See DeepSeek-V3.1-Terminus 5.5bit MLX in action - demonstration video
Notes
- No longer available on HF due to storage restrictions, please contact if you require this version
- Original DeepSeek V3.1 performed better in our testing
- Runs on a single M3 Ultra 512GB RAM using Inferencer app
- Memory usage: ~480 GB
- Expect ~15-20 tokens/s
- Quantized with a modified version of MLX 0.27
- For more details see demonstration video or visit DeepSeek-V3.1-Terminus.
q5.5bit quant typically achieves 1.141 perplexity in our testing
| Quantization | Perplexity |
|---|---|
| q2.5 | 41.293 |
| q3.5 | 1.900 |
| q4.5 | 1.168 |
| q5.5 | 1.141 |
| q6.5 | 1.128 |
| q8.5 | 1.128 |
Disclaimer
We are not the creator, originator, or owner of any model listed. Each model is created and provided by third parties. Models may not always be accurate or contextually appropriate. You are responsible for verifying the information before making important decisions. We are not liable for any damages, losses, or issues arising from its use, including data loss or inaccuracies in AI-generated content.
- Downloads last month
- 163
Model tree for inferencerlabs/DeepSeek-V3.1-Terminus-MLX-5.5bit
Base model
deepseek-ai/DeepSeek-V3.1-Base Quantized
deepseek-ai/DeepSeek-V3.1-Terminus