Model Card on HF 🤗: https://huggingface.co/anthracite-forge/magnum-v3-27b-kto-r3
Magnum V3 27B Kto R3 Benchmarks
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)
Magnum V3 27B Kto R3 Parameters and Internals
| LLM Name | Magnum V3 27B Kto R3 |
| Repository 🤗 | https://huggingface.co/anthracite-forge/magnum-v3-27b-kto-r3 |
| Model Size | 27b |
| Required VRAM | 53.9 GB |
| Updated | 2025-01-14 |
| Maintainer | anthracite-forge |
| Model Type | gemma2 |
| Model Files | |
| Model Architecture | Gemma2ForCausalLM |
| Context Length | 8192 |
| Model Max Length | 8192 |
| Transformers Version | 4.43.2 |
| Tokenizer Class | GemmaTokenizer |
| Padding Token | <pad> |
| Vocabulary Size | 256000 |
| Torch Data Type | bfloat16 |
Best Alternatives to Magnum V3 27B Kto R3
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| SystemGemma2 27B It | 32K / 54.7 GB | 96 | 1 |
| SystemGemma2 27B It | 32K / 54.7 GB | 53 | 3 |
| Gemma 2 27B It | 8K / 54.7 GB | 91565 | 502 |
| Gemma 2 27B | 8K / 108.3 GB | 31751 | 193 |
| ...s Horizon AI Korean Superb 27B | 8K / 54.8 GB | 123 | 1 |
| Magnum V3 27B Kto | 8K / 108.2 GB | 4182 | 12 |
| ...izon AI Korean Avengers V3 27B | 8K / 54.8 GB | 15 | 0 |
| GemmaStock1 27B | 8K / 54.8 GB | 8 | 0 |
| ...izon AI Korean Avengers V2 27B | 8K / 54.8 GB | 18 | 0 |
| Gemma2Slerp2 27B | 8K / 54.8 GB | 10 | 0 |
Note: green Score (e.g. "73.2") means that the model is better than anthracite-forge/magnum-v3-27b-kto-r3 .
Rank the Magnum V3 27B Kto R3 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! 🌟
What open-source LLMs or SLMs are you in search of? 41636 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227