Model comparison
parakeet-tdt-0.6b-v3 vs nemotron-3.5-asr-streaming-0.6b
At Q4_K_M, parakeet-tdt-0.6b-v3 is the smaller download — 485,425,504 bytes against 495,831,520.
From the file· summed bytes, KV per layer
Side by side
| parakeet-tdt-0.6b-v3 | nemotron-3.5-asr-streaming-0.6b | |
|---|---|---|
| Parameters | 627M | 638M |
| Architecture | parakeet | parakeet |
| Layers | — | — |
| Native context | — | — |
| Mixture of experts | no | no |
| Quantizations published | 10 | 8 |
| Smallest quantization | 0.39 GiB | 0.38 GiB |
| Q4_K_M | 0.45 GiB | 0.46 GiB |
| Licence | cc-by-4.0 | other |
KV cache by context
the term that decides long-context viability
| Context | parakeet-tdt-0.6b-v3 | nemotron-3.5-asr-streaming-0.6b | Ratio |
|---|---|---|---|
| 4,096 | — | — | — |
| 8,192 | — | — | — |
| 16,384 | — | — | — |
| 32,768 | — | — | — |
| 65,536 | — | — | — |
| 131,072 | — | — | — |