CLM Playground

Mock encoder. This server is running tools/playground_mock.py, which replaces Qwen3-8B with character n-grams — the numbers below are lexical-overlap noise, not CLM predictions.

GPU encoder is not running. Start it and this page reconnects on its own:

vllm serve Qwen/Qwen3-8B --served-model-name qwen3-8b --runner pooling \
     --enable-prefix-caching --max-model-len 2048 --gpu-memory-utilization 0.35 --port 8090

State

JSON 0 chars

Questions one call answers all of them against the same state

Model

Answers appear here.

Press ⌘↵ to run, or load an example from the menu above.