Wrappers around LLM API models and embeddings clients.
Project description
Langchain LLM API
A Langchain compatible implementation which enables the integration with LLM-API
The main reason for implementing this package is to be able to use Langchain with any model run locally.
Usage
You can install this as a python library using the command (until it's integrated with langchain itself)
pip install langchain-llm-api
To use this langchain implementation with the LLM-API:
from langchain_llm_api import LLMAPI, APIEmbeddings
llm = LLMAPI(
params={"temp": 0.2},
verbose=True
)
llm("What is the capital of France?")
Or with streaming:
from langchain_llm_api import LLMAPI, APIEmbeddings
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
llm = LLMAPI(
params={"temp": 0.2},
verbose=True,
streaming=True,
callback_manager=CallbackManager([StreamingStdOutCallbackHandler()])
)
llm("What is the capital of France?")
Check LLM-API for the possible models and thier params
Embeddings
to use the embeddings endpoint:
emb = APIEmbeddings(
host_name="your api host name",
params = {"n_predict": 300, "temp": 0.2, ...}
)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file langchain_llm_api-0.0.1.tar.gz
.
File metadata
- Download URL: langchain_llm_api-0.0.1.tar.gz
- Upload date:
- Size: 3.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.4.2 CPython/3.8.16 Linux/5.15.0-1035-azure
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 14fcb392805691022ab4274189515866bf1bb9fc58eb90803a468ea31f799015 |
|
MD5 | 2b1cb6bbffe02deaa010a36129002c71 |
|
BLAKE2b-256 | b4ddd45b26379aa1fba996064071a31adeb2393e50f2f74d5e82817d34659e32 |
File details
Details for the file langchain_llm_api-0.0.1-py3-none-any.whl
.
File metadata
- Download URL: langchain_llm_api-0.0.1-py3-none-any.whl
- Upload date:
- Size: 5.3 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.4.2 CPython/3.8.16 Linux/5.15.0-1035-azure
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | bb46a951251b448429db98ece1f6cf609103a11e38dd8d73c000244a0485462d |
|
MD5 | 24ea18de31f623cb9e452e177e635ffd |
|
BLAKE2b-256 | 6bcd79526ec075e366576f87d397e1cf636da88a7dc03e026e6bed21ad72d004 |