Yayi 7B Benchmarks
ARC: 46.33 vs 96.7 (so35)-52.1%
HellaSwag: 61.72 vs 95.3 (gpt4)-35.2%
MMLU: 36.34 vs 88.3 (so35)-58.8%
TruthfulQA: 43.7 vs 59 (gpt4)-25.9%
WinoGrande: 62.27 vs 87.5 (gpt4)-28.8%
GSM8K: 0.91 vs 96.4 (so35)-99.1%
LLME Score: 0.17713
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)
Yayi 7B Parameters and Internals
| LLM Name | Yayi 7B |
| Repository 🤗 | https://huggingface.co/wenge-research/yayi-7b |
| Model Size | 7b |
| Required VRAM | 28.2 GB |
| Updated | 2024-09-07 |
| Maintainer | wenge-research |
| Model Type | bloom |
| Model Files | |
| Supported Languages | zh en |
| Model Architecture | BloomForCausalLM |
| Transformers Version | 4.28.1 |
| Tokenizer Class | BloomTokenizer |
| Padding Token | <pad> |
| Vocabulary Size | 250684 |
| Torch Data Type | bfloat16 |
Best Alternatives to Yayi 7B
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| ...lama 2 7B Chat Hf Text Teacher | 0K / 2.2 GB | 6 | 0 |
| ...Qa Llama 2 7B Chat Hf Uld Loss | 0K / 2.2 GB | 5 | 0 |
| ...tral 7B Instruct V0.2 Uld Loss | 0K / 2.2 GB | 5 | 0 |
| ... 7B Instruct V0.2 Text Teacher | 0K / 2.2 GB | 5 | 0 |
| Gogpt 7B Bloom | 0K / 32.3 GB | 671 | 3 |
| Phoenix Inst Chat 7B | 0K / 16.2 GB | 1320 | 43 |
| Vietcuna 7B 2k5 | 0K / 14.2 GB | 5 | 0 |
| Bloom Xp3 | 0K / 32.4 GB | 8 | 0 |
| Vietcuna 7B V3 | 0K / 14.2 GB | 133 | 8 |
| Huozi 7B Rlhf | 0K / 15.6 GB | 7 | 3 |
Note: green Score (e.g. "73.2") means that the model is better than wenge-research/yayi-7b .
Rank the Yayi 7B 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? 35693 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v2024072803