🔌
New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
// RUN — FITS
Can the Jetson Thor T2000 run CodeT5-base?
Yes — here's how fast and up to what context, with real measured numbers.
CodeT5-base on the Jetson Thor T2000
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 190k | 227.3-270.9 t/s | ~16.6-19.8 (slow) |
| Q8 | yes | 193k | 454.6-541.8 t/s | ~33.2-39.5 (slow) |
| Q4 | yes | 195k | 811.8-967.5 t/s | ~59.3-70.6 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
FAQ
Can the Jetson Thor T2000 run CodeT5-base?▶
Yes. At Q4 it generates ~811.8-967.5 tok/s and fits up to 195k context; beyond that it spills to SSD swap and slows to ~59.3-70.6 tok/s.
How much context fits?▶
At Q4, up to about 195k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.