torch/dynamo/torch.compile causes triton.compiler.errors.CompilationError

Open

#87 opened on Mar 25, 2025 6 comments 0 reactions 0 assignees View on GitHub

Dominant language: Python
Stars: 2.1k
Forks: 196
PR merge metrics: No merged PRs in 30d

Description

**Issue**
This problem has been troubling me for a long time. While training the BLT model, I encountered an error during torch.compile/_dynamo of flex_attention. The error traceback indicated a NameError for 's10' that causes triton.compiler.errors.CompilationError. I hope everyone can help me solve this issue, thank you.
The environment was set up using .`create_env.sh`

**Error Traceback**
```
[rank0]: Traceback (most recent call last):
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/output_graph.py", line 1446, in _call_user_compiler
[rank0]: compiled_fn = compiler_fn(gm, self.example_inputs())
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/repro/after_dynamo.py", line 129, in __call__
[rank0]: compiled_gm = compiler_fn(gm, example_inputs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/__init__.py", line 2235, in __call__
[rank0]: return compile_fx(model_, inputs_, config_patches=self.config)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/compile_fx.py", line 1521, in compile_fx
[rank0]: return aot_autograd(
[rank0]: ^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/backends/common.py", line 72, in __call__
[rank0]: cg = aot_module_simplified(gm, example_inputs, **self.kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_functorch/aot_autograd.py", line 1071, in aot_module_simplified
[rank0]: compiled_fn = dispatch_and_compile()
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_functorch/aot_autograd.py", line 1056, in dispatch_and_compile
[rank0]: compiled_fn, _ = create_aot_dispatcher_function(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_functorch/aot_autograd.py", line 522, in create_aot_dispatcher_function
[rank0]: return _create_aot_dispatcher_function(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_functorch/aot_autograd.py", line 759, in _create_aot_dispatcher_function
[rank0]: compiled_fn, fw_metadata = compiler_fn(
[rank0]: ^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_functorch/_aot_autograd/jit_compile_runtime_wrappers.py", line 586, in aot_dispatch_autograd
[rank0]: compiled_fw_func = aot_config.fw_compiler(fw_module, adjusted_flat_args)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/compile_fx.py", line 1350, in fw_compiler_base
[rank0]: return _fw_compiler_base(model, example_inputs, is_inference)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/compile_fx.py", line 1421, in _fw_compiler_base
[rank0]: return inner_compile(
[rank0]: ^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/compile_fx.py", line 475, in compile_fx_inner
[rank0]: return wrap_compiler_debug(_compile_fx_inner, compiler_name="inductor")(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/repro/after_aot.py", line 85, in debug_wrapper
[rank0]: inner_compiled_fn = compiler_fn(gm, example_inputs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/compile_fx.py", line 661, in _compile_fx_inner
[rank0]: compiled_graph = FxGraphCache.load(
[rank0]: ^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/codecache.py", line 1370, in load
[rank0]: compiled_graph = compile_fx_fn(
[rank0]: ^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/compile_fx.py", line 570, in codegen_and_compile
[rank0]: compiled_graph = fx_codegen_and_compile(gm, example_inputs, **fx_kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/compile_fx.py", line 878, in fx_codegen_and_compile
[rank0]: compiled_fn = graph.compile_to_fn()
[rank0]: ^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/graph.py", line 1913, in compile_to_fn
[rank0]: return self.compile_to_module().call
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/graph.py", line 1839, in compile_to_module
[rank0]: return self._compile_to_module()
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/graph.py", line 1867, in _compile_to_module
[rank0]: mod = PyCodeCache.load_by_key_path(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/codecache.py", line 2876, in load_by_key_path
[rank0]: mod = _reload_python_module(key, path)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/runtime/compile_tasks.py", line 45, in _reload_python_module
[rank0]: exec(code, mod.__dict__, mod.__dict__)
[rank0]: File "/tmp/torchinductor_guo/su/csuznihssg2brk4uicpccso4qteblry2inofleo47kgyiopzdpcr.py", line 474, in
[rank0]: async_compile.wait(globals())
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/async_compile.py", line 276, in wait
[rank0]: scope[key] = result.result()
[rank0]: ^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/codecache.py", line 3341, in result
[rank0]: result = self.future.result()
[rank0]: ^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/concurrent/futures/_base.py", line 456, in result
[rank0]: return self.__get_result()
[rank0]: ^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/concurrent/futures/_base.py", line 401, in __get_result
[rank0]: raise self._exception
[rank0]: torch._inductor.compile_worker.subproc_pool.SubprocException: An exception occurred in a subprocess:

[rank0]: triton.compiler.errors.CompilationError: at 56:48:
[rank0]: m = offs_m
[rank0]: n = offs_n

[rank0]: post_mod_scores = (qk)

[rank0]: if CHECK_BLOCK_BOUNDARY:
[rank0]: # Mask out the elements that are out of the KV_LEN for non divisible seqlen.
[rank0]: post_mod_scores = tl.where(offs_n < KV_LEN, post_mod_scores, float("-inf"))

[rank0]: if not IS_FULL_BLOCKS:
[rank0]: mask_mod_output = tl.load(in_ptr8 + (m)*s10 + (n) + (off_z)*s10*s9)
[rank0]: ^
[rank0]: NameError('s10 is not defined')

[rank0]: The above exception was the direct cause of the following exception:

[rank0]: triton.compiler.errors.CompilationError: at 57:28:
[rank0]: acc, l_i, m_i,
[rank0]: # Offsets
[rank0]: off_z, off_h, offs_m, offs_n,
[rank0]: MATMUL_PRECISION, RCP_LN2,
[rank0]: IS_FULL_BLOCKS,
[rank0]: )
[rank0]: else:
[rank0]: # Benchmark shows even we applied mod & mask to each block for non divisible seqlen,
[rank0]: # it's on par or slightly faster than only applying to the last block in fwd.
[rank0]: # However, we choose different strategy for bwd, where we only apply mod & mask
[rank0]: # to the last block because it's faster a lot.
[rank0]: acc, l_i, m_i = forward_block_mn(
[rank0]: ^

[rank0]: The above exception was the direct cause of the following exception:

[rank0]: Traceback (most recent call last):
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/compile_worker/subproc_pool.py", line 270, in do_job
[rank0]: result = job()
[rank0]: ^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/runtime/compile_tasks.py", line 68, in _worker_compile_triton
[rank0]: load_kernel().precompile(warm_cache_only=True)
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/runtime/triton_heuristics.py", line 244, in precompile
[rank0]: compiled_binary, launcher = self._precompile_config(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_inductor/runtime/triton_heuristics.py", line 428, in _precompile_config
[rank0]: triton.compile(*compile_args, **compile_kwargs),
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/triton/compiler/compiler.py", line 276, in compile
[rank0]: module = src.make_ir(options, codegen_fns, context)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/triton/compiler/compiler.py", line 113, in make_ir
[rank0]: return ast_to_ttir(self.fn, self, context=context, options=options, codegen_fns=codegen_fns)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: triton.compiler.errors.CompilationError: at 148:20:
[rank0]: )
[rank0]: V_block_ptr = tl.make_block_ptr(
[rank0]: base=V,
[rank0]: shape=(KV_LEN, V_HEAD_DIM),
[rank0]: strides=(stride_vn, stride_vk),
[rank0]: offsets=(kv_start, 0),
[rank0]: block_shape=(BLOCK_N, V_HEAD_DIM),
[rank0]: order=(1, 0)
[rank0]: )
[rank0]: offs_n = kv_start + tl.arange(0, BLOCK_N)

[rank0]: acc, l_i, m_i = forward_inner(
[rank0]: ^

[rank0]: The above exception was the direct cause of the following exception:

[rank0]: Traceback (most recent call last):
[rank0]: File "", line 198, in _run_module_as_main
[rank0]: File "", line 88, in _run_code
[rank0]: File "/home/guo/Desktop/Project/blt-main/bytelatent/train.py", line 812, in
[rank0]: main()
[rank0]: File "/home/guo/Desktop/Project/blt-main/bytelatent/train.py", line 808, in main
[rank0]: train(train_args)
[rank0]: File "/home/guo/Desktop/Project/blt-main/bytelatent/train.py", line 475, in train
[rank0]: pred = model(
[rank0]: ^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1736, in _wrapped_call_impl
[rank0]: return self._call_impl(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1844, in _call_impl
[rank0]: return inner()
[rank0]: ^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1790, in inner
[rank0]: result = forward_call(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/guo/Desktop/Project/blt-main/bytelatent/model/blt.py", line 1023, in forward
[rank0]: output, _ = self.local_decoder(
[rank0]: ^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1736, in _wrapped_call_impl
[rank0]: return self._call_impl(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1747, in _call_impl
[rank0]: return forward_call(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/guo/Desktop/Project/blt-main/bytelatent/model/local_models.py", line 390, in forward
[rank0]: h_cross = self.cross_attn_layers[i](
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1736, in _wrapped_call_impl
[rank0]: return self._call_impl(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1844, in _call_impl
[rank0]: return inner()
[rank0]: ^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1790, in inner
[rank0]: result = forward_call(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/guo/Desktop/Project/blt-main/bytelatent/model/latent_transformer.py", line 105, in forward
[rank0]: output = flex_attention_comp(xq, xk, xv, block_mask=mask)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/eval_frame.py", line 465, in _fn
[rank0]: return fn(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/convert_frame.py", line 1269, in __call__
[rank0]: return self._torchdynamo_orig_callable(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/convert_frame.py", line 1064, in __call__
[rank0]: result = self._inner_convert(
[rank0]: ^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/convert_frame.py", line 526, in __call__
[rank0]: return _compile(
[rank0]: ^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/convert_frame.py", line 924, in _compile
[rank0]: guarded_code = compile_inner(code, one_graph, hooks, transform)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/convert_frame.py", line 666, in compile_inner
[rank0]: return _compile_inner(code, one_graph, hooks, transform)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_utils_internal.py", line 87, in wrapper_function
[rank0]: return function(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/convert_frame.py", line 699, in _compile_inner
[rank0]: out_code = transform_code_object(code, transform)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/bytecode_transformation.py", line 1322, in transform_code_object
[rank0]: transformations(instructions, code_options)
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/convert_frame.py", line 219, in _fn
[rank0]: return fn(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/convert_frame.py", line 634, in transform
[rank0]: tracer.run()
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/symbolic_convert.py", line 2796, in run
[rank0]: super().run()
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/symbolic_convert.py", line 983, in run
[rank0]: while self.step():
[rank0]: ^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/symbolic_convert.py", line 895, in step
[rank0]: self.dispatch_table[inst.opcode](self, inst)
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/symbolic_convert.py", line 2987, in RETURN_VALUE
[rank0]: self._return(inst)
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/symbolic_convert.py", line 2972, in _return
[rank0]: self.output.compile_subgraph(
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/output_graph.py", line 1117, in compile_subgraph
[rank0]: self.compile_and_call_fx_graph(tx, list(reversed(stack_values)), root)
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/output_graph.py", line 1369, in compile_and_call_fx_graph
[rank0]: compiled_fn = self.call_user_compiler(gm)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/output_graph.py", line 1416, in call_user_compiler
[rank0]: return self._call_user_compiler(gm)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/anaconda3/envs/blt/lib/python3.11/site-packages/torch/_dynamo/output_graph.py", line 1465, in _call_user_compiler
[rank0]: raise BackendCompilerFailed(self.compiler_fn, e) from e
[rank0]: torch._dynamo.exc.BackendCompilerFailed: backend='inductor' raised:
[rank0]: SubprocException: An exception occurred in a subprocess:

[rank0]: triton.compiler.errors.CompilationError: at 56:48:
[rank0]: m = offs_m
[rank0]: n = offs_n

[rank0]: post_mod_scores = (qk)

[rank0]: if not IS_FULL_BLOCKS:
[rank0]: mask_mod_output = tl.load(in_ptr8 + (m)*s10 + (n) + (off_z)*s10*s9)
[rank0]: ^
[rank0]: NameError('s10 is not defined')

[rank0]: The above exception was the direct cause of the following exception:

[rank0]: acc, l_i, m_i = forward_inner(
[rank0]: ^

[rank0]: Set TORCH_LOGS="+dynamo" and TORCHDYNAMO_VERBOSE=1 for more information

[rank0]: You can suppress this exception and fall back to eager by setting:
[rank0]: import torch._dynamo
[rank0]: torch._dynamo.config.suppress_errors = True

Process blt_data_loader:
Traceback (most recent call last):
File "/usr/local/anaconda3/envs/blt/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap
self.run()
File "/usr/local/anaconda3/envs/blt/lib/python3.11/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/guo/Desktop/Project/blt-main/bytelatent/data/iterators/multiprocess_iterator.py", line 84, in start_work_from_state
batch_queue.put(item, timeout=0.1)
File "", line 2, in put
File "/usr/local/anaconda3/envs/blt/lib/python3.11/multiprocessing/managers.py", line 822, in _callmethod
kind, result = conn.recv()
^^^^^^^^^^^
File "/usr/local/anaconda3/envs/blt/lib/python3.11/multiprocessing/connection.py", line 250, in recv
buf = self._recv_bytes()
^^^^^^^^^^^^^^^^^^
File "/usr/local/anaconda3/envs/blt/lib/python3.11/multiprocessing/connection.py", line 430, in _recv_bytes
buf = self._recv(4)
^^^^^^^^^^^^^
File "/usr/local/anaconda3/envs/blt/lib/python3.11/multiprocessing/connection.py", line 399, in _recv
raise EOFError
EOFError
```

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

torch/dynamo/torch.compile causes triton.compiler.errors.CompilationError

Description

Contributor guide

Assessment

Get new issues in your inbox