groq_llm(LLM_MODEL_NAME, LLM_MODEL_TEMPERATURE, GROQ_API_KEY)
123
 
124
- from langchain_community.document_loaders import TextLoader
125
-
126
- initialize()
127
 
128
 
129
  # Gradio utils
@@ -140,14 +147,6 @@ def add_text(history, text):
140
 
141
  import gradio as gr
142
 
143
-
144
- def bot_kadi(history):
145
- user_query = history[-1][0]
146
- response = rag_workflow(user_query)
147
- history[-1] = (user_query, response)
148
-
149
- yield history
150
-
151
  def main():
152
  with gr.Blocks() as demo:
153
  gr.Markdown("## KadiAPY - AI Coding-Assistant")
 
52
  api = HfApi()
53
 
54
 
55
+ def initialize():
56
+ global vectorstore, chunks, llm
57
+
58
+
59
+ download_gitlab_repo_to_hfspace(GITLAB_API_URL, GITLAB_PROJECT_ID, GITLAB_PROJECT_VERSION)
60
+
61
+ code_texts, code_references = extract_repo_files(DATA_DIR, ['kadi_apy'], [])
62
+ #doc_texts, doc_references = extract_files_and_filepath_from_dir(DATA_DIR, ['docs/source/'], [])
63
+ doc_texts, doc_references = extract_repo_files(DATA_DIR, [], [])
64
+
65
+ print("LEEEEEEEEEEEENGTH of code_texts: ", len(code_texts))
66
+ print("LEEEEEEEEEEEENGTH of doc_files: ", len(doc_texts))
67
+
68
+ code_chunks = chunk_pythoncode_and_add_metadata(code_texts, code_references)
69
+ doc_chunks = chunk_text_and_add_metadata(doc_texts, doc_references, CHUNK_SIZE, CHUNK_OVERLAP)
70
+
71
+ print(f"Total number of code_chunks: {len(code_chunks)}")
72
+ print(f"Total number of doc_chunks: {len(doc_chunks)}")
73
+
74
+ vectorstore = setup_vectorstore(doc_chunks + code_chunks, EMBEDDING_MODEL_NAME, VECTORSTORE_DIRECTORY)
75
+ llm = get_groq_llm(LLM_MODEL_NAME, LLM_MODEL_TEMPERATURE, GROQ_API_KEY)
76
+
77
+
78
+ initialize()
79
+
80
  def rag_workflow(query):
81
  """
82
  RAGChain class to perform the complete RAG workflow.
 
123
 
124
  return response
125
 
 
 
 
 
126
 
127
+ def bot_kadi(history):
128
+ user_query = history[-1][0]
129
+ response = rag_workflow(user_query)
130
+ history[-1] = (user_query, response)
 
 
 
 
 
 
 
 
 
 
131
 
132
+ yield history
 
133
 
 
 
 
134
 
135
 
136
  # Gradio utils
 
147
 
148
  import gradio as gr
149
 
 
 
 
 
 
 
 
 
150
  def main():
151
  with gr.Blocks() as demo:
152
  gr.Markdown("## KadiAPY - AI Coding-Assistant")