card: measured numbers on 2x RTX 3090 + config note

README.md

@@ -48,12 +48,27 @@ vllm serve /models/Nex-N2.5-mini-Int4-Int8Mix \
   --default-chat-template-kwargs '{"enable_thinking": false}'
 ```
 
-Measured numbers from that box are in the repo README once the lane is up (KV pool, count-to-100 and prose tok/s). This checkpoint has no MTP head, so no speculative decoding.
+Measured on that box (2026-09-08, single stream, temperature 0, thinking off, no draft):
+
+| | |
+|---|---|
+| Weights per card | 9.98 GiB (Marlin WNA16 kernels) |
+| KV pool | **1,867,521 tokens** at 262,144 context, 7.12x concurrency (fp8 e5m2 KV; 10 attention layers, 30 Gated DeltaNet layers) |
+| Count to 100 | 185 tok/s, correct 1..100 |
+| Prose (450 tokens) | 187 tok/s |
+| Code (450 tokens) | 186 tok/s |
+| Load | 27 s weights, 67 s torch.compile, serving in about 4 minutes |
+
+This checkpoint has no MTP head, so no speculative decoding. Recipe, launcher, logs and the boot ladder: https://github.com/tonyd2wild/Nex-N2.5-mini-Int4-Int8Mix-2x-RTX-3090
+
+### Config note (fixed 2026-09-08)
+
+The `quantization_config` targets and ignore patterns are written as `re:.*layers[.]...` on purpose. vLLM matches quantization targets against its own module names (`language_model.model.layers.N...`), not the checkpoint names (`model.language_model.layers.N...`). The first upload pinned the patterns to the checkpoint prefix; vLLM then matched nothing, built the experts unquantized in fp16 and ran out of memory before loading a weight. The prefix-agnostic patterns match both namings, so the checkpoint loads in vLLM and in transformers.
 
 ## Caveats
 
 - Data-free INT4 is a size play, not an accuracy claim; run your own evals for your workload.
-- Quantized with our shard-streaming quantizer (`nex_quant_stream.py`, adapted from the GLM-5.3 Int4-Int8Mix recipe at https://github.com/tonyd2wild/GLM-5.3-Int4-Int8Mix-TP4-4x-DGX-Spark).
+- Quantized with our shard-streaming quantizer (`quant/nex_quant_stream.py` in the recipe repo, adapted from the GLM-5.3 Int4-Int8Mix recipe at https://github.com/tonyd2wild/GLM-5.3-Int4-Int8Mix-TP4-4x-DGX-Spark).
 
 ## Credits