main
Add support for request rescheduling (#319)
Auto-format (#261)
Add Correlation Id string support for BLS (#344)
Fix error handling for GPU tensors (#249)
Fix the lifetime of InferPayload (#241)
Exposing trace context to python backend (#346)
Fix warning for GPU tensors (#330)
Fix GPU buffers when dynamic batching is enabled (#199)
Update name of ipc message type (#329)
Add the new build (#1)
Optimize GPU tensor support for Python backend (#293)
Add GitHub action to format and lint code (#265)
Model loading API for Python BLS (#262)
Python Backend Windows Support (#294)
Add Python backend request cancellation (#304)
Enable Python execute() to return Triton error code (#292)
Add new shared memory manager (#128)
Enhancement for BLS decoupled support (#208)
Add request statistic reporting for decoupled mode (#163)
Fix segfault for decoupled models (#327)
Fix response iterator memory leak (#302)
Decoupled Async Execute (#350)
Randomize Python backend shared memory region naming (#351)