[REQUEST] Imporve compatibility with reasoning template And/Or Implement API response field "reasoning_content" like R1 API/llama.cpp

[Originalimoc]

Problem

The chat templates of R1/QwQ auto append a tag in front of generation to force thinking but this breaks some UI, eg. ggml-org/llama.cpp#11861

Solution

Detect the thinking template and return "" in advance of generated content to client.
For reference:
ggml-org/llama.cpp#11607
Solve #294 #299 too.

Alternatives

Hmm.. None?

Acknowledgements

• I have looked for similar requests before submitting this one.
• I understand that the developers have lives and my issue will be answered when possible.
• I understand the developers of this program are human, and I will make my requests politely.

[Originalimoc]

Continue discussion in #295