What are the hardware requirements for Starling LM 7B Beta?

Starling LM 7B Beta requires approximately 14.4 GB of VRAM and supports a context window of 8K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Starling LM 7B Beta and how large is it?

Starling LM 7B Beta is developed by Nexusflow, a model with 7b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Starling LM 7B Beta perform on standard benchmarks?

Starling LM 7B Beta has the following published scores: MMLU 65.14, LMSYS ELO 1134. Compare against reference models on this page or on the LLM Explorer leaderboards.

Starling LM 7B Beta by Nexusflow — VRAM 14.4GB, 8K context

Name: Starling LM 7B Beta
Rating: 3 (1 reviews)
Author: Nexusflow

Starling LM 7B Beta is an open-source language model by Nexusflow. Features: 7b LLM, VRAM: 14.4GB, Context: 8K, License: apache-2.0, HF Score: 69.9, LLM Explorer Score: 0.28, ELO: 1134, Arc: 67.2, HellaSwag: 83.5, MMLU: 65.1, TruthfulQA: 55.5, WinoGrande: 81.3, GSM8K: 66.6.

Arxiv:1909.08593 Autotrain compatible Conversational Dataset:berkeley-nest/nectar En Endpoints compatible Mistral Region:us Reward model Rlaif Rlhf Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/Nexusflow/Starling-LM-7B-beta

Starling LM 7B Beta Benchmarks

LMSys ELO: 1134 vs 1272 (so35)^-10.8%

ARC: 67.24 vs 96.7 (so35)^-30.5%

HellaSwag: 83.47 vs 95.3 (gpt4)^-12.4%

MMLU: 65.14 vs 88.3 (so35)^-26.2%

TruthfulQA: 55.47 vs 59 (gpt4)^-6%

WinoGrande: 81.29 vs 87.5 (gpt4)^-7.1%

GSM8K: 66.64 vs 96.4 (so35)^-30.9%

LLME Score: 0.27864

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

Starling LM 7B Beta (Nexusflow/Starling-LM-7B-beta)

🌟 Advertise your project 🚀

Starling LM 7B Beta Parameters and Internals

Model Type

Language Model, finetuned with RLHF / RLAIF

Use Cases

Considerations:

Please use the provided chat template to ensure optimal performance. The model can occasionally produce verbose outputs. Setting temperature to 0 can mitigate verbosity.

Additional Notes

The model is hosted on LMSYS Chatbot Arena for free testing.

Supported Languages

en (Proficient)

Training Details

Data Sources:

berkeley-nest/Nectar

Methodology:

Trained by Reinforcement Learning from AI Feedback (RLAIF) with policy optimization method PPO.

Input Output

Input Format:

Chat templates following Openchat-3.5-0106.

Accepted Modalities:

text

Output Format:

Text output based on input templates.

Performance Tips:

Follow the exact chat template for improved performance.

LLM Name	Starling LM 7B Beta
Repository 🤗	https://huggingface.co/Nexusflow/Starling-LM-7B-beta
Model Size	7b
Required VRAM	14.4 GB
Updated	2025-09-23
Maintainer	Nexusflow
Model Type	mistral
Model Files	4.9 GB: 1-of-3 5.0 GB: 2-of-3 4.5 GB: 3-of-3
Supported Languages	en
Model Architecture	MistralForCausalLM
License	apache-2.0
Context Length	8192
Model Max Length	8192
Transformers Version	4.37.1
Tokenizer Class	LlamaTokenizer
Padding Token	<\|end_of_turn\|>
Vocabulary Size	32002
Torch Data Type	bfloat16

Quantized Models of the Starling LM 7B Beta

Model	Likes	Downloads	VRAM
Starling LM 7B Beta 8bit	2	14	7 GB
Starling LM 7B Beta AWQ	3	9	4 GB
Starling LM 7B Beta 4bit	4	38	4 GB
Starling LM 7B Beta GPTQ	4	3	4 GB

Best Alternatives to Starling LM 7B Beta

Best Alternatives	Context / RAM	Downloads	Likes
...Nemo Instruct 2407 Abliterated	1000K / 24.5 GB	254	20
MegaBeam Mistral 7B 512K	512K / 14.4 GB	11443	54
SpydazWeb AI HumanAI RP	512K / 14.4 GB	21	1
SpydazWeb AI HumanAI 002	512K / 14.4 GB	18	1
...daz Web AI ChatML 512K Project	512K / 14.5 GB	12	0
MegaBeam Mistral 7B 300K	282K / 14.4 GB	3779	16
MegaBeam Mistral 7B 300K	282K / 14.4 GB	7968	16
Hebrew Mistral 7B 200K	256K / 30 GB	1316	15
Astral 256K 7B V2	250K / 14.4 GB	10	0
Astral 256K 7B	250K / 14.4 GB	5	0

Rank the Starling LM 7B Beta Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

What open-source LLMs or SLMs are you in search of? 53254 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

Starling LM 7B Beta by Nexusflow