Mistral 22B V0.2 By Vezora: Benchmarks, Features and Detailed Analysis. Insights on Mistral 22B V0.2.

Autotrain compatible Endpoints compatible Mistral Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/Vezora/Mistral-22B-v0.2

Mistral 22B V0.2 Benchmarks

ARC: 52.39 vs 96.7 (so35)^-45.8%

HellaSwag: 78.63 vs 95.3 (gpt4)^-17.5%

MMLU: 54.63 vs 88.3 (so35)^-38.1%

TruthfulQA: 48.84 vs 59 (gpt4)^-17.2%

WinoGrande: 76.16 vs 87.5 (gpt4)^-13%

GSM8K: 26.99 vs 96.4 (so35)^-72%

LLME Score: 0.18334

^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)

Mistral 22B V0.2 (Vezora/Mistral-22B-v0.2)

🌟 Advertise your project 🚀

Mistral 22B V0.2 Parameters and Internals

Model Type

dense model, MOE to Dense conversion

Use Cases

Applications:

Coding, Math proficiency

Limitations:

This model is highly uncensored

Additional Notes

The model requires Guanaco prompt format for optimal results. It's still experimental and v0.3 is under training.

Training Details

Data Volume:

8x more data than v0.1

Methodology:

Experimental model, MOE to Dense conversion

Context Length:

32000

Input Output

Input Format:

Guanaco prompt format

Release Notes

Version:

v.02

Date:

April 13

Notes:

Handcrafted experimental model, trained on more data, capabilities improved

LLM Name	Mistral 22B V0.2
Repository 🤗	https://huggingface.co/Vezora/Mistral-22B-v0.2
Model Size	22b
Required VRAM	44.7 GB
Updated	2025-09-14
Maintainer	Vezora
Model Type	mistral
Model Files	4.9 GB: 1-of-9 5.0 GB: 2-of-9 5.0 GB: 3-of-9 4.9 GB: 4-of-9 5.0 GB: 5-of-9 5.0 GB: 6-of-9 4.9 GB: 7-of-9 5.0 GB: 8-of-9 5.0 GB: 9-of-9
Model Architecture	MistralForCausalLM
License	apache-2.0
Context Length	65536
Model Max Length	65536
Transformers Version	4.40.0.dev0
Tokenizer Class	LlamaTokenizer
Vocabulary Size	32000
Torch Data Type	bfloat16

Quantized Models of the Mistral 22B V0.2

Model	Likes	Downloads	VRAM
Mistral 22B V0.2 GGUF	0	38	8 GB
Mistral 22B V0.2 AWQ	2	6	12 GB

Best Alternatives to Mistral 22B V0.2

Best Alternatives	Context / RAM	Downloads	Likes
MS Schisandra 22B V0.2	128K / 44.7 GB	5	9
...ntheon RP Pure 1.6.2 22B Small	128K / 44.7 GB	10	32
MS Meadowlark 22B	128K / 44.7 GB	79	15
...rker The Final Abomination 22B	128K / 44.7 GB	7	6
...er The Final Transgression 22B	128K / 44.7 GB	10	3
...Darker The Final Directive 22B	128K / 44.7 GB	7	0
The Omega Directive M 22B V1.0	128K / 44.7 GB	9	2
Retrograde Omega M 22B V1.0	128K / 44.7 GB	8	0
Beeper King 22B	128K / 44.7 GB	8	7
... V4x1.6.2RP Cydonia VXXX 22B 8	128K / 44.7 GB	7	5

Note: green Score (e.g. "73.2") means that the model is better than Vezora/Mistral-22B-v0.2.

Rank the Mistral 22B V0.2 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? 51368 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 v20241124

Support LLM Explorer