aljaz/dev-gpt
1
0
Fork 0
mirror of https://github.com/aljazceru/dev-gpt.git synced 2025-12-20 15:14:20 +01:00
Code Issues Packages Projects Releases Wiki Activity

⏪ fix: self healing response parser

Browse Source
This commit is contained in:
Florian Hönicke
2023-05-27 01:47:04 +02:00
parent 699757419d
commit e72bbeea17
1 changed files with 7 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
dev_gpt/options/generate/templates_user.py
View File

@@ -52,30 +52,19 @@ template_generate_possible_packages = PromptTemplate.from_template(
'''Here is the task description of the problem you need to solve: '''Here is the task description of the problem you need to solve:
"{description}" "{description}"
1. Write down up to 3 different strategies to solve the task. For each strategy write down how it solves the core problems. 1. Write down up to 3 different strategies to solve the task. For each strategy write down how it solves the core problems.
Note that packages are preferred over external apis except if it is mentioned in the description. Note that packages are preferred over external apis except they are mentioned in the description.
2. For each strategy list up to 3 Python packages that are specifically designed or have functionalities to solve the complete core problems. 2. Write down the strategies.json which is a list of required python package combinations.
3. For each package think if it fulfills the following requirements:
a) specifically designed or have functionalities to solve the complete core problem.
b) has a stable api among different versions
c) does not have system requirements
d) can solve the task when running in a docker container
e) the implementation of the core problem using the package would obey the following rules:
''' + not_allowed_function_string + '''
When answering, just write "yes" or "no".
4. For each approach, list the required python package combinations as discibed in the following.
You must output the package combinations as json wrapped into triple backticks ``` and name it **strategies.json**. \
Note that you can also leave a list empty to indicate that one of the strategies does not require any package and can be done in plain python. Note that you can also leave a list empty to indicate that one of the strategies does not require any package and can be done in plain python.
Write the output using double asterisks and triple backticks like this: Note that if gpt_3_5_turbo or google_custom_search are mentioned in the description, then they must be part of each strategy in strategies.json.
Example:
**strategies.json** **strategies.json**
``` ```
[ [
["package1", "package2", "package3"], ["package1", "package2", "package3"],
["package4", "package5"], ["package4", "package5"],
["package6", "package7", "package8", "package9"], ["package6", "package7", "gpt_3_5_turbo"],
[], [],
["package10"] ["google_custom_search"]
] ]
```''') ```''')
@@ -100,7 +89,7 @@ When you get asked something like 'Who was having a date with <X>?', then you an
You must not answer something else - only the json. You must not answer something else - only the json.
\'\'\') \'\'\')
generated_string = gpt_3_5_turbo(prompt_string="example user prompt") # prompt_string is the only parameter generated_string = gpt_3_5_turbo("example user prompt") # prompt_string is the only parameter
``` ```
""" """