Thanh Vinh Vo
commited on
Commit
·
bacedbc
1
Parent(s):
e326b5f
update
Browse files
app.py
CHANGED
|
@@ -233,9 +233,11 @@ class BasicAgent:
|
|
| 233 |
- Writing code to solve problem.
|
| 234 |
- Browse the web to find information.
|
| 235 |
- Solving chess problems.
|
|
|
|
| 236 |
This agent follows rules below:
|
| 237 |
1. Take the question literally! Do not add any additional information or assumptions.
|
| 238 |
2. `wikipedia` Python library is provided that makes it easy to to interact with Wikipedia pages.
|
|
|
|
| 239 |
""",
|
| 240 |
verbosity_level=0,
|
| 241 |
max_steps=10,
|
|
@@ -282,7 +284,6 @@ class BasicAgent:
|
|
| 282 |
"{question}""{file}"
|
| 283 |
Please follow rules below:
|
| 284 |
1. Take the question literally! Do not add any additional information or assumptions.
|
| 285 |
-
2. `wikipedia` Python library is provided that makes it easy to to interact with Wikipedia pages.
|
| 286 |
"""
|
| 287 |
result = self.manager_agent.run(prompt)
|
| 288 |
print(f"Agent responded with: {result}")
|
|
|
|
| 233 |
- Writing code to solve problem.
|
| 234 |
- Browse the web to find information.
|
| 235 |
- Solving chess problems.
|
| 236 |
+
- Parsing Wikipedia pages.
|
| 237 |
This agent follows rules below:
|
| 238 |
1. Take the question literally! Do not add any additional information or assumptions.
|
| 239 |
2. `wikipedia` Python library is provided that makes it easy to to interact with Wikipedia pages.
|
| 240 |
+
3. `extract_table_from_html` tool is provided that makes it easy to extract tables from Wikipedia HTML pages.
|
| 241 |
""",
|
| 242 |
verbosity_level=0,
|
| 243 |
max_steps=10,
|
|
|
|
| 284 |
"{question}""{file}"
|
| 285 |
Please follow rules below:
|
| 286 |
1. Take the question literally! Do not add any additional information or assumptions.
|
|
|
|
| 287 |
"""
|
| 288 |
result = self.manager_agent.run(prompt)
|
| 289 |
print(f"Agent responded with: {result}")
|