Update MTP recipe: HTTP server example + vLLM caveat

README.md

@@ -167,13 +167,28 @@ out = target.generate(
 print(tok.decode(out[0][ids.shape[1]:], skip_special_tokens=True))
 ```
 
-For production throughput, use vLLM — it implements the drafter's centroid-masking optimization (sparse lm_head over ~4K candidates instead of ~262K vocab, ~45x reduction):
+For a self-hosted OpenAI-compatible HTTP endpoint, wrap the pair in a small FastAPI server that holds both models resident and exposes `/v1/chat/completions`. Example: [`scripts/08_mtp_server.py`](scripts/08_mtp_server.py) in the source repo, callable as:
+
+```bash
+curl http://localhost:8765/v1/chat/completions -d '{
+  "model": "igorls/gemma4-e4b-classifier",
+  "messages": [{"role":"user","content":"What is the capital of France?"}],
+  "max_tokens": 16,
+  "use_mtp": true
+}'
+```
+
+### vLLM (future)
+
+vLLM is the right inference stack for production throughput — it implements the drafter's centroid-masking optimization (sparse lm_head over ~4K candidates instead of ~262K vocab, ~45x reduction in lm_head compute):
 
 ```bash
 vllm serve igorls/gemma4-e4b-classifier \
   --speculative-config '{"model": "google/gemma-4-E4B-it-assistant", "num_speculative_tokens": 4}'
 ```
 
+**However**, as of May 2026 (vLLM 0.20.2, latest on PyPI), this fails: the drafter's `Gemma4AssistantConfig` is not yet registered in vLLM's `AutoModel` mapping. The vLLM Gemma 4 recipes page documents the feature but it's ahead of the released version. Track [vllm-project/vllm](https://github.com/vllm-project/vllm/) for the release that lands `Gemma4Assistant` support; once available, the command above should work as-is against this model.
+
 ## License
 
 Inherited from the base model: [Gemma Terms of Use](https://ai.google.dev/gemma/terms). By using this model you agree to those terms.