Spaces:
Runtime error
Runtime error
syedmudassir16
commited on
Update app.py
Browse files
app.py
CHANGED
@@ -14,7 +14,6 @@ import json
|
|
14 |
import gradio as gr
|
15 |
import re
|
16 |
from threading import Thread
|
17 |
-
from transformers.agents import Tool, HfEngine, ReactJsonAgent
|
18 |
|
19 |
class DocumentRetrievalAndGeneration:
|
20 |
def __init__(self, embedding_model_name, lm_model_id, data_folder):
|
@@ -23,7 +22,6 @@ class DocumentRetrievalAndGeneration:
|
|
23 |
self.gpu_index = self.create_faiss_index()
|
24 |
self.tokenizer, self.model = self.initialize_llm(lm_model_id)
|
25 |
self.retriever_tool = self.create_retriever_tool()
|
26 |
-
self.agent = self.create_agent()
|
27 |
|
28 |
def load_documents(self, folder_path):
|
29 |
loader = DirectoryLoader(folder_path, loader_cls=TextLoader)
|
@@ -89,22 +87,11 @@ class DocumentRetrievalAndGeneration:
|
|
89 |
return "Text generation process encountered an error"
|
90 |
|
91 |
def create_retriever_tool(self):
|
92 |
-
class RetrieverTool
|
93 |
-
|
94 |
-
description = "Retrieves documents from the knowledge base that are semantically similar to the input query."
|
95 |
-
inputs = {
|
96 |
-
"query": {
|
97 |
-
"type": "text",
|
98 |
-
"description": "The query to perform. Use affirmative form rather than a question.",
|
99 |
-
}
|
100 |
-
}
|
101 |
-
output_type = "text"
|
102 |
-
|
103 |
-
def __init__(self, parent, **kwargs):
|
104 |
-
super().__init__(**kwargs)
|
105 |
self.parent = parent
|
106 |
|
107 |
-
def
|
108 |
similarityThreshold = 1
|
109 |
query_embedding = self.parent.embeddings.encode(query, convert_to_tensor=True).cpu().numpy()
|
110 |
distances, indices = self.parent.gpu_index.search(np.array([query_embedding]), k=3)
|
@@ -117,22 +104,23 @@ class DocumentRetrievalAndGeneration:
|
|
117 |
|
118 |
return RetrieverTool(self)
|
119 |
|
120 |
-
def create_agent(self):
|
121 |
-
llm_engine = HfEngine("meta-llama/Meta-Llama-3.1-8B-Instruct")
|
122 |
-
return ReactJsonAgent(tools=[self.retriever_tool], llm_engine=llm_engine, max_iterations=4, verbose=2)
|
123 |
-
|
124 |
def run_agentic_rag(self, question: str) -> str:
|
125 |
-
|
126 |
-
|
|
|
|
|
|
|
|
|
|
|
127 |
Respond only to the question asked, be concise and relevant.
|
128 |
-
If you can't find information,
|
129 |
-
Make sure to cover the question completely by calling the retriever tool several times with semantically different queries.
|
130 |
-
Your queries should be in affirmative form, not questions.
|
131 |
|
132 |
-
Question:
|
133 |
-
|
134 |
|
135 |
-
|
|
|
|
|
136 |
|
137 |
def query_and_generate_response(self, query):
|
138 |
# Standard RAG
|
|
|
14 |
import gradio as gr
|
15 |
import re
|
16 |
from threading import Thread
|
|
|
17 |
|
18 |
class DocumentRetrievalAndGeneration:
|
19 |
def __init__(self, embedding_model_name, lm_model_id, data_folder):
|
|
|
22 |
self.gpu_index = self.create_faiss_index()
|
23 |
self.tokenizer, self.model = self.initialize_llm(lm_model_id)
|
24 |
self.retriever_tool = self.create_retriever_tool()
|
|
|
25 |
|
26 |
def load_documents(self, folder_path):
|
27 |
loader = DirectoryLoader(folder_path, loader_cls=TextLoader)
|
|
|
87 |
return "Text generation process encountered an error"
|
88 |
|
89 |
def create_retriever_tool(self):
|
90 |
+
class RetrieverTool:
|
91 |
+
def __init__(self, parent):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
92 |
self.parent = parent
|
93 |
|
94 |
+
def run(self, query: str) -> str:
|
95 |
similarityThreshold = 1
|
96 |
query_embedding = self.parent.embeddings.encode(query, convert_to_tensor=True).cpu().numpy()
|
97 |
distances, indices = self.parent.gpu_index.search(np.array([query_embedding]), k=3)
|
|
|
104 |
|
105 |
return RetrieverTool(self)
|
106 |
|
|
|
|
|
|
|
|
|
107 |
def run_agentic_rag(self, question: str) -> str:
|
108 |
+
retriever_output = self.retriever_tool.run(question)
|
109 |
+
|
110 |
+
enhanced_prompt = f"""Using the following information retrieved from the knowledge base:
|
111 |
+
|
112 |
+
{retriever_output}
|
113 |
+
|
114 |
+
Give a comprehensive answer to the question below.
|
115 |
Respond only to the question asked, be concise and relevant.
|
116 |
+
If you can't find information, say "No relevant information found."
|
|
|
|
|
117 |
|
118 |
+
Question: {question}
|
119 |
+
Answer:"""
|
120 |
|
121 |
+
input_ids = self.tokenizer.encode(enhanced_prompt, return_tensors="pt").to(self.model.device)
|
122 |
+
|
123 |
+
return self.generate_response_with_timeout(input_ids)
|
124 |
|
125 |
def query_and_generate_response(self, query):
|
126 |
# Standard RAG
|