A plug-and-play SDK for de-hallucinating outputs from LLMs
Project description
Overview
HAPI is a plug-and-play SDK to dehallucinate your LLM generation in real time
- Hallucination Detection: Identifies and flags AI-generated misinformation.
- Confidence Filtering: Evaluates reliability of generated content.
- Easy Integration: Will support any open-source model soon
Supported models - more coming soon!
- Llama-2-7b-chat (meta-llama/Llama-2-7b-chat)
- Falcon40b (tiiuae/falcon-40b)
- Llama-2-7b (meta-llama/Llama-2-7b-hf)
- MPT7B (mosaicml/mpt-7b)
Installation
pip install HAPI_SDK
Quick Start
from hapi_sdk import DeHallucinationClient
client = DeHallucinationClient(model_id="meta-llama/Llama-2-7b-chat-hf", license_key="your-license-key")
prompt = "Explain quantum entanglement in simple terms."
output, flagged_sentences = client.dehallucinate1(prompt)
print("Filtered Output:", output)
print("Flagged Sentences:", flagged_sentences)
Contact with any suggestions!
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
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file hapi_sdk-0.1.9.tar.gz.
File metadata
- Download URL: hapi_sdk-0.1.9.tar.gz
- Upload date:
- Size: 15.9 MB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.5
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
c5b64cfbca502b66ebaa7d84e7710ed5362bc3be2c4db10c95bb5a5c0cfce323
|
|
| MD5 |
731efc6f3a923dfc77647182715a91c2
|
|
| BLAKE2b-256 |
66e5ddc77e84fe2234cc2b4a373ebf417c113ee15bfb125beb98a5f1aa39b231
|
File details
Details for the file hapi_sdk-0.1.9-py3-none-any.whl.
File metadata
- Download URL: hapi_sdk-0.1.9-py3-none-any.whl
- Upload date:
- Size: 67.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.5
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
147d2ee7fed78f58480cff187c189e3b2b49da2a0869fd4209d0b499a078addb
|
|
| MD5 |
e6c723f89b2ab0b97d550847716baad2
|
|
| BLAKE2b-256 |
df6292c915100fbd53e6ee875c0029f9e800319292f5e9cbf32258938383c39a
|
