dsh-live-perf-gauges

中文

An automotive-style live performance strip for DeepSeek Harness Web: decode throughput, TTFT, prefill throughput, and output-token progress under the composer.

What it measures

• Decode tok/s refreshes every 250 ms during generation. It is explicitly marked ~ while it is inferred from streamed characters; it switches to exact provider output tokens when the step finishes.
• TTFT is the elapsed time from step/start to the first streamed text, reasoning, or tool-argument delta.
• Prefill tok/s is prompt tokens divided by TTFT. It is estimated from DSH's context-pressure projection while a request is live, then corrected with provider-reported prompt tokens (input + cache read + cache write) at completion.
• Output follows the same estimate-then-exact rule.

The plugin never contacts a network service and never stores message text, prompts, tools, or credentials. It only projects timestamps, counts, and final token usage already present in the session event stream.

Install

dsh plugin --profile web add github:TnzGit/dsh-live-perf-gauges

Restart the running dsh web service, then hard-refresh the browser. The strip appears below the composer after the next model step begins.

Compatibility

Designed for the DSH 0.1.x Web profile and its standard conversation.composer.dock, sessionProjections, and contextPressure interfaces.

Development

npm test

This repository ships prebuilt dependency-free host and client entries, so users do not need an install-time build step.

License

MIT