What are the hardware requirements for Taiwan LLM 8x7B DPO?

Taiwan LLM 8x7B DPO requires approximately 90 GB of VRAM and supports a context window of 32K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Taiwan LLM 8x7B DPO and how large is it?

Taiwan LLM 8x7B DPO is developed by yentinglin, a model with 46.7b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Taiwan LLM 8x7B DPO perform on standard benchmarks?

Taiwan LLM 8x7B DPO has the following published scores: MMLU 72.43. Compare against reference models on this page or on the LLM Explorer leaderboards.

Taiwan LLM 8x7B DPO by yentinglin — VRAM 90GB, 32K context

Name: Taiwan LLM 8x7B DPO
Author: yentinglin

Taiwan LLM 8x7B DPO is an open-source language model by yentinglin. Features: 46.7b LLM, VRAM: 90GB, Context: 32K, License: apache-2.0, MoE, LLM Explorer Score: 0.12, Arc: 71, HellaSwag: 87.2, MMLU: 72.4, GSM8K: 71.1.

Arxiv:2311.17487 Conversational Endpoints compatible Mixtral Moe Region:us Safetensors Sharded Tensorflow Zh

Model Card on HF 🤗: https://huggingface.co/yentinglin/Taiwan-LLM-8x7B-DPO

Taiwan LLM 8x7B DPO Parameters and Internals

Model Type

Mixtral MoE, text generation, fine-tuned

Supported Languages

Traditional Chinese (primary (zh-tw))

LLM Name	Taiwan LLM 8x7B DPO
Repository 🤗	https://huggingface.co/yentinglin/Taiwan-LLM-8x7B-DPO
Model Size	46.7b
Required VRAM	90 GB
Updated	2026-08-04
Maintainer	yentinglin
Model Type	mixtral
Model Files	2.0 GB: 1-of-48 2.0 GB: 2-of-48 2.0 GB: 3-of-48 2.0 GB: 4-of-48 2.0 GB: 5-of-48 2.0 GB: 6-of-48 2.0 GB: 7-of-48 2.0 GB: 8-of-48 2.0 GB: 9-of-48 2.0 GB: 10-of-48 2.0 GB: 11-of-48 2.0 GB: 12-of-48 2.0 GB: 13-of-48 2.0 GB: 14-of-48 2.0 GB: 15-of-48 2.0 GB: 16-of-48 2.0 GB: 17-of-48 2.0 GB: 18-of-48 2.0 GB: 19-of-48 2.0 GB: 20-of-48 2.0 GB: 21-of-48 2.0 GB: 22-of-48 2.0 GB: 23-of-48 2.0 GB: 24-of-48 2.0 GB: 25-of-48 2.0 GB: 26-of-48 2.0 GB: 27-of-48 2.0 GB: 28-of-48 2.0 GB: 29-of-48 2.0 GB: 30-of-48 2.0 GB: 31-of-48 2.0 GB: 32-of-48 2.0 GB: 33-of-48 2.0 GB: 34-of-48 2.0 GB: 35-of-48 2.0 GB: 36-of-48 2.0 GB: 37-of-48 2.0 GB: 38-of-48 2.0 GB: 39-of-48 2.0 GB: 40-of-48 2.0 GB: 41-of-48 2.0 GB: 42-of-48 2.0 GB: 43-of-48 2.0 GB: 44-of-48 2.0 GB: 45-of-48
Supported Languages	zh
Model Architecture	MixtralForCausalLM
License	apache-2.0
Context Length	32768
Model Max Length	32768
Transformers Version	4.37.2
Vocabulary Size	32002
Torch Data Type	bfloat16

Rank the Taiwan LLM 8x7B DPO 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 Taiwan LLM 8x7B DPO

Best Alternatives	Context / RAM	Downloads	Likes
Mixtral 8x7B Instruct V0.1	32K / 93.6 GB	842893	4720
Nous Hermes 2 Mixtral 8x7B DPO	32K / 93.6 GB	40865	457
Mixtral 8x7B V0.1	32K / 93.6 GB	49231	1820
Sensualize Mixtral Bf16	32K / 93.6 GB	0	0
Skadi Mixtral V1	32K / 93.5 GB	0	0
Franziska Mixtral V1	32K / 93.5 GB	0	0
Typhon Mixtral V1	32K / 93.4 GB	0	0
GritLM 8x7B KTO	32K / 93.6 GB	7976	3
Mixtral 8x7B Instruct V0.1	32K / 93.6 GB	7	0
Smaug Mixtral V0.1	32K / 187.7 GB	8437	12

Note: green Score (e.g. "73.2") means that the model is better than yentinglin/Taiwan-LLM-8x7B-DPO.