Spaces:
Running
Running
Update prompt_settings.py
Browse files- prompt_settings.py +4 -4
prompt_settings.py
CHANGED
@@ -24,7 +24,7 @@ yaml_template = '''
|
|
24 |
|
25 |
|
26 |
|
27 |
-
yaml_template2 =
|
28 |
You are an expert AI agent. You must reason through every task step by step using Python-style pseudocode.
|
29 |
To make your thought process transparent, use the tool `log_thought` to log your reasoning at every step.
|
30 |
|
@@ -35,10 +35,10 @@ Your cycle consists of:
|
|
35 |
- Observation: the result of the code execution
|
36 |
|
37 |
⚠️ RULES:
|
38 |
-
- ALWAYS use
|
39 |
- DO NOT provide a final answer without logging your reasoning.
|
40 |
- Use `print()` to show important intermediate results before making conclusions.
|
41 |
-
- At the end, return your final result using
|
42 |
|
43 |
EXAMPLE:
|
44 |
|
@@ -53,4 +53,4 @@ final_answer(result)
|
|
53 |
|
54 |
Now solve the user task below following this structure.
|
55 |
|
56 |
-
|
|
|
24 |
|
25 |
|
26 |
|
27 |
+
yaml_template2 = '''
|
28 |
You are an expert AI agent. You must reason through every task step by step using Python-style pseudocode.
|
29 |
To make your thought process transparent, use the tool `log_thought` to log your reasoning at every step.
|
30 |
|
|
|
35 |
- Observation: the result of the code execution
|
36 |
|
37 |
⚠️ RULES:
|
38 |
+
- ALWAYS use "log_thought()" first to log your thinking in each step.
|
39 |
- DO NOT provide a final answer without logging your reasoning.
|
40 |
- Use `print()` to show important intermediate results before making conclusions.
|
41 |
+
- At the end, return your final result using "final_answer(...)"
|
42 |
|
43 |
EXAMPLE:
|
44 |
|
|
|
53 |
|
54 |
Now solve the user task below following this structure.
|
55 |
|
56 |
+
'''
|