What are the hardware requirements for Kilo 2x8B?

Kilo 2x8B requires approximately 27.5 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 Kilo 2x8B and how large is it?

Kilo 2x8B is developed by Alsebay, a model with 13.7b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Kilo 2x8B perform on standard benchmarks?

Kilo 2x8B has the following published scores: MMLU 64.15. Compare against reference models on this page or on the LLM Explorer leaderboards.

Kilo 2x8B by Alsebay — VRAM 27.5GB, 8K context

Name: Kilo 2x8B
Author: Alsebay

Kilo 2x8B is an open-source language model by Alsebay. Features: 13.7b LLM, VRAM: 27.5GB, Context: 8K, License: cc-by-nc-4.0, MoE, LLM Explorer Score: 0.13, Arc: 61, HellaSwag: 81.9, MMLU: 64.2, GSM8K: 52.

Autotrain compatible Base model:merge:sao10k/l3-sol... Base model:merge:vicgalle/role... Base model:sao10k/l3-solana-8b... Base model:vicgalle/roleplay-l... Endpoints compatible Merge Mixtral Moe Region:us Roleplay Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/Alsebay/Kilo-2x8B

Kilo 2x8B Parameters and Internals

Model Type

moe, merge

Additional Notes

Just some experiment, nothing else. I think it not too good for roleplay, you should try it

LLM Name	Kilo 2x8B
Repository 🤗	https://huggingface.co/Alsebay/Kilo-2x8B
Base Model(s)	Roleplay Llama 3 8B L3 Solana 8B V1 vicgalle/Roleplay-Llama-3-8B Sao10K/L3-Solana-8B-v1
Model Size	13.7b
Required VRAM	27.5 GB
Updated	2024-12-03
Maintainer	Alsebay
Model Type	mixtral
Model Files	1.1 GB: 1-of-15 2.0 GB: 2-of-15 2.0 GB: 3-of-15 2.0 GB: 4-of-15 2.0 GB: 5-of-15 2.0 GB: 6-of-15 2.0 GB: 7-of-15 2.0 GB: 8-of-15 2.0 GB: 9-of-15 2.0 GB: 10-of-15 2.0 GB: 11-of-15 2.0 GB: 12-of-15 2.0 GB: 13-of-15 2.0 GB: 14-of-15 0.4 GB: 15-of-15
Model Architecture	MixtralForCausalLM
License	cc-by-nc-4.0
Context Length	8192
Model Max Length	8192
Transformers Version	4.40.0
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|begin_of_text\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the Kilo 2x8B 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 Kilo 2x8B

Best Alternatives	Context / RAM	Downloads	Likes
L3.1 MoE 2x8B V0.2	128K / 27.3 GB	9	9
... Deepseek DeepHermes E32 13.7B	128K / 54.9 GB	5	0
HAI SER	128K / 27.3 GB	7	16
L3.1 Celestial Stone 2x8B	128K / 27.3 GB	312	3
...ma 3 2x8B Instruct MoE 64K Ctx	64K / 27.3 GB	12	4
Penny Llama3 2x8b	8K / 27.3 GB	6	1
Defne Llama3 2x8B	8K / 27.4 GB	15	6
Llama 3 Elyza Youko MoE 2x8B	8K / 27.3 GB	10	0
Llama 3 ELYZA Hermes 2x8B	8K / 27.4 GB	3	1
Inixion 2x8B V2	8K / 27.4 GB	6	2

Note: green Score (e.g. "73.2") means that the model is better than Alsebay/Kilo-2x8B.