Franken MoE 18B V0.1 By MaziyarPanahi: Benchmarks, Features and Detailed Analysis. Insights on Franken MoE 18B V0.1.

Autotrain compatible Endpoints compatible Merge Mixtral Moe Region:us Safetensors Sharded Tensorflow

Franken MoE 18B V0.1 Benchmarks

ARC: 72.1 vs 96.7 (so35)^-25.4%

HellaSwag: 88.3 vs 95.3 (gpt4)^-7.3%

MMLU: 65.01 vs 88.3 (so35)^-26.4%

TruthfulQA: 67.51 vs 59 (gpt4)^14.4%

WinoGrande: 83.74 vs 87.5 (gpt4)^-4.3%

GSM8K: 69.45 vs 96.4 (so35)^-28%

LLME Score: 0.20673

^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)

Franken MoE 18B V0.1 Parameters and Internals

LLM Name	Franken MoE 18B V0.1
Repository 🤗	https://huggingface.co/MaziyarPanahi/Franken-MoE-18B-v0.1
Model Size	18.5b
Required VRAM	37.1 GB
Updated	2024-09-07
Maintainer	MaziyarPanahi
Model Type	mixtral
Model Files	10.0 GB: 1-of-4 10.0 GB: 2-of-4 10.0 GB: 3-of-4 7.1 GB: 4-of-4
Model Architecture	MixtralForCausalLM
License	apache-2.0
Context Length	32768
Model Max Length	32768
Transformers Version	4.37.2
Tokenizer Class	LlamaTokenizer
Padding Token	<s>
Vocabulary Size	32000
Torch Data Type	bfloat16

Franken MoE 18B V0.1 (MaziyarPanahi/Franken-MoE-18B-v0.1)

Quantized Models of the Franken MoE 18B V0.1

Model	Likes	Downloads	VRAM
Franken MoE 18B V0.1 GGUF	0	142	6 GB

Best Alternatives to Franken MoE 18B V0.1

Best Alternatives	Context / RAM	Downloads	Likes
Lumina 3.5	32K / 37.1 GB	2497	0
Topxtral 4x7B V0.1	32K / 37.1 GB	5123	4
EastAsia 4x7B MoE Experiment	32K / 37.1 GB	670	1
Hyperion 3.0 Mixtral 3x7B	32K / 37.1 GB	62	4
Blitz AI MoE V0.7	32K / 37.1 GB	60	1
Blitz AI MoE V0.4	32K / 37.1 GB	58	1
HeroBophades 3x7B	32K / 37.1 GB	7	1
Wizard Kun Lake 3x7B MoE	32K / 37.1 GB	6	1
NaruMOE 3x7B V2	32K / 37.1 GB	5	0
MoE 3x7b QA Code Inst	32K / 37 GB	78	4

Note: green Score (e.g. "73.2") means that the model is better than MaziyarPanahi/Franken-MoE-18B-v0.1.

Rank the Franken MoE 18B V0.1 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? 35693 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 v2024072803

Support LLM Explorer