$ the-wire · showcase
PyTorch's Caffe2 cleanup hits a wall as internal dependencies force three reverts
By RepoJournal · Filed · About PyTorch · Composed from the cited sources · methodology
PyTorch's attempt to shed its Caffe2 legacy just got blocked by internal dependencies, and the resulting reverts ripple across the cleanup effort.
Three reverts in a row [1][2][3] pulled back placeholder removals, with the merge bot citing that "lots of internal code still depends on this header" [1] and similar issues for timer.h [2]. Meanwhile, an Inductor patch tightens the serialization boundary , a fake tensor caching fix targets 325 uncached dispatches on Qwen3-8B [4], TensorRT adds a pin-drift detector across eleven files [5], and a probe loop fix handles a c10::Error change in PyTorch 2.13 [6]. Torchtitan flags that zero-valid-token RL batches still step the optimizer [7], and adds validation for MRoPE and activation memory budgets [8][9]. The Caffe2 reverts need coordination, but the core caching and validation fixes are safe to take.
Action items
- → Check for internal references to caffe2/core/macros.h or timer.h before any cleanup PRs pytorch/pytorch [immediate]
- → Review the fake tensor cache fix for performance gains on large prefill traces pytorch/pytorch [plan]
- → Update TensorRT to handle c10::Error in bind_nccl_comm() for PyTorch >= 2.13 pytorch/TensorRT [immediate]
- → Audit RL training loops for zero-valid-token batches that skip optimizer steps pytorch/torchtitan [immediate]
References
- [1] Revert "Remove the caffe2/core/macros.h placeholder (#194230)" pytorch/pytorch ↗
- [2] Revert "Remove caffe2/core/timer.h (#194228)" pytorch/pytorch ↗
- [3] Revert "Remove unused caffe2/utils/string_utils (#185125)" pytorch/pytorch ↗
- [4] [fake tensor] Cache _unsafe_view dispatch (#193765) pytorch/pytorch ↗
- [5] Detect ExecuTorch pin drift across the repository ↗ pytorch/TensorRT
- [6] [fix] MD-TRT: handle c10::Error in bind_nccl_comm() probe loop ↗ pytorch/TensorRT
- [7] [Bug][RL] Zero-valid-token batches still step optimizer and advance policy version ↗ pytorch/torchtitan
- [8] [Bug] Validate MRoPE dimension contracts ↗ pytorch/torchtitan
- [9] [Bug] Validate activation memory budget (#4291) ↗ pytorch/torchtitan
