aljaz/dev-gpt
1
0
Fork 0
mirror of https://github.com/aljazceru/dev-gpt.git synced 2025-12-23 16:44:20 +01:00
Code Issues Packages Projects Releases Wiki Activity

👨‍💻 fix: prompt schema

Browse Source
This commit is contained in:
Florian Hönicke
2023-04-20 15:22:54 +02:00
parent 042a4164ee
commit 42cb4fc143
2 changed files with 15 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/options/generate/generator.py
View File

@@ -24,7 +24,7 @@ class Generator:
self.test_description = test_description
def extract_content_from_result(self, plain_text, file_name, match_single_block=False):
pattern = fr"^\*\*{file_name}\*\*\n```(?:\w+\n)?([\s\S]*?)```"
pattern = fr"^\*\*{file_name}\*\*\n```(?:\w+\n)?([\s\S]*?)\n```" # the \n at the end makes sure that ``` within the generated code is not matched
match = re.search(pattern, plain_text, re.MULTILINE)
if match:
return match.group(1).strip()

29
src/options/generate/templates_user.py
View File

@@ -297,20 +297,7 @@ The playground must look like it was made by a professional designer.
All the ui elements are well thought out to make them visually appealing and easy to use.
The playground contains many emojis that fit the theme of the playground and has an emoji as favicon.
The playground encourages the user to deploy their own microservice by clicking on this link: https://github.com/jina-ai/gptdeploy
The playground provides a javascript code snippet that can be used to send requests to the executor. It looks like this:
```javascript
(async () => {{
const deploymentId = '<your_deployment_id>';
const url = `https://gptdeploy-${deploymentId}.wolf.jina.ai/`;
const docArray = [{{ text: '...' }}];
const res = await fetch(url, {{ method: 'POST', headers: {{ 'Content-Type': 'application/json' }}, body: JSON.stringify(docArray) }});
console.log((await res.json())[0].text);
}})();
```
Make sure you put the right example data into the text attribute of the document.
This is an example how you can connect to the executor assuming the document (d) is already defined:
The playground uses the following code to send a request to the microservice:
```
from jina import Client, Document, DocumentArray
client = Client(host='http://localhost:8080')
@@ -318,7 +305,19 @@ response = client.post('/', inputs=DocumentArray([d])) # always use '/'
print(response[0].text) # can also be blob in case of image/audio..., this should be visualized in the streamlit app
```
Note that the response will always be in response[0].text
You must provide the complete app.py file with the exact same syntax to wrap the code.
The playground displays a code block containing the microservice specific curl code that can be used to send the request to the microservice.
Example:
with st.expander("See curl command"):
st.code(
'curl -X POST -H "Content-Type: application/json" -d \'[{{"text": "<different for each microservice>"}}]\' https://gptdeploy-<your_deployment_id>.wolf.jina.ai/'
language='bash'
)
You must provide the complete app.py file using the following syntax to wrap the code:
**app.py**
```python
...
```
The playground (app.py) must always use the host on http://localhost:8080.
The playground (app.py) must not let the user configure the host on the UI.
The playground (app.py) must not import the executor.