Arxiv:2402.05044 Ai-safety Autotrain compatible Benchmark Dataset:lmsys/lmsys-chat-1m Dataset:lmsys/toxic-chat Dataset:pku-alignment/beaverta... En Endpoints compatible Evaluation Judge Mistral Region:us Safetensors Safety Safetyguard Salad-bench Sharded Tensorflow
Model Card on HF 🤗: https://huggingface.co/OpenSafetyLab/MD-Judge-v0.1
MD Judge V0.1 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)
MD Judge V0.1 Parameters and Internals
| Model Type |
|
| LLM Name | MD Judge V0.1 |
| Repository 🤗 | https://huggingface.co/OpenSafetyLab/MD-Judge-v0.1 |
| Model Size | 1m |
| Required VRAM | 14.4 GB |
| Updated | 2025-02-22 |
| Maintainer | OpenSafetyLab |
| Model Type | mistral |
| Model Files | |
| Supported Languages | en |
| Model Architecture | MistralForCausalLM |
| License | apache-2.0 |
| Context Length | 32768 |
| Model Max Length | 32768 |
| Transformers Version | 4.36.1 |
| Tokenizer Class | LlamaTokenizer |
| Padding Token | <unk> |
| Vocabulary Size | 32000 |
| Torch Data Type | float16 |
Best Alternatives to MD Judge V0.1
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Ganga 1B | 128K / 4 GB | 278 | 15 |
| Nucleus 1B Alpha 1 | 32K / 2.3 GB | 173 | 12 |
| Experimental Mistral 1B V00 | 32K / 3.5 GB | 161 | 2 |
| Mallam 1.1B 20K Instructions | 32K / 2.2 GB | 72 | 1 |
| Lmlab Mistral 1B Untrained | 32K / 4.5 GB | 35 | 7 |
| Empty Phi 1B | 3K / 5.8 GB | 10 | 0 |
| Phi Filtered 500M C1 | 2K / 0.7 GB | 169 | 0 |
| Nucleus 1B GPTQ | 32K / 1 GB | 66 | 1 |
| Mistral 1B GPTQ | 32K / 1 GB | 66 | 0 |
Note: green Score (e.g. "73.2") means that the model is better than OpenSafetyLab/MD-Judge-v0.1 .
Rank the MD Judge 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! 🌟
What open-source LLMs or SLMs are you in search of? 43470 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227