ZayaXMLToolParser.__init__() parameter mismatch when using tool calls with Zyphra/ZAYA1-8B
I am deploying Zyphra/ZAYA1-8B with vLLM and enabling auto tool calling. The server starts fine, but when my Agent makes a tool call request, the following error occurs:
TypeError: ZayaXMLToolParser.init() takes 2 positional arguments but 3 were given
File "/root/anaconda3/envs/zaya/lib/python3.13/site-packages/vllm/entrypoints/serve/render/serving.py", line 605, in preprocess_chat
request = tool_parser(tokenizer, request.tools).adjust_request(
~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: ZayaXMLToolParser.init() takes 2 positional arguments but 3 were given
Startup command:
export CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7
nohup vllm serve /data/models/ZAYA1-8B --port 8000
--served-model-name ZAYA1-8B
--mamba-cache-dtype float32
--dtype bfloat16
--reasoning-parser qwen3
--enable-auto-tool-choice
--max-model-len 131072
--data-parallel-size 8
--enable-expert-parallel
--max-num-seqs 5
--gpu-memory-utilization 0.92
--tool-call-parser zaya_xml &
