Model comparison
Falcon3-1B-Instruct vs ced-base
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| Falcon3-1B-Instruct | ced-base | |
|---|---|---|
| Parameters | 1.7B | 86M |
| Architecture | llama | ced |
| Layers | 18 | — |
| Native context | 8,192 | — |
| Mixture of experts | no | no |
| Quantizations published | 30 | 3 |
| Smallest quantization | 0.21 GiB | 0.12 GiB |
| Q4_K_M | 0.98 GiB | — |
| Licence | other | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | Falcon3-1B-Instruct | ced-base | Ratio |
|---|---|---|---|
| 4,096 | 0.28 GiB | — | — |
| 8,192 | 0.56 GiB | — | — |
| 16,384 | 1.13 GiB | — | — |
| 32,768 | 2.25 GiB | — | — |
| 65,536 | 4.50 GiB | — | — |
| 131,072 | 9.00 GiB | — | — |