🔌
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 Apple M2 Max 32GB run ProBERTa?
Yes — here's how fast and up to what context, with real measured numbers.
ProBERTa on the Apple M2 Max 32GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 415k | 3090.9-3727.3 t/s | ~77.3-93.2 (slow) |
| Q8 | yes | 416k | 6181.8-7454.5 t/s | ~154.5-186.4 (slow) |
| Q4 | yes | 416k | 11039-13311.7 t/s | ~276-332.8 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
FAQ
Can the Apple M2 Max 32GB run ProBERTa?▶
Yes. At Q4 it generates ~11039-13311.7 tok/s and fits up to 416k context; beyond that it spills to SSD swap and slows to ~276-332.8 tok/s.
How much context fits?▶
At Q4, up to about 416k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.