SCB 10X · 70.6B · Dense
Largest Typhoon model, built on Llama 3.1 — strong Thai math/code benchmarks
0 downloads
0 likes
2024-12 88K context
Use Cases
chat multilingual reasoning code
| Quant | Bits | VRAM | Quality | Status |
|---|---|---|---|---|
| Q2_K | 2 | 23.1 GB | low | — |
| Q3_K_M | 3 | 32.1 GB | moderate | — |
| Q4_K_M | 4 | 36.7 GB | good | — |
| Q5_K_M | 5 | 45.7 GB | good | — |
| Q6_K | 6 | 54.7 GB | excellent | — |
| Q8_0 | 8 | 72.8 GB | excellent | — |
| F16 | 16 | 145.2 GB | lossless | — |
GPUs that run Typhoon 2 70B
Typhoon 2 70B needs about 36.7 GB of VRAM at Q4_K_M, so it runs comfortably on a GPU with 48 GB of VRAM or more (16 of the GPUs we track). Common cards, smallest first:
| VRAM | GPUs |
|---|---|
| 48 GB | RTX 6000 Ada , RTX 5880 Ada , RTX A6000 |
| 80 GB | A100 , H100 |
| 96 GB | RTX PRO 6000 , GH200 , Ryzen AI MAX+ 395 |
| 128 GB | DGX Spark |
About this model
Largest Typhoon model, built on Llama 3.1 — strong Thai math/code benchmarks