Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
|
@@ -254,7 +254,7 @@ def load_knowledge_base():
|
|
| 254 |
try:
|
| 255 |
loader = TextLoader("./data_source/time_to_rethink_trust_book.md")
|
| 256 |
documents = loader.load()
|
| 257 |
-
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=
|
| 258 |
docs = text_splitter.split_documents(documents)
|
| 259 |
return docs
|
| 260 |
except Exception as e:
|
|
@@ -361,7 +361,7 @@ prompt_message = f"""
|
|
| 361 |
* It should properly follow format of blogs , reports and articles for any organization. No need to literally follow the knowledge base here .
|
| 362 |
* Do not mention name of trust bucket [ stability, development , relationship , competence , benefit , vision] in your response .
|
| 363 |
|
| 364 |
-
If query is related to trust builders or proof points , refer to knowledge base :[knowledge_base] and follow the instructions :
|
| 365 |
* Generate a compelling copy which contains short facts and dot points only , based upon guidlines & objective of trust bucket mentioned in [{input}].
|
| 366 |
* Return around 5-6 proof points with figures , but do not elaborate them .
|
| 367 |
* Strictly Do not mention "Trust" , "Stability" , "Development" , "Relationship" , "Competence" , "benefit" , "vision" words in your response .
|
|
|
|
| 254 |
try:
|
| 255 |
loader = TextLoader("./data_source/time_to_rethink_trust_book.md")
|
| 256 |
documents = loader.load()
|
| 257 |
+
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=150)
|
| 258 |
docs = text_splitter.split_documents(documents)
|
| 259 |
return docs
|
| 260 |
except Exception as e:
|
|
|
|
| 361 |
* It should properly follow format of blogs , reports and articles for any organization. No need to literally follow the knowledge base here .
|
| 362 |
* Do not mention name of trust bucket [ stability, development , relationship , competence , benefit , vision] in your response .
|
| 363 |
|
| 364 |
+
If query is related to trust builders , trust statements or proof points , refer to knowledge base :[knowledge_base] and follow the instructions :
|
| 365 |
* Generate a compelling copy which contains short facts and dot points only , based upon guidlines & objective of trust bucket mentioned in [{input}].
|
| 366 |
* Return around 5-6 proof points with figures , but do not elaborate them .
|
| 367 |
* Strictly Do not mention "Trust" , "Stability" , "Development" , "Relationship" , "Competence" , "benefit" , "vision" words in your response .
|