Skip to content

PyTorch 2.1.1 Release, bug fix release
Compare
Choose a tag to compare
@jerryzh168 jerryzh168 released this 15 Nov 22:59
· 9458 commits to main since this release
v2.1.1
4c55dc5
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.

This release is meant to fix the following issues (regressions / silent correctness):

  • Remove spurious warning in comparison ops (#112170)
  • Fix segfault in foreach_* operations when input list length does not match (#112349)
  • Fix cuda driver API to load the appropriate .so file (#112996)
  • Fix missing CUDA initialization when calling FFT operations (#110326)
  • Ignore beartype==0.16.0 within the onnx package as it is incompatible (#111861)
  • Fix the behavior of torch.new_zeros in onnx due to TorchScript behavior change (#111694)
  • Remove unnecessary slow code in torch.distributed.checkpoint.optimizer.load_sharded_optimizer_state_dict (#111687)
  • Add planner argument to torch.distributed.checkpoint.optimizer.load_sharded_optimizer_state_dict (#111393)
  • Continue if param not exist in sharded load in torch.distributed.FSDP (#109116)
  • Fix handling of non-contiguous bias_mask in torch.nn.functional.scaled_dot_product_attention (#112673)
  • Fix the meta device implementation for nn.functional.scaled_dot_product_attention (#110893)
  • Fix copy from mps to cpu device when storage_offset is non-zero (#109557)
  • Fix segfault in torch.sparse.mm for non-contiguous inputs (#111742)
  • Fix circular import between Dynamo and einops (#110575)
  • Verify flatbuffer module fields are initialized for mobile deserialization (#109794)

The #110961 contains all relevant pull requests related to this release as well as links to related issues.

Assets 3