Model comparison
GLM-5.2-FP8-Uncensored vs GLM-5.2
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| GLM-5.2-FP8-Uncensored | GLM-5.2 | |
|---|---|---|
| Parameters | 753B | 753B |
| Architecture | glm-dsa | glm-dsa |
| Layers | — | 78 |
| Native context | — | 1,048,576 |
| Mixture of experts | no | yes, 256 experts |
| Quantizations published | 7 | 20 |
| Smallest quantization | 276.94 GiB | 169.33 GiB |
| Q4_K_M | 423.33 GiB | — |
| Licence | mit | mit |
KV cache by context
the term that decides long-context viability
| Context | GLM-5.2-FP8-Uncensored | GLM-5.2 | Ratio |
|---|---|---|---|
| 4,096 | — | 0.34 GiB | — |
| 8,192 | — | 0.69 GiB | — |
| 16,384 | — | 1.37 GiB | — |
| 32,768 | — | 2.74 GiB | — |
| 65,536 | — | 5.48 GiB | — |
| 131,072 | — | 10.97 GiB | — |