Model comparison
starcoder2-7b vs llama-3-youko-8b
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| starcoder2-7b | llama-3-youko-8b | |
|---|---|---|
| Parameters | 7.2B | 8.0B |
| Architecture | starcoder2 | llama |
| Layers | 32 | 32 |
| Native context | 16,384 | 8,192 |
| Mixture of experts | no | no |
| Quantizations published | 12 | 2 |
| Smallest quantization | 2.54 GiB | 5.34 GiB |
| Q4_K_M | 4.10 GiB | — |
| Licence | bigcode-openrail-m | llama3 |
KV cache by context
the term that decides long-context viability
| Context | starcoder2-7b | llama-3-youko-8b | Ratio |
|---|---|---|---|
| 4,096 | — | 0.50 GiB | — |
| 8,192 | — | 1.00 GiB | — |
| 16,384 | — | 2.00 GiB | — |
| 32,768 | — | 4.00 GiB | — |
| 65,536 | — | 8.00 GiB | — |
| 131,072 | — | 16.00 GiB | — |