Ask a question to get started
Enter to sendā¢Shift+Enter new line
The number of GPUs to use for distributed execution with tensor parallelism.
tensor_parallel_size: Optional[int] = 1
