https://github.com/jm12138/wenxinworkshop-python-sdk
一个文心千帆平台的第三方 Python SDK。A third-party Python SDK for a WenxinWorkshop.
https://github.com/jm12138/wenxinworkshop-python-sdk
api ernie-bot llm python sdk wenxin
Last synced: about 1 year ago
JSON representation
一个文心千帆平台的第三方 Python SDK。A third-party Python SDK for a WenxinWorkshop.
- Host: GitHub
- URL: https://github.com/jm12138/wenxinworkshop-python-sdk
- Owner: jm12138
- License: apache-2.0
- Created: 2023-08-07T14:43:30.000Z (almost 3 years ago)
- Default Branch: main
- Last Pushed: 2023-08-23T12:01:05.000Z (almost 3 years ago)
- Last Synced: 2025-03-30T13:01:54.233Z (over 1 year ago)
- Topics: api, ernie-bot, llm, python, sdk, wenxin
- Language: Python
- Homepage:
- Size: 40 KB
- Stars: 11
- Watchers: 1
- Forks: 5
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# WenxinWorkshop Python SDK
* A third-party Python SDK for a WenxinWorkshop.
## Quick Start
* Install wenxinworkshop SDK
```bash
$ pip install git+https://github.com/jm12138/WenxinWorkshop-Python-SDK
```
* Import wenxinworkshop SDK
```python
from wenxinworkshop import LLMAPI, EmbeddingAPI, PromptTemplateAPI
from wenxinworkshop import Message, Messages, Texts
```
* Set API key and Secret key
```python
api_key = '...'
secret_key = '...'
```
* LLM chat
```python
# create a LLM API
erniebot = LLMAPI(
api_key=api_key,
secret_key=secret_key,
url=LLMAPI.ERNIEBot
)
# create a message
message = Message(
role='user',
content='你好！'
)
# create messages
messages: Messages = [message]
# get response from LLM API
response = erniebot(
messages=messages,
temperature=None,
top_p=None,
penalty_score=None,
stream=None,
user_id=None,
chunk_size=512
)
# print response
print(response)
# get response stream from LLM API
response_stream = erniebot(
messages=messages,
temperature=None,
top_p=None,
penalty_score=None,
stream=True,
user_id=None,
chunk_size=512
)
# print response stream
for item in response_stream:
print(item, end='')
```
* Embedding
```python
# create a Embedding API
ernieembedding = EmbeddingAPI(
api_key=api_key,
secret_key=secret_key,
url=EmbeddingAPI.EmbeddingV1
)
# create texts
texts: Texts = [
'你好！',
'你好吗？',
'你是谁？'
]
# get embeddings from Embedding API
response = ernieembedding(
texts=texts,
user_id=None
)
# print embeddings
print(response)
```
* Get prompt template
```python
# create a Prompt Template API
prompttemplate = PromptTemplateAPI(
api_key=api_key,
secret_key=secret_key,
url=PromptTemplateAPI.PromptTemplate
)
# get prompt template from Prompt Template API
response = prompttemplate(
template_id=1968,
content='侏罗纪世界'
)
# print prompt template
print(response)
```