Llama-3-Patronus-Lynx-8B-Instruct-v1.1

Name: Llama-3-Patronus-Lynx-8B-Instruct-v1.1
Author: Hugging Face

cc-by-nc-4.08B

Lynx is an open-source hallucination evaluation model. Patronus-Lynx-8B-Instruct-v1.1 was trained on a mix of datasets including CovidQA, PubmedQA, DROP, RAGTruth. The datasets contain a mix of hand-annotated and synthetic data. The maximum sequence length is 128000 tokens.

Context

131K

Capabilities

Show unsupported (5)

Reasoning & Interaction

Tool Calling

Infrastructure

Streaming

Fine-tuning

Details

Model IDPatronusAI/Llama-3-Patronus-Lynx-8B-Instruct-v1.1

ProviderHugging Face

Creatorpatronusai

FamilyLlama-3

Licensecc-by-nc-4.0

Status—

Typechat

Input modalitiestext

Output modalitiestext

Open weightYes

Parameters8B

Architecturetransformer

Context window131K tokens

Max input—

Max output—

Reasoning tokens—

Knowledge cutoff—

Release date2024-07-24

Sourceofficial

Tools

Function Callingfunction_calling

Call external functions and APIs

Family Comparison: Llama-3

Model	Caps	Perf	Speed	Context	Max out
Llama-3-8b-Ita	TSF	—	—	8K	—
Meta-Llama-3-8B-Instruct-abliterated-v3	TSF	—	—	8K	—
Llama-3-8B-Instruct-Gradient-1048k	TSF	—	—	1.0M	—
Llama-3-8B-Instruct-RR	TS	—	—	8K	—
Meta-Llama-3-70B-Instruct	TSF	—	—	8K	8K
Meta-Llama-3-8B-Instruct	TSF	—	—	8K	8K
Meta-Llama-3-70B-Instruct	TSF	—	—	8K	—
Meta-Llama-3-8B-Instruct	TSF	—	—	8K	—
Llama-3-Patronus-Lynx-8B-Instruct-v1.1	TSF	—	—	131K	—
Meta-Llama-3-8B-Instruct-FP8-KV	TS	—	—	8K	—
Llama-3-Swallow-8B-Instruct-v0.1	TSF	—	—	8K	—
Llama-3-SauerkrautLM-8b-Instruct	TSF	—	—	8K	—
Llama-3-Taiwan-8B-Instruct	TSF	—	—	8K	—

API

GET/v1/models/huggingface/PatronusAI/Llama-3-Patronus-Lynx-8B-Instruct-v1.1