Melvin56/ko-r1-7b-v2.0.3-GGUF
Original Model : OLAIR/ko-r1-7b-v2.0.3
All quants are made using the imatrix.
| Model | Size (GB) |
|---|---|
| Q2_K_S | 2.83 |
| Q2_K | 3.01 |
| Q3_K_M | 3.81 |
| Q3_K_L | 4.09 |
| Q4_K_M | 4.68 |
| Q5_K_M | 5.44 |
| Q6_K | 6.25 |
| Q8_0 | 8.1 |
| F16 | 15.2 |
| CPU (AVX2) | CPU (ARM NEON) | Metal | cuBLAS | rocBLAS | SYCL | CLBlast | Vulkan | Kompute | |
|---|---|---|---|---|---|---|---|---|---|
| K-quants | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ 🐢5 | ✅ 🐢5 | ❌ |
| I-quants | ✅ 🐢4 | ✅ 🐢4 | ✅ 🐢4 | ✅ | ✅ | Partial¹ | ❌ | ❌ | ❌ |
✅: feature works
🚫: feature does not work
❓: unknown, please contribute if you can test it youself
🐢: feature is slow
¹: IQ3_S and IQ1_S, see #5886
²: Only with -ngl 0
³: Inference is 50% slower
⁴: Slower than K-quants of comparable size
⁵: Slower than cuBLAS/rocBLAS on similar cards
⁶: Only q8_0 and iq4_nl