What are the hardware requirements for MixTAO 7Bx2 MoE V8.1?

MixTAO 7Bx2 MoE V8.1 requires approximately 25.8 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 MixTAO 7Bx2 MoE V8.1 and how large is it?

MixTAO 7Bx2 MoE V8.1 is developed by zhengr, a model with 12.9b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does MixTAO 7Bx2 MoE V8.1 perform on standard benchmarks?

MixTAO 7Bx2 MoE V8.1 has the following published scores: MMLU 64.92. Compare against reference models on this page or on the LLM Explorer leaderboards.

MixTAO 7Bx2 MoE V8.1 by zhengr — VRAM 25.8GB, 32K context

Name: MixTAO 7Bx2 MoE V8.1
Rating: 1 (3 reviews)
Author: zhengr

MixTAO 7Bx2 MoE V8.1 is an open-source language model by zhengr. Features: 12.9b LLM, VRAM: 25.8GB, Context: 32K, License: apache-2.0, MoE, LLM Explorer Score: 0.2, Arc: 73.8, HellaSwag: 89.2, MMLU: 64.9, GSM8K: 71.1.

Autotrain compatible Conversational Endpoints compatible Mixtral Model-index Moe Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/zhengr/MixTAO-7Bx2-MoE-v8.1

MixTAO 7Bx2 MoE V8.1 Parameters and Internals

Model Type

MoE, text generation

Additional Notes

MixTAO-7Bx2-MoE is a Mixture of Experts model mainly used for large model technology experiments.

Input Output

Input Format:

Alpaca Prompt Template

Accepted Modalities:

text

Output Format:

Text response

LLM Name	MixTAO 7Bx2 MoE V8.1
Repository 🤗	https://huggingface.co/zhengr/MixTAO-7Bx2-MoE-v8.1
Model Size	12.9b
Required VRAM	25.8 GB
Updated	2024-09-16
Maintainer	zhengr
Model Type	mixtral
Model Files	1.9 GB: 1-of-13 2.0 GB: 2-of-13 2.0 GB: 3-of-13 2.0 GB: 4-of-13 1.9 GB: 5-of-13 2.0 GB: 6-of-13 2.0 GB: 7-of-13 2.0 GB: 8-of-13 2.0 GB: 9-of-13 2.0 GB: 10-of-13 2.0 GB: 11-of-13 2.0 GB: 12-of-13 2.0 GB: 13-of-13
Model Architecture	MixtralForCausalLM
License	apache-2.0
Context Length	32768
Model Max Length	32768
Transformers Version	4.38.1
Tokenizer Class	LlamaTokenizer
Padding Token	<s>
Vocabulary Size	32000
Torch Data Type	bfloat16

Rank the MixTAO 7Bx2 MoE V8.1 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

Quantized Models of the MixTAO 7Bx2 MoE V8.1

Model	Likes	Downloads	VRAM
MixTAO 7Bx2 MoE V8.1 GGUF	12	624	4 GB

Best Alternatives to MixTAO 7Bx2 MoE V8.1

Best Alternatives	Context / RAM	Downloads	Likes
MixTAO 7Bx2 MoE V8.1	32K / 25.8 GB	15381	56
Inf Silent Kunoichi V0.1 2x7B	32K / 25.6 GB	8	0
Inf Silent Kunoichi V0.2 2x7B	32K / 25.6 GB	9	1
LogoS 7Bx2 MoE 13B V0.2	32K / 25.9 GB	7978	10
MultiMash8 13B Slerp	32K / 25.7 GB	5	0
MultiMash11 13B Slerp	32K / 25.7 GB	9	0
MultiMash9 13B Slerp	32K / 25.7 GB	5	0
MultiMash10 13B Slerp	32K / 25.7 GB	8	0
Multimash3 12B Slerp	32K / 25.7 GB	8	0
MixTaoTruthful 13B Slerp	32K / 25.7 GB	5	0

Note: green Score (e.g. "73.2") means that the model is better than zhengr/MixTAO-7Bx2-MoE-v8.1.