Issue Station
Sign In
Tools hereToolbox

pytorch - 💡(How to fix) Fix DTensor native sharding propagation cache can reuse stale SymInt OutputSharding across compiled_autograd ShapeEnv boundaries

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…

Error Message

[rank0]: Traceback (most recent call last): [rank0]: File "/torch/src/a.py", line 53, in <module> [rank0]: run_case() [rank0]: File "/torch/src/a.py", line 48, in run_case [rank0]: res.sum().backward() [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_tensor.py", line 631, in backward [rank0]: torch.autograd.backward( [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/autograd/init.py", line 381, in backward [rank0]: _engine_run_backward( [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/autograd/graph.py", line 869, in _engine_run_backward [rank0]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/compiled_autograd.py", line 1145, in runtime_wrapper [rank0]: out = compiled_fn( [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 472, in call [rank0]: return super().call(*args, **kwargs) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1779, in _wrapped_call_impl [rank0]: return self._call_impl(*args, **kwargs) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1790, in _call_impl [rank0]: return forward_call(*args, **kwargs) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 1024, in compile_wrapper [rank0]: return fn(*args, **kwargs) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 949, in call_wrapped [rank0]: return self._wrapped_call(self, *args, **kwargs) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 461, in call [rank0]: raise e [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 447, in call [rank0]: return super(self.cls, obj).call(*args, **kwargs) # type: ignore[misc] [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1779, in _wrapped_call_impl [rank0]: return self._call_impl(*args, **kwargs) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1790, in _call_impl [rank0]: return forward_call(*args, **kwargs) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 2316, in call [rank0]: result = self._torchdynamo_orig_backend( [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 729, in call [rank0]: result = _compile( [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1827, in _compile [rank0]: guarded_code, tracer_output = compile_inner(code, one_graph, hooks) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_utils_internal.py", line 96, in wrapper_function [rank0]: return function(*args, **kwargs) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1500, in compile_inner [rank0]: return _compile_inner(code, one_graph, hooks) [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1634, in _compile_inner [rank0]: check_fn = dynamo_output.build_guards( [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 904, in build_guards [rank0]: return CheckFunctionManager( [rank0]: File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/guards.py", line 3928, in init [rank0]: raise AssertionError( [rank0]: AssertionError: Guard failed on the same frame it was created. This is a bug - please create an issue.Guard fail reason: !1/2/0: sizes[0].size()[1] == inputs[1].size()[0] # sum_backward0 = torch__dynamo_compiled_autograd_ops_SumBackward0([getitem_10], [True], [unwrap_maybe_dynamic_int]); getitem_10 = unwrap_maybe_dynamic_int = None # <eval_with_key>.9:16 in forward (_dynamo/compiled_autograd.py:251 in call)

Fix Action

Fix / Workaround

CPU: Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 46 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 112 On-line CPU(s) list: 0-111 Vendor ID: GenuineIntel Model name: Intel(R) Xeon(R) Gold 6330 CPU @ 2.00GHz CPU family: 6 Model: 106 Thread(s) per core: 2 Core(s) per socket: 28 Socket(s): 2 Stepping: 6 Frequency boost: enabled CPU max MHz: 3100.0000 CPU min MHz: 800.0000 BogoMIPS: 4000.00 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 invpcid_single ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local wbnoinvd dtherm ida arat pln pts avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq rdpid md_clear pconfig flush_l1d arch_capabilities Virtualization: VT-x L1d cache: 2.6 MiB (56 instances) L1i cache: 1.8 MiB (56 instances) L2 cache: 70 MiB (56 instances) L3 cache: 84 MiB (2 instances) NUMA node(s): 2 NUMA node0 CPU(s): 0-27,56-83 NUMA node1 CPU(s): 28-55,84-111 Vulnerability Gather data sampling: Mitigation; Microcode Vulnerability Itlb multihit: Not affected Vulnerability L1tf: Not affected Vulnerability Mds: Not affected Vulnerability Meltdown: Not affected Vulnerability Mmio stale data: Mitigation; Clear CPU buffers; SMT vulnerable Vulnerability Retbleed: Not affected Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization Vulnerability Spectre v2: Mitigation; Enhanced IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected

Code Example

import torch
import torch._inductor as inductor
import torch.distributed as dist
from torch._dynamo import compiled_autograd
from torch.distributed.tensor import DeviceMesh, DTensor, Shard
from torch.testing._internal.distributed.fake_pg import FakeStore


def ca_compiler(gm):
    def inner(gm_, example_inputs_):
        return inductor.compile(gm_, example_inputs_)

    return torch.compile(gm, backend=inner, fullgraph=True, dynamic=True)


def run_case():
    torch._dynamo.reset()

    with compiled_autograd._enable(ca_compiler):
        mesh = DeviceMesh("cpu", torch.arange(2))

        def fn(x, y):
            out = x.sin()
            y.add_(2)
            return out

        opt_fn = torch.compile(fn, backend="aot_eager", fullgraph=True)

        x_ref = DTensor.from_local(
            torch.randn(4), mesh, [Shard(0)], run_check=False
        ).requires_grad_(True)
        y_ref = DTensor.from_local(
            torch.randn(4), mesh, [Shard(0)], run_check=False
        ).requires_grad_(False)

        x = x_ref.clone().detach().requires_grad_(True)
        y = y_ref.clone().detach().requires_grad_(False)

        ref = fn(x_ref.clone(), y_ref)
        res = opt_fn(x.clone(), y)

        torch.testing.assert_close(res, ref)

        ref.sum().backward()
        res.sum().backward()


dist.init_process_group("fake", store=FakeStore(), rank=0, world_size=2)
try:
    run_case()
    run_case()
finally:
    dist.destroy_process_group()

---

[rank0]: Traceback (most recent call last):
[rank0]:   File "/torch/src/a.py", line 53, in <module>
[rank0]:     run_case()
[rank0]:   File "/torch/src/a.py", line 48, in run_case
[rank0]:     res.sum().backward()
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_tensor.py", line 631, in backward
[rank0]:     torch.autograd.backward(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/autograd/__init__.py", line 381, in backward
[rank0]:     _engine_run_backward(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/autograd/graph.py", line 869, in _engine_run_backward
[rank0]:     return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/compiled_autograd.py", line 1145, in runtime_wrapper
[rank0]:     out = compiled_fn(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 472, in __call__
[rank0]:     return super().__call__(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1779, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1790, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 1024, in compile_wrapper
[rank0]:     return fn(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 949, in call_wrapped
[rank0]:     return self._wrapped_call(self, *args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 461, in __call__
[rank0]:     raise e
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 447, in __call__
[rank0]:     return super(self.cls, obj).__call__(*args, **kwargs)  # type: ignore[misc]
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1779, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1790, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 2316, in __call__
[rank0]:     result = self._torchdynamo_orig_backend(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 729, in __call__
[rank0]:     result = _compile(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1827, in _compile
[rank0]:     guarded_code, tracer_output = compile_inner(code, one_graph, hooks)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_utils_internal.py", line 96, in wrapper_function
[rank0]:     return function(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1500, in compile_inner
[rank0]:     return _compile_inner(code, one_graph, hooks)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1634, in _compile_inner
[rank0]:     check_fn = dynamo_output.build_guards(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 904, in build_guards
[rank0]:     return CheckFunctionManager(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/guards.py", line 3928, in __init__
[rank0]:     raise AssertionError(
[rank0]: AssertionError: Guard failed on the same frame it was created. This is a bug - please create an issue.Guard fail reason: !1/2/0: sizes[0].size()[1] == inputs[1].size()[0]  # sum_backward0 = torch__dynamo_compiled_autograd_ops_SumBackward0([getitem_10], [True], [unwrap_maybe_dynamic_int]);  getitem_10 = unwrap_maybe_dynamic_int = None  # <eval_with_key>.9:16 in forward (_dynamo/compiled_autograd.py:251 in __call__)
RAW_BUFFERClick to expand / collapse

🐛 Describe the bug

I can reproduce a pure PyTorch issue involving DTensor, torch.compile, and compiled autograd. This does not require any third-party backend. The repro works on CPU with torch==2.11.0+cpu.

The failure happens when a sharded DTensor participates in a compiled forward, followed by compiled-autograd backward. A stale DTensor native sharding propagation cache entry appears to be reused across compiled-autograd ShapeEnv boundaries. This can incorrectly equate a DTensor global size with a local shard size.

Minimal repro:


import torch
import torch._inductor as inductor
import torch.distributed as dist
from torch._dynamo import compiled_autograd
from torch.distributed.tensor import DeviceMesh, DTensor, Shard
from torch.testing._internal.distributed.fake_pg import FakeStore


def ca_compiler(gm):
    def inner(gm_, example_inputs_):
        return inductor.compile(gm_, example_inputs_)

    return torch.compile(gm, backend=inner, fullgraph=True, dynamic=True)


def run_case():
    torch._dynamo.reset()

    with compiled_autograd._enable(ca_compiler):
        mesh = DeviceMesh("cpu", torch.arange(2))

        def fn(x, y):
            out = x.sin()
            y.add_(2)
            return out

        opt_fn = torch.compile(fn, backend="aot_eager", fullgraph=True)

        x_ref = DTensor.from_local(
            torch.randn(4), mesh, [Shard(0)], run_check=False
        ).requires_grad_(True)
        y_ref = DTensor.from_local(
            torch.randn(4), mesh, [Shard(0)], run_check=False
        ).requires_grad_(False)

        x = x_ref.clone().detach().requires_grad_(True)
        y = y_ref.clone().detach().requires_grad_(False)

        ref = fn(x_ref.clone(), y_ref)
        res = opt_fn(x.clone(), y)

        torch.testing.assert_close(res, ref)

        ref.sum().backward()
        res.sum().backward()


dist.init_process_group("fake", store=FakeStore(), rank=0, world_size=2)
try:
    run_case()
    run_case()
finally:
    dist.destroy_process_group()

Actual result:

Sometimes the first compiled backward already fails after ref.sum().backward() has populated the DTensor sharding cache. In other runs, the first run_case() passes and the second one fails after re-entering compiled autograd.

Observed errors include:

[rank0]: Traceback (most recent call last):
[rank0]:   File "/torch/src/a.py", line 53, in <module>
[rank0]:     run_case()
[rank0]:   File "/torch/src/a.py", line 48, in run_case
[rank0]:     res.sum().backward()
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_tensor.py", line 631, in backward
[rank0]:     torch.autograd.backward(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/autograd/__init__.py", line 381, in backward
[rank0]:     _engine_run_backward(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/autograd/graph.py", line 869, in _engine_run_backward
[rank0]:     return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/compiled_autograd.py", line 1145, in runtime_wrapper
[rank0]:     out = compiled_fn(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 472, in __call__
[rank0]:     return super().__call__(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1779, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1790, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 1024, in compile_wrapper
[rank0]:     return fn(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 949, in call_wrapped
[rank0]:     return self._wrapped_call(self, *args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 461, in __call__
[rank0]:     raise e
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/fx/graph_module.py", line 447, in __call__
[rank0]:     return super(self.cls, obj).__call__(*args, **kwargs)  # type: ignore[misc]
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1779, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1790, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 2316, in __call__
[rank0]:     result = self._torchdynamo_orig_backend(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 729, in __call__
[rank0]:     result = _compile(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1827, in _compile
[rank0]:     guarded_code, tracer_output = compile_inner(code, one_graph, hooks)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_utils_internal.py", line 96, in wrapper_function
[rank0]:     return function(*args, **kwargs)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1500, in compile_inner
[rank0]:     return _compile_inner(code, one_graph, hooks)
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 1634, in _compile_inner
[rank0]:     check_fn = dynamo_output.build_guards(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 904, in build_guards
[rank0]:     return CheckFunctionManager(
[rank0]:   File "/torch/venv3/pytorch/lib/python3.10/site-packages/torch/_dynamo/guards.py", line 3928, in __init__
[rank0]:     raise AssertionError(
[rank0]: AssertionError: Guard failed on the same frame it was created. This is a bug - please create an issue.Guard fail reason: !1/2/0: sizes[0].size()[1] == inputs[1].size()[0]  # sum_backward0 = torch__dynamo_compiled_autograd_ops_SumBackward0([getitem_10], [True], [unwrap_maybe_dynamic_int]);  getitem_10 = unwrap_maybe_dynamic_int = None  # <eval_with_key>.9:16 in forward (_dynamo/compiled_autograd.py:251 in __call__)

Versions

PyTorch version: 2.11.0+cpu Is debug build: False CUDA used to build PyTorch: None ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.5 LTS (x86_64) GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 Clang version: 14.0.0-1ubuntu1.1 CMake version: version 4.3.1 Libc version: glibc-2.35

Python version: 3.10.12 (main, May 27 2025, 17:12:29) [GCC 11.4.0] (64-bit runtime) Python platform: Linux-5.4.0-172-generic-x86_64-with-glibc2.35 Is CUDA available: False CUDA runtime version: No CUDA CUDA_MODULE_LOADING set to: N/A GPU models and configuration: No CUDA Nvidia driver version: No CUDA cuDNN version: No CUDA Is XPU available: False HIP runtime version: N/A MIOpen runtime version: N/A Is XNNPACK available: True Caching allocator config: N/A

CPU: Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 46 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 112 On-line CPU(s) list: 0-111 Vendor ID: GenuineIntel Model name: Intel(R) Xeon(R) Gold 6330 CPU @ 2.00GHz CPU family: 6 Model: 106 Thread(s) per core: 2 Core(s) per socket: 28 Socket(s): 2 Stepping: 6 Frequency boost: enabled CPU max MHz: 3100.0000 CPU min MHz: 800.0000 BogoMIPS: 4000.00 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 invpcid_single ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local wbnoinvd dtherm ida arat pln pts avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq rdpid md_clear pconfig flush_l1d arch_capabilities Virtualization: VT-x L1d cache: 2.6 MiB (56 instances) L1i cache: 1.8 MiB (56 instances) L2 cache: 70 MiB (56 instances) L3 cache: 84 MiB (2 instances) NUMA node(s): 2 NUMA node0 CPU(s): 0-27,56-83 NUMA node1 CPU(s): 28-55,84-111 Vulnerability Gather data sampling: Mitigation; Microcode Vulnerability Itlb multihit: Not affected Vulnerability L1tf: Not affected Vulnerability Mds: Not affected Vulnerability Meltdown: Not affected Vulnerability Mmio stale data: Mitigation; Clear CPU buffers; SMT vulnerable Vulnerability Retbleed: Not affected Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization Vulnerability Spectre v2: Mitigation; Enhanced IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected

Versions of relevant libraries: [pip3] mypy_extensions==1.1.0 [pip3] numpy==1.26.4 [pip3] optree==0.19.0 [pip3] torch==2.11.0+cpu [pip3] torch_mlu==1.32.0+torch2.11.0.1b645a [pip3] torchaudio==2.11.0+cpu [pip3] torchvision==0.26.0+cpu

cc @awgu @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @pragupta @msaroufim @dcci @aditvenk @weifengpy @tianyu-l @XilunWu @SherlockNoMad @ppwwyyxx

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#api#task chaining#parallel task#integration issue#index setup

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

pytorch - 💡(How to fix) Fix DTensor native sharding propagation cache can reuse stale SymInt OutputSharding across compiled_autograd ShapeEnv boundaries