What are the hardware requirements for SJT 8B?

SJT 8B requires approximately 17.8 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed SJT 8B and how large is it?

SJT 8B is developed by Sakalti, a model with 8b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate SJT 8B?

SJT 8B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

SJT 8B by Sakalti — VRAM 17.8GB, 128K context

Name: SJT 8B
Author: Sakalti

SJT 8B is an open-source language model by Sakalti. Features: 8b LLM, VRAM: 17.8GB, Context: 128K, Merged, LLM Explorer Score: 0.25.

Merged Model Base model:finetune:sakalti/mo... Base model:sakalti/model-3 Conversational Endpoints compatible Qwen2 Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/Sakalti/SJT-8B

SJT 8B Parameters and Internals

LLM Name	SJT 8B
Repository 🤗	https://huggingface.co/Sakalti/SJT-8B
Base Model(s)	Model 3 Sakalti/model-3
Merged Model	Yes
Model Size	8b
Required VRAM	17.8 GB
Updated	2026-07-16
Maintainer	Sakalti
Model Type	qwen2
Model Files	1.1 GB: 1-of-34 1.1 GB: 2-of-34 0.5 GB: 3-of-34 0.5 GB: 4-of-34 0.4 GB: 5-of-34 0.5 GB: 6-of-34 0.4 GB: 7-of-34 0.5 GB: 8-of-34 0.4 GB: 9-of-34 0.5 GB: 10-of-34 0.4 GB: 11-of-34 0.5 GB: 12-of-34 0.5 GB: 13-of-34 0.5 GB: 14-of-34 0.5 GB: 15-of-34 0.5 GB: 16-of-34 0.5 GB: 17-of-34 0.5 GB: 18-of-34 0.5 GB: 19-of-34 0.5 GB: 20-of-34 0.5 GB: 21-of-34 0.5 GB: 22-of-34 0.5 GB: 23-of-34 0.5 GB: 24-of-34 0.5 GB: 25-of-34 0.5 GB: 26-of-34 0.5 GB: 27-of-34 0.5 GB: 28-of-34 0.5 GB: 29-of-34 0.5 GB: 30-of-34 0.5 GB: 31-of-34 0.5 GB: 32-of-34 0.5 GB: 33-of-34 0.5 GB: 34-of-34
Model Architecture	Qwen2ForCausalLM
Context Length	131072
Model Max Length	131072
Transformers Version	4.46.3
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	152064
Torch Data Type	float16
Errors	replace

Rank the SJT 8B capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

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

Best Alternatives to SJT 8B

Best Alternatives	Context / RAM	Downloads	Likes
Q2.5 TheGrimoire 7B Base Test0	128K / 15.2 GB	612	0
L3.1 TheGrimoire 8B Base Test0	128K / 15.2 GB	179	0
... 1.5B DPO Tulu3 Decontaminated	128K / 3.1 GB	18	0
Sailor2 L 8B	128K / 17.1 GB	8	0
Kepler 8B Instruct V2	32K / 15.4 GB	0	1
Sailor2 8B Chat	4K / 17.1 GB	661	19
Squeal Ai 8m Base	1K / 0 GB	479	0
SupaMATH	128K / 5.6 GB	5	0
Sailor2 8B Chat Uncensored	32K / 17.1 GB	93	2
BloomVN 8B Chat	32K / 17.1 GB	20	5

Note: green Score (e.g. "73.2") means that the model is better than Sakalti/SJT-8B.