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

Magnum V2.5 12B Kto by anthracite-org

 »  All LLMs  »  anthracite-org  »  Magnum V2.5 12B Kto   URL Share it on

  5134   
Base model:anthracite-org/magn... Base model:finetune:anthracite...   Chat   Conversational   De   En   Es   Fr   It   Ja   Mistral   Pt   Region:us   Ru   Safetensors   Sharded   Tensorflow   Zh
Model Card on HF 🤗: https://huggingface.co/anthracite-org/magnum-v2.5-12b-kto 

Magnum V2.5 12B Kto Benchmarks

MMLU Pro: 24.61
GPQA: 5.82
MUSR: 9.98
BBH: 29.63
IFEval: 38.66 vs 88 (so35)-56.1%
MATH Lvl 5: 4.61
LLME Score: 0.32734
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Magnum V2.5 12B Kto (anthracite-org/magnum-v2.5-12b-kto)

Magnum V2.5 12B Kto Parameters and Internals

Model Type 
text generation, chat
Additional Notes 
KTO is an experimental release, part of a series of models. It's fine-tuned on top of Magnum-12b-v2. Experimental data was used for initial testing, with plans to scale up.
Supported Languages 
en (supported), fr (supported), de (supported), es (supported), it (supported), pt (supported), ru (supported), zh (supported), ja (supported)
Training Details 
Data Sources:
Stheno dataset (filtered), kalomaze/Opus_Instruct_25k, Nopm/Opus_WritingStruct, Gryphe/Sonnet3.5-SlimOrcaDedupCleaned, kalomaze/Opus_Instruct_3k
Methodology:
hybrid reinforcement learning strategy of KTO + DPOP using rejected data sampled from the original model as rejected and data from original finetuning dataset as chosen.
Input Output 
Input Format:
Instruct tuned with the ChatML formatting.
LLM NameMagnum V2.5 12B Kto
Repository 🤗https://huggingface.co/anthracite-org/magnum-v2.5-12b-kto 
Base Model(s)  Magnum 12B V2   anthracite-org/magnum-12b-v2
Model Size12b
Required VRAM24.5 GB
Updated2025-01-22
Maintaineranthracite-org
Model Typemistral
Model Files  4.9 GB: 1-of-5   4.9 GB: 2-of-5   4.9 GB: 3-of-5   4.9 GB: 4-of-5   4.9 GB: 5-of-5
Supported Languagesen fr de es it pt ru zh ja
Model ArchitectureMistralForCausalLM
Licenseapache-2.0
Context Length1024000
Model Max Length1024000
Transformers Version4.43.3
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<pad>
Vocabulary Size131072
Torch Data Typefloat16

Best Alternatives to Magnum V2.5 12B Kto

Best Alternatives
Context / RAM
Downloads
Likes
...r Nemo 12B Instruct R 21 09 241000K / 24.5 GB836599
...s PersonalityEngine V1.1.0 12B1000K / 24.5 GB33524
Mistral Nemo Kartoffel 12B1000K / 24.5 GB602
Saiga Nemo 12b1000K / 24.5 GB21399633
Mistral Nemo Bophades3 12B1000K / 24.5 GB522
Captain Eris Violet V0.420 12B1000K / 24.5 GB24820
MN 12B Mag Mell R11000K / 24.5 GB341889
Dolphin 2.9.3 Mistral Nemo 12B1000K / 24.5 GB803197
Captain BMO 12B1000K / 24.5 GB110818
SauerkrautLM Nemo 12B Instruct1000K / 24.5 GB601422
Note: green Score (e.g. "73.2") means that the model is better than anthracite-org/magnum-v2.5-12b-kto.

Rank the Magnum V2.5 12B Kto 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? 41728 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