Qwen3.5-9B bounded adapter training — peak dedicated GPU
Qwen3.5-9B bounded adapter training — peak dedicated GPU: 9.78 GiB on a 12 GB card.
Single reading
10,500,022,272 bytes
Protocol
Test definition
GPU memory telemetry from the project's watcher / allocator counters. 16 updates, 990 s wall; no host fallback.
Limits
Limitations
- Sampled gauges; sub-sample transients not exhaustively proven. Physical card is 12,227 MiB; allocator and dedicated readings are different counters.
Log
Recorded values
| Date | Value | Evidence | Samples | Source |
|---|---|---|---|---|
| 21 September 2026 | 10,500,022,272 bytes | MEASURED | — | docs/ledger/FINAL_LM_QWEN35_BOUNDED_ADAPTER_RESULT.json |
Source paths refer to the private JARVIS repository and runtime records. They are listed for audit and are not published.