frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 12:54:11 +08:00
Code Issues Packages Projects Releases Wiki Activity

[torchfuzz] ones over zero (#164002)

Browse Source 
reduces likelihood of divide by zero errors. long term we'll probably want to just fuzz these values entirely

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164002
Approved by: https://github.com/pianpwk
ghstack dependencies: #163743, #163812, #163890
This commit is contained in:
bobrenjc93
2025-09-26 15:21:31 -07:00
committed by PyTorch MergeBot
parent b48a3d0a38
commit 3a115da3e6
2 changed files with 15 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
tools/experimental/dynamic_shapes/torchfuzz/operators/constant.py
View File

@ -76,19 +76,20 @@ class ConstantOperator(Operator):
import torch
default_values = {
torch.float16: 0.0,
torch.float32: 0.0,
torch.float64: 0.0,
torch.bfloat16: 0.0,
torch.int8: 0,
torch.int16: 0,
torch.int32: 0,
torch.int64: 0,
torch.bool: False,
torch.complex64: 0.0,
torch.complex128: 0.0,
torch.float16: 1.0,
torch.float32: 1.0,
torch.float64: 1.0,
torch.bfloat16: 1.0,
torch.int8: 1,
torch.int16: 1,
torch.int32: 1,
torch.int64: 1,
torch.bool: True,
torch.complex64: 1.0,
torch.complex128: 1.0,
}
fill_value = default_values.get(output_spec.dtype, 0)
fill_value = default_values.get(output_spec.dtype, 1)
tensor_creation = (
f"torch.full({size_str}, {fill_value}, dtype={dtype_str})"
)

4
tools/experimental/dynamic_shapes/torchfuzz/tensor_fuzzer.py
View File

@ -373,7 +373,7 @@ def fuzz_tensor(
# Handle empty tensor case
if len(size) == 0:
return torch.zeros((), dtype=dtype), seed
return torch.ones((), dtype=dtype), seed
# Calculate required storage size for the custom stride
required_storage = _compute_storage_size_needed(size, stride)
@ -400,7 +400,7 @@ def fuzz_tensor(
base_tensor = torch.randint(-100, 100, (required_storage,), dtype=dtype)
else:
# Use zeros (default behavior)
base_tensor = torch.zeros(required_storage, dtype=dtype)
base_tensor = torch.ones(required_storage, dtype=dtype)
# Create strided tensor view
strided_tensor = torch.as_strided(base_tensor, size, stride)