Thomas Viehmann
Thomas Viehmann
When comparing Thunder Torch Executor to Torch Eager, the ResNet18 gradients are not close for FP32.
https://pytorch.org/docs/stable/generated/torch.autograd.gradcheck.gradcheck.html says > Note > The default values are designed for input of double precision. This check will likely fail if input is of less precision, e.g., FloatTensor. however, the...
To my mind, we should support direct indexing with a single `int | slice | Ellipsis | list[int] | Tensor | None` and with a `tuple[int | slice | Ellipsis...
> The clang indexing operations should (probably) be consistent with NumPy, but the torch indexing operations would, ideally, be consistent with PyTorch's behavior, and that may require clang indexing operations...
``` In [11]: implemented - tested_instructions Out[11]: {'LOAD_ASSERTION_ERROR', 'LOAD_METHOD'} In [12]: {*dis.opmap} - implemented Out[12]: {'CACHE', 'CALL_INTRINSIC_2', 'CHECK_EG_MATCH', 'CLEANUP_THROW', 'INSTRUMENTED_CALL', 'INSTRUMENTED_CALL_FUNCTION_EX', 'INSTRUMENTED_END_FOR', 'INSTRUMENTED_END_SEND', 'INSTRUMENTED_FOR_ITER', 'INSTRUMENTED_INSTRUCTION', 'INSTRUMENTED_JUMP_BACKWARD', 'INSTRUMENTED_JUMP_FORWARD', 'INSTRUMENTED_LINE', 'INSTRUMENTED_LOAD_SUPER_ATTR', 'INSTRUMENTED_POP_JUMP_IF_FALSE',...
`CLEANUP_THROW` is only seen in exception handling of yield_from, which would need improvments #815, `LOAD_ASSERTION_ERROR` is not seen in tests because PyTest rewriting asserts before execution of tests, `INTERPRETER_EXIT` is...
We even have 3.13 as of #1906
Note that #461 was about avoiding the inf recursion. This is from the dataclass decorator(?) setting `__dataclass_params__` on the `BaseModelOutput` class (to `_DataclassParams(init=True,repr=True,eq=True,order=False,unsafe_hash=False,frozen=False)`) The trouble very likely stems from either...
Note also that the above (the creation of the BaseModelOutput class) is triggered by the lazy loading of transformers leading to the importing of the `modeling_outputs` module being done by...
Here is an even more minimal repro (different error message due to different setattr method) ``` class A: pass def fn(x): A.x = x fn(1) # works as expected print(A.x)...
Nit: we probably should go to 2.5.1