2026-10-07T19:28:58Z warm run 1: 1.27 s to ready, 9107.7 MB/s effective (llama-server -m ~/models/gguf/gpt-oss-20b-MXFP4.gguf --port 8082 --jinja -c 32768 --alias gpt-oss-20b)
2026-10-07T19:28:59Z warm run 2: 1.20 s to ready, 9623.8 MB/s effective (llama-server -m ~/models/gguf/gpt-oss-20b-MXFP4.gguf --port 8082 --jinja -c 32768 --alias gpt-oss-20b)
2026-10-07T19:29:00Z warm run 3: 1.21 s to ready, 9512.8 MB/s effective (llama-server -m ~/models/gguf/gpt-oss-20b-MXFP4.gguf --port 8082 --jinja -c 32768 --alias gpt-oss-20b)
