Model comparison
starcoder2-3b vs Qwen3-VL-8B-Instruct-abliterated-v1
At Q4_K_M, starcoder2-3b is the smaller download — 1,848,976,448 bytes against 5,027,785,792.
From the file· summed bytes, KV per layer
Side by side
| starcoder2-3b | Qwen3-VL-8B-Instruct-abliterated-v1 | |
|---|---|---|
| Parameters | 3.0B | 8.8B |
| Architecture | starcoder2 | qwen3vl |
| Layers | 30 | — |
| Native context | 16,384 | — |
| Mixture of experts | no | no |
| Quantizations published | 37 | 48 |
| Smallest quantization | 1.07 GiB | 1.97 GiB |
| Q4_K_M | 1.72 GiB | 4.68 GiB |
| Licence | bigcode-openrail-m | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | starcoder2-3b | Qwen3-VL-8B-Instruct-abliterated-v1 | Ratio |
|---|---|---|---|
| 4,096 | — | — | — |
| 8,192 | — | — | — |
| 16,384 | — | — | — |
| 32,768 | — | — | — |
| 65,536 | — | — | — |
| 131,072 | — | — | — |