benchmark 20260731_150000 router batch 400

benchmark-20260731_150000.yaml

@@ -0,0 +1,26 @@
+backend: router-batch-probe
+benchmark_id: hf-benchmark-run-20260731-150000
+completed_at: "2026-07-31T15:00:00Z"
+model: Nanthasit/router-400-batch-20260731
+models_probed:
+- Nanthasit/sakthai-plus-1.5b
+- Nanthasit/sakthai-coder-browser
+- Nanthasit/sakthai-context-1.5b-merged-v2
+- Nanthasit/sakthai-context-0.5b-tools-sft
+- Nanthasit/sakthai-context-0.5b-tools-grpo-pilot-lora16
+- Nanthasit/sakthai-plus-1.5b-lora
+- Nanthasit/sakthai-plus-1.5b-coder
+- Nanthasit/sakthai-context-7b-tools-grpo
+notes: >-
+  Router probe sweep. api-inference.huggingface.co is decommissioned (NXDOMAIN).
+  All router probes returned HTTP 400 Bad Request. Local inference backend is not
+  available in this environment: llama.cpp is missing and transformers/torch CPU
+  stack is not installed. No executable benchmark runnable this turn.
+run_id: 20260731_150000
+scores:
+  router_400: true
+status: 400
+summary: >-
+  Confirmed router does not support these Nanthasit text-generation models.
+  No backend available to run inference locally, so no new benchmarkable run
+  produced this turn.