Helion 4x34B By Weyaxi: Benchmarks, Features and Detailed Analysis. Insights on Helion 4x34B.

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

Model Card on HF 🤗: https://huggingface.co/Weyaxi/Helion-4x34B

Helion 4x34B Benchmarks

ARC: 69.71 vs 96.7 (so35)^-27.9%

HellaSwag: 85.28 vs 95.3 (gpt4)^-10.5%

MMLU: 77.33 vs 88.3 (so35)^-12.4%

TruthfulQA: 63.91 vs 59 (gpt4)^8.3%

WinoGrande: 84.37 vs 87.5 (gpt4)^-3.6%

GSM8K: 72.25 vs 96.4 (so35)^-25.1%

LLME Score: 0.18579

^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)

Helion 4x34B Parameters and Internals

Model Type

text-generation

Additional Notes

This is a mixture of experts model (MOE) and allows utilizing various prompt templates from other similar models. Quantized versions are also available for optimized use.

LLM Name	Helion 4x34B
Repository 🤗	https://huggingface.co/Weyaxi/Helion-4x34B
Model Size	113.7b
Required VRAM	227.7 GB
Updated	2025-02-22
Maintainer	Weyaxi
Model Type	mixtral
Model Files	9.8 GB: 1-of-23 10.0 GB: 2-of-23 10.0 GB: 3-of-23 10.0 GB: 4-of-23 10.0 GB: 5-of-23 10.0 GB: 6-of-23 10.0 GB: 7-of-23 10.0 GB: 8-of-23 10.0 GB: 9-of-23 10.0 GB: 10-of-23 10.0 GB: 11-of-23 10.0 GB: 12-of-23 10.0 GB: 13-of-23 10.0 GB: 14-of-23 10.0 GB: 15-of-23 10.0 GB: 16-of-23 10.0 GB: 17-of-23 10.0 GB: 18-of-23 10.0 GB: 19-of-23 10.0 GB: 20-of-23 10.0 GB: 21-of-23 10.0 GB: 22-of-23 7.9 GB: 23-of-23
Model Architecture	MixtralForCausalLM
License	other
Context Length	200000
Model Max Length	200000
Transformers Version	4.36.2
Tokenizer Class	LlamaTokenizer
Padding Token	<s>
Vocabulary Size	64000
Torch Data Type	bfloat16

Quantized Models of the Helion 4x34B

Model	Likes	Downloads	VRAM
Helion 4x34B GPTQ	3	33	58 GB
Helion 4x34B GGUF	3	26	41 GB

Best Alternatives to Helion 4x34B

Best Alternatives	Context / RAM	Downloads	Likes
Astralis 4x34B	195K / 227.7 GB	1762	3
APUS XDAN 4.0 MoE V2	32K / 227.6 GB	13	0

Rank the Helion 4x34B 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? 43508 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.

Release v20241227

Support LLM Explorer