LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

MiS Firefly V0.2 22B by invisietch

 »  All LLMs  »  invisietch  »  MiS Firefly V0.2 22B   URL Share it on

  162   
  Merged Model   Autotrain compatible   Axolotl   Conversational   En   Endpoints compatible   Mistral   Model-index   Not-for-all-audiences   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/invisietch/MiS-Firefly-v0.2-22B 

MiS Firefly V0.2 22B Benchmarks

MMLU Pro: 29.11
GPQA: 7.27
MUSR: 17.81
BBH: 36.08
IFEval: 53.71 vs 88 (so35)-39%
MATH Lvl 5: 16.54
LLME Score: 0.30977
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
MiS Firefly V0.2 22B (invisietch/MiS-Firefly-v0.2-22B)

MiS Firefly V0.2 22B Parameters and Internals

Model Type 
text-generation
Use Cases 
Areas:
creative writing, roleplay
Additional Notes 
Fix for the quantization issue in Firefly v0.1. Largely uncensored, supports context up to 32,768 tokens.
Training Details 
Data Volume:
474 million token text completion dataset + 172 million token instruct dataset
Methodology:
Finetune with coherence, decensorship & roleplay
Context Length:
32768
Training Time:
32 hours on 4x Nvidia A100 80GB
Hardware Used:
4x Nvidia A100 80GB
LLM NameMiS Firefly V0.2 22B
Repository 🤗https://huggingface.co/invisietch/MiS-Firefly-v0.2-22B 
Merged ModelYes
Model Size22b
Required VRAM44.5 GB
Updated2025-03-22
Maintainerinvisietch
Model Typemistral
Model Files  3.9 GB: 1-of-12   3.9 GB: 2-of-12   3.9 GB: 3-of-12   3.9 GB: 4-of-12   3.9 GB: 5-of-12   3.9 GB: 6-of-12   3.9 GB: 7-of-12   3.9 GB: 8-of-12   3.9 GB: 9-of-12   3.9 GB: 10-of-12   3.9 GB: 11-of-12   1.6 GB: 12-of-12
Supported Languagesen
Model ArchitectureMistralForCausalLM
Licenseother
Context Length32768
Model Max Length32768
Transformers Version4.44.2
Tokenizer ClassLlamaTokenizer
Padding Token</s>
Vocabulary Size32768
Torch Data Typebfloat16

Best Alternatives to MiS Firefly V0.2 22B

Best Alternatives
Context / RAM
Downloads
Likes
MS Schisandra 22B V0.2128K / 44.7 GB399
...ntheon RP Pure 1.6.2 22B Small128K / 44.7 GB11829
MS Meadowlark 22B128K / 44.7 GB15414
... V4x1.6.2RP Cydonia VXXX 22B 8128K / 44.7 GB275
Beeper King 22B128K / 44.7 GB217
... V4x1.6.2RP Cydonia VXXX 22B 6128K / 44.7 GB193
MS Dampf 2409 22B128K / 44.7 GB160
Kyne R2 22B128K / 44.7 GB200
MS Fujin 2409 22B128K / 44.7 GB140
MS A Coolyte 2409 22B128K / 44.7 GB130
Note: green Score (e.g. "73.2") means that the model is better than invisietch/MiS-Firefly-v0.2-22B.

Rank the MiS Firefly V0.2 22B 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? 45389 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