llama.cpp on a jailbroken PS5: run any GGUF model natively on the console, starting with Qwen 3.8
App: ready only once the server serves metrics, not its 503 while the model loads
Mert Cobanov760bbf8209e8
diff --git a/ps5/app/stats.cpp b/ps5/app/stats.cppindex e6bee86..56d2cf2 100644--- a/ps5/app/stats.cpp+++ b/ps5/app/stats.cpp@@ -340,7 +340,9 @@ void StatsCollector::run() {const double wall = now_s(), cpu = cpu_seconds();const std::string m = http_get(port_, "/metrics");- const bool up = !m.empty();+ // Only real metrics: while the model loads the server already answers,+ // with 503 "Loading model", and that is not ready.+ const bool up = m.find("llamacpp:") != std::string::npos;const double busy_s = metric(m, "prompt_seconds_total") + metric(m, "tokens_predicted_seconds_total");const double tokens = metric(m, "tokens_predicted_total");const double gen_s = metric(m, "tokens_predicted_seconds_total");