mirror of
https://github.com/kennethreitz/instructor.git
synced 2026-06-05 22:50:18 +00:00
Jason Liu
59 lines
1.3 KiB
Markdown
59 lines
1.3 KiB
Markdown
Often times not only do you want the base model but may also want the original response from the API. You can do this by retrieving the `raw_response`, since the `raw_response` is also a pydantic model, you can use any of the pydantic model methods on it.
|
|
|
|
```python
|
|
import instructor
|
|
|
|
from openai import OpenAI
|
|
from pydantic import BaseModel
|
|
|
|
client = instructor.patch(OpenAI())
|
|
|
|
class UserExtract(BaseModel):
|
|
name: str
|
|
age: int
|
|
|
|
user: UserExtract = client.chat.completions.create(
|
|
model="gpt-3.5-turbo",
|
|
response_model=UserExtract,
|
|
messages=[
|
|
{"role": "user", "content": "Extract jason is 25 years old"},
|
|
],
|
|
)
|
|
|
|
print(user._raw_response)
|
|
```
|
|
|
|
**Output:**
|
|
|
|
```python
|
|
{
|
|
"id": "chatcmpl-8bHUPGZc9vAXBraJlebf8ciz4AMuh",
|
|
"choices": [
|
|
{
|
|
"finish_reason": "stop",
|
|
"index": 0,
|
|
"message": {
|
|
"content": null,
|
|
"role": "assistant",
|
|
"function_call": {
|
|
"arguments": "{\n \"name\": \"Jason\",\n \"age\": 25\n}",
|
|
"name": "UserExtract"
|
|
},
|
|
"tool_calls": null
|
|
},
|
|
"logprobs": null
|
|
}
|
|
],
|
|
"created": 1703896057,
|
|
"model": "gpt-3.5-turbo-0613",
|
|
"object": "chat.completion",
|
|
"system_fingerprint": null,
|
|
"usage": {
|
|
"completion_tokens": 16,
|
|
"prompt_tokens": 73,
|
|
"total_tokens": 89
|
|
}
|
|
}
|
|
|
|
```
|