Some LLMs include the thinking process in their outputs. This feature is designed to structurally parse the output, separating the model's thinking process (think) from the final output (content) and storing them in the [object Object] and [object Object] fields, respectively.
[object Object]: stores the model's internal reasoning, analysis, and logic judgment before generating the final answer.[object Object]: stores the model's final output answer or decision.
- The Atlas 800I A2 inference server, Atlas 800I A3 SuperPoD server, and Atlas 300I Duo inference card support this feature.
- Currently, only the Qwen3-32B, Qwen3-235B-A22B, Qwen3-30B-A3B, DeepSeek-R1, and DeepSeek-V3.1 models support this feature.
- To enable reasoning analysis for the DeepSeek-V3.1 model, include the following field in the request:
[object Object], or add[object Object]to the[object Object]file. - Currently, only the OpenAI inference API is supported.
describes the parameters required for enabling the thinking analysis feature.
Table 1 [object Object] in [object Object] [object Object][object Object]
Open the
[object Object]file of the server.- Installation using the
[object Object]package:
[object Object]- Installation using the
[object Object]package:
[object Object]- Installation using the
Set serving parameters. Add the
[object Object]field to the[object Object]file of the server according to . For details about the serving parameters, see . The following is an example of parameter settings.The following uses Qwen3-32B as an example:
[object Object][object Object]
Start the service.
- Installation using the
[object Object]package:
[object Object]- Installation using the
[object Object]package:
[object Object]- Installation using the
Send a request. For details about the parameters, see "Cluster Management Components" > "Coordinator" > "RESTful APIs" > "User-Side APIs" > "OpenAI Inference APIs" in MindIE Motor Development Guide.