Attn: Cerebras Any attempt to deal with "<think>" in the code gets it replaced with "<tool_call>". Both in inference.cerebras.ai chat and API. Same model on

Attn: Cerebras Any attempt to deal with "<think>" in the code gets it replaced with "<tool_call>". Both in inference.cerebras.ai chat and API. Same model on chat.qwen.ai doesn't do it.