My first attempt at quantization using PrismQuant
No warranty, nothing. Feel free to test it out and let me know how it go.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Introducing PrismaQuant | 166 | 7283 | June 12, 2026 | |
| Qwen3.6 27B Binary/Ternary release by Prism-ML | 7 | 1589 | July 18, 2026 | |
| Introducing PrismaScout -- PrismaQuant v2! | 138 | 9946 | July 21, 2026 | |
| What's the best speed we can get with Qwen 3.6 27B without quantizing? | 64 | 24283 | July 6, 2026 | |
| MiniMax M2.7 TQ3 - A TurboQuant 3-bit quantized version of MiniMax-M2.7 for single DGX Spark | 8 | 3566 | June 21, 2026 | |
| Qwen/Qwen3.5-122B-A10B - Alibaba/Qwen thought about us... :-D | 340 | 17903 | March 24, 2026 | |
| Qwen3.5-122B-A10B on single Spark: 15 → 21.5 tok/s with hybrid GPTQ-INT4 + FP8 dense layers (https://github.com/rmstxrx/vllm-hybrid-quant) | 9 | 948 | March 20, 2026 | |
| Introducing Spark Auto Round /w OpenCode Instruct dataset | 110 | 4387 | July 17, 2026 | |
| Benchmark Report: unsloth/Qwen3.6-35B-A3B-NVFP4-Fast vs nvidia/Qwen3.6-35B-A3B-NVFP4 | 5 | 1261 | July 20, 2026 | |
| Qwen 3.5 SLM on DGX GB10 | 12 | 655 | March 3, 2026 |