Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

hf_T5_generate inference accuracy fail_to_run #1392

Open
kaileiyx opened this issue Feb 21, 2025 · 0 comments
Open

hf_T5_generate inference accuracy fail_to_run #1392

kaileiyx opened this issue Feb 21, 2025 · 0 comments
Assignees
Labels
Milestone

Comments

@kaileiyx
Copy link

🐛 Describe the bug

xpu eval hf_T5_generate
ERROR:common:
Traceback (most recent call last):
File "/home/sdp/actions-runner/_work/torch-xpu-ops/pytorch/benchmarks/dynamo/common.py", line 2256, in check_accuracy
new_result = self.run_n_iterations(
File "/home/sdp/actions-runner/_work/torch-xpu-ops/pytorch/benchmarks/dynamo/common.py", line 1943, in run_n_iterations
model_iter_fn(mod, inputs, collect_outputs=False)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 589, in _fn
raise e.remove_dynamo_frames() from None # see TORCHDYNAMO_VERBOSE=1
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/compile_fx.py", line 754, in _compile_fx_inner
raise InductorError(e, currentframe()).with_traceback(
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/compile_fx.py", line 739, in _compile_fx_inner
mb_compiled_graph = fx_codegen_and_compile(
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/compile_fx.py", line 1407, in fx_codegen_and_compile
return scheme.codegen_and_compile(gm, example_inputs, inputs_to_check, graph_kwargs)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/compile_fx.py", line 1127, in codegen_and_compile
compiled_fn = graph.compile_to_module().call
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/graph.py", line 1989, in compile_to_module
return self._compile_to_module()
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/graph.py", line 1995, in _compile_to_module
self.codegen_with_cpp_wrapper() if self.cpp_wrapper else self.codegen()
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/graph.py", line 1904, in codegen
self._update_scheduler()
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/graph.py", line 1898, in _update_scheduler
self.scheduler = Scheduler(self.operations)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/scheduler.py", line 1947, in init
self._init(nodes)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/scheduler.py", line 1965, in _init
self.nodes = [self.create_scheduler_node(n) for n in nodes]
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/scheduler.py", line 1965, in
self.nodes = [self.create_scheduler_node(n) for n in nodes]
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/scheduler.py", line 2102, in create_scheduler_node
return SchedulerNode(self, node)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/scheduler.py", line 984, in init
self._compute_attrs()
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/scheduler.py", line 992, in _compute_attrs
self._sizes, self._body = self.node.simplify_and_reorder(
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/ir.py", line 4108, in simplify_and_reorder
) = self.get_default_sizes_body()
File "", line 6, in get_default_sizes_body_cache_on_self
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/ir.py", line 4061, in get_default_sizes_body
body = LoopBody(
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/loop_body.py", line 117, in init
self._init_with_tracing(fn, args)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/loop_body.py", line 131, in _init_with_tracing
self.root_block = LoopBodyBlock(self, fn, args) # traces
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/loop_body.py", line 460, in init
ops.output(fn(*args))
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/ir.py", line 1075, in store_reduction
self.inner_fn(vars, reduction_vars),
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/lowering.py", line 5620, in loader
return inner_loader(new_index)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/lowering.py", line 5859, in fn
return pow_recursive(loader(idx), b, a.get_dtype())
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/lowering.py", line 3340, in fn
var_index = indices_loader(idx[:indices_ndim])
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/ir.py", line 2455, in loader
return inner(reindex(idx))
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/lowering.py", line 1282, in inner_fn
ops.masked(
File "", line 272, in masked
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/virtualized.py", line 286, in _default
return OpsWrapper._wrap(getattr(_ops, name)(*new_args, **new_kwargs))
File "", line 272, in masked
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/index_propagation.py", line 295, in _default
return self.fallback(name, args, kwargs)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/index_propagation.py", line 269, in fallback
return self.wrap(getattr(self._inner, name)(*new_args, **new_kwargs))
File "", line 272, in masked
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/loop_body.py", line 498, in _default
return getattr(self._inner, name)(*args, **kwargs)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/loop_body.py", line 625, in masked
self.body.subblocks[name] = LoopBodyBlock(self.body, masked_body, [])
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/loop_body.py", line 460, in init
ops.output(fn(*args))
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/lowering.py", line 1284, in
lambda: inputs_loadersi,
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/lowering.py", line 1251, in inner_fn
idx_dim = ops.index_expr(idx[dim], torch.int64)
File "", line 188, in index_expr
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/virtualized.py", line 286, in _default
return OpsWrapper._wrap(getattr(_ops, name)(*new_args, **new_kwargs))
File "", line 188, in index_expr
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/index_propagation.py", line 305, in _default
return self.propagate_sympy(name, args, kwargs)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/index_propagation.py", line 282, in propagate_sympy
new_expr = getattr(SymPyOps, name)(*new_args, **new_kwargs)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/index_propagation.py", line 88, in index_expr
return TypedExpr(value, dtype)
File "", line 5, in init
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/torch/_inductor/index_propagation.py", line 67, in post_init
self.expr = dtype_to_type(self.dtype)(self.expr)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/sympy/core/expr.py", line 308, in int
r = self.round(2)
File "/home/sdp/miniforge3/envs/e2e_ci/lib/python3.10/site-packages/sympy/core/expr.py", line 3838, in round
raise TypeError(
torch._inductor.exc.InductorError: TypeError: Expected a number but got Identity:

Set TORCH_LOGS="+dynamo" and TORCHDYNAMO_VERBOSE=1 for more information

TorchDynamo optimized model failed to run because of following error
fail_to_run

Versions

Envirnoments:
Device: PVC 1100
torch-xpu-ops: https://github.com/etaf/pytorch-inductor-xpu/commits/xpu_inductor_windows/
python: 3.10
TRITON_COMMIT_ID: c4a79a1960ba1c247c2548cbd3abf6a728b3ce6f
TORCH_COMMIT_ID: 1ba49a78edafa61e2ce4f80d147576e66566eec3
TORCHBENCH_COMMIT_ID: 373ffb19dc470f4423a3176a4133f8f4b3cdb5bd
TORCHVISION_COMMIT_ID: d23a6e1664d20707c11781299611436e1f0c104f
TORCHAUDIO_COMMIT_ID: f084f34bbb743fada85f66b0ed8041387565e69c
DRIVER_VERSION: 1.23.10.49.231129.50
KERNEL_VERSION: 5.15.0-73-generic #80 SMP Mon May 15 15:18:26 UTC 2023
BUNDLE_VERSION: 2025.0.1.20241113
OS_PRETTY_NAME: Ubuntu 22.04.2 LTS
GCC_VERSION: 11

@daisyden daisyden added this to the PT2.7 milestone Feb 24, 2025
@daisyden daisyden added the E2E label Feb 24, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

3 participants