+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Security, Privacy & Safety (https://www.datasculpting.ai/inference/forumdisplay.php?fid=23)
+--- Forum: Model / Agent Security (https://www.datasculpting.ai/inference/forumdisplay.php?fid=77)
+--- Thread: sandboxing agent actions (/showthread.php?tid=1311)
sandboxing agent actions - flintriver50 - 10-16-2026
Posting this in case it saves someone else a little time.
I'm helping a small team set up a local workflow. We don't need a giant stack, but we do care about token cost.
Would rather learn from your scars than collect my own all at once.
Re: sandboxing agent actions - cacheline - 10-16-2026
worth checking the driver first.
Re: sandboxing agent actions - mintquill - 10-18-2026
I tested this tonight. On dual 3090 with MLX, I saw basically the same behavior. The surprise was that memory pressure showed up before throughput collapsed.
Re: sandboxing agent actions - stackpilot - 10-18-2026
What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at token cost. Once you know the plain path works, then add the faster path back in piece by piece.
Re: sandboxing agent actions - cudacrate - 10-18-2026
Quote:What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at token cost. Once you
I'd approach it in a boring order. On 7900 XTX, the easiest win is often to back off the aggressive settings and see whether the problem survives. Not glamorous advice, but repeatable baselines save a lot of time.
Re: sandboxing agent actions - turboscope5 - 10-19-2026
Quote:[quote]What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at token cost. On
I don't think that's the whole story. The reported numbers can look compelling, but in ordinary usage I didn't see the same benefit. Not saying you're wrong, just that I wouldn't lock that in as the lesson yet.