A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
Add documentation for granite 4 tool calling (#26175)
Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
Maximilien de Bayser committed
Oct 5, 2025 at 04:35 UTC
e0986ea07bd7f1f01cd3c54d91523aea91589b59
1 file changed
+5
-1
docs/features/tool_calling.md
+5
-1
@@ -191,10 +191,14 @@ VLLM also provides a pythonic and JSON-based chat template for Llama 4, but pyth
191
192
For Llama 4 model, use `--tool-call-parser llama4_pythonic --chat-template examples/tool_chat_template_llama4_pythonic.jinja`.
193
194
-#### IBM Granite
194
+### IBM Granite
195
196
Supported models:
197
198
+* `ibm-granite/granite-4.0-h-small` and other Granite 4.0 models
199
+
200
+ Recommended flags: `--tool-call-parser hermes`
201
+
202
* `ibm-granite/granite-3.0-8b-instruct`
203
204
Recommended flags: `--tool-call-parser granite --chat-template examples/tool_chat_template_granite.jinja`