Rate this Page
★ ★ ★ ★ ★
fullgraph=False">
error_on_graph_break">

Toggling error_on_graph_break#

Created On: Sep 03, 2025 | Last Updated On: Sep 03, 2025

Summary:

  • When fullgraph=False, we can use torch._dynamo.error_on_graph_break() for more flexibility in dealing with graph breaks.

So far, we have introduced two ways in dealing with graph breaks in torch.compile:

  1. fullgraph=True errors on the first graph break and additionally guarantees that only one graph is traced from the code.

  2. fullgraph=False continues tracing even when encountering graph breaks.

What if we want to disallow graph breaks for most of the code, but there are a few problematic functions where the graph breaks are hard to remove, and we are okay with having those graph breaks? We can use torch._dynamo.error_on_graph_break() to achieve this.

torch.compile has an error_on_graph_break setting (initially set to False). If a graph break or compiler error occurs in code while error_on_graph_break is set to False, then torch.compile will attempt to continue compilation after the graph break/error. If error_on_graph_break is set to True, then torch.compile will abort compilation and propagate the error to user code.

A significant difference between error_on_graph_break=True and fullgraph=True is that the former does not guarantee that a single graph will be captured. error_on_graph_break can be arbitrarily toggled during compile time by using the torch._dynamo.error_on_graph_break() context manager/decorator. In comparison, once fullgraph is set to True, it cannot be set back to False. Finally, error_on_graph_break has lower precedence than fullgraph - error_on_graph_break only takes effect when fullgraph=False.

error_on_graph_break(False) example#

@torch._dynamo.error_on_graph_break(False)
def code_with_a_difficult_graph_break(x):
    x = x + 1
    torch._dynamo.graph_break()
    return x + 2

def inner(x):
    return code_with_a_difficult_graph_break(x)

# NOTE: fullgraph=False
@torch._dynamo.error_on_graph_break(True)
@torch.compile
def fn(x):
    return inner(x)

# No error, but there is a graph break
fn(torch.randn(3))

Graph break in user code at /tmp/ipykernel_880/1452578661.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
User code traceback:
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel_launcher.py", line 18, in <module>
    app.launch_new_instance()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/traitlets/config/application.py", line 1075, in launch_instance
    app.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelapp.py", line 739, in start
    self.io_loop.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/tornado/platform/asyncio.py", line 211, in start
    self.asyncio_loop.run_forever()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 603, in run_forever
    self._run_once()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 1909, in _run_once
    handle._run()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/events.py", line 80, in _run
    self._context.run(self._callback, *self._args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 519, in dispatch_queue
    await self.process_one()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 508, in process_one
    await dispatch(*args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 400, in dispatch_shell
    await result
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 368, in execute_request
    await super().execute_request(stream, ident, parent)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 767, in execute_request
    reply_content = await reply_content
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 455, in do_execute
    res = shell.run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/zmqshell.py", line 577, in run_cell
    return super().run_cell(*args, **kwargs)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3006, in run_cell
    result = self._run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3061, in _run_cell
    result = runner(coro)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/async_helpers.py", line 129, in _pseudo_sync_runner
    coro.send(None)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3266, in run_cell_async
    has_raised = await self.run_ast_nodes(code_ast.body, cell_name,
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3445, in run_ast_nodes
    if await self.run_code(code, result, async_=asy):
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3505, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "/tmp/ipykernel_880/1452578661.py", line 17, in <module>
    fn(torch.randn(3))
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 209, in inner
    return func(*args, **kwargs)
  File "/tmp/ipykernel_880/1452578661.py", line 14, in fn
    return inner(x)
  File "/tmp/ipykernel_880/1452578661.py", line 8, in inner
    return code_with_a_difficult_graph_break(x)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 209, in inner
    return func(*args, **kwargs)
  File "/tmp/ipykernel_880/1452578661.py", line 4, in code_with_a_difficult_graph_break
    torch._dynamo.graph_break()

Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/1452578661.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/1452578661.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/1452578661.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html

tensor([3.7641, 3.6966, 2.6646])

Using error_on_graph_break(False) under error_on_graph_break(True) is helpful for when we want to minimize graph breaks (i.e. follow the fullgraph=True programming model), but there are some sections of code with non-performance-critical graph breaks that are difficult to work around.

error_on_graph_break() can be used as a context manager as well:

# NOTE: fullgraph=False
@torch._dynamo.error_on_graph_break(True)
@torch.compile
def fn(x):
    x = x + 1
    with torch._dynamo.error_on_graph_break(False):
        torch._dynamo.graph_break()  # no error
    return x + 2

# No error, but there is a graph break
fn(torch.randn(3))

Graph break in user code at /tmp/ipykernel_880/737485247.py:7
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
User code traceback:
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel_launcher.py", line 18, in <module>
    app.launch_new_instance()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/traitlets/config/application.py", line 1075, in launch_instance
    app.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelapp.py", line 739, in start
    self.io_loop.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/tornado/platform/asyncio.py", line 211, in start
    self.asyncio_loop.run_forever()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 603, in run_forever
    self._run_once()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 1909, in _run_once
    handle._run()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/events.py", line 80, in _run
    self._context.run(self._callback, *self._args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 519, in dispatch_queue
    await self.process_one()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 508, in process_one
    await dispatch(*args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 400, in dispatch_shell
    await result
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 368, in execute_request
    await super().execute_request(stream, ident, parent)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 767, in execute_request
    reply_content = await reply_content
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 455, in do_execute
    res = shell.run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/zmqshell.py", line 577, in run_cell
    return super().run_cell(*args, **kwargs)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3006, in run_cell
    result = self._run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3061, in _run_cell
    result = runner(coro)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/async_helpers.py", line 129, in _pseudo_sync_runner
    coro.send(None)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3266, in run_cell_async
    has_raised = await self.run_ast_nodes(code_ast.body, cell_name,
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3445, in run_ast_nodes
    if await self.run_code(code, result, async_=asy):
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3505, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "/tmp/ipykernel_880/737485247.py", line 11, in <module>
    fn(torch.randn(3))
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 209, in inner
    return func(*args, **kwargs)
  File "/tmp/ipykernel_880/737485247.py", line 7, in fn
    torch._dynamo.graph_break()  # no error

tensor([3.2377, 5.2136, 3.8709])

You can use monkey patching to toggle error_on_graph_break for code where you cannot edit the source (e.g. framework code):

class ThirdPartyModule(torch.nn.Module):
    def forward(self, x):
        x = x + 1
        torch._dynamo.graph_break()
        return x + 2

tp_mod = ThirdPartyModule()
tp_mod.forward = torch._dynamo.error_on_graph_break(False)(tp_mod.forward)

@torch._dynamo.error_on_graph_break(True)
@torch.compile
def fn(x):
    return tp_mod.forward(x)

# No error, but there is a graph break
fn(torch.randn(3))

Graph break in user code at /tmp/ipykernel_880/2112598647.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
User code traceback:
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel_launcher.py", line 18, in <module>
    app.launch_new_instance()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/traitlets/config/application.py", line 1075, in launch_instance
    app.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelapp.py", line 739, in start
    self.io_loop.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/tornado/platform/asyncio.py", line 211, in start
    self.asyncio_loop.run_forever()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 603, in run_forever
    self._run_once()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 1909, in _run_once
    handle._run()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/events.py", line 80, in _run
    self._context.run(self._callback, *self._args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 519, in dispatch_queue
    await self.process_one()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 508, in process_one
    await dispatch(*args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 400, in dispatch_shell
    await result
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 368, in execute_request
    await super().execute_request(stream, ident, parent)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 767, in execute_request
    reply_content = await reply_content
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 455, in do_execute
    res = shell.run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/zmqshell.py", line 577, in run_cell
    return super().run_cell(*args, **kwargs)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3006, in run_cell
    result = self._run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3061, in _run_cell
    result = runner(coro)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/async_helpers.py", line 129, in _pseudo_sync_runner
    coro.send(None)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3266, in run_cell_async
    has_raised = await self.run_ast_nodes(code_ast.body, cell_name,
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3445, in run_ast_nodes
    if await self.run_code(code, result, async_=asy):
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3505, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "/tmp/ipykernel_880/2112598647.py", line 16, in <module>
    fn(torch.randn(3))
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 209, in inner
    return func(*args, **kwargs)
  File "/tmp/ipykernel_880/2112598647.py", line 13, in fn
    return tp_mod.forward(x)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 209, in inner
    return func(*args, **kwargs)
  File "/tmp/ipykernel_880/2112598647.py", line 4, in forward
    torch._dynamo.graph_break()

Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/2112598647.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/2112598647.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html

tensor([4.0035, 2.7495, 4.1951])

error_on_graph_break(True) example#

@torch._dynamo.error_on_graph_break(True)
def inner2(x):
    x = x + 1
    torch._dynamo.graph_break()  # error
    return x + 2

def inner(x):
    return inner2(x)

# fullgraph=False, error_on_graph_break=False
@torch.compile
def fn(x):
    x = x + 4
    torch._dynamo.graph_break()  # no error
    return inner(x)

try:
    fn(torch.randn(3))
except Exception as e:
    print(e)

Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html

from user code:
   File "/tmp/ipykernel_880/2379101916.py", line 15, in torch_dynamo_resume_in_fn_at_14
    return inner(x)
  File "/tmp/ipykernel_880/2379101916.py", line 8, in inner
    return inner2(x)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 209, in inner
    return func(*args, **kwargs)
  File "/tmp/ipykernel_880/2379101916.py", line 4, in inner2
    torch._dynamo.graph_break()  # error

Set TORCHDYNAMO_VERBOSE=1 for the internal stack trace (please do this especially if you're reporting a bug to PyTorch). For even more developer context, set TORCH_LOGS="+dynamo"

Graph break in user code at /tmp/ipykernel_880/2379101916.py:14
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
User code traceback:
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel_launcher.py", line 18, in <module>
    app.launch_new_instance()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/traitlets/config/application.py", line 1075, in launch_instance
    app.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelapp.py", line 739, in start
    self.io_loop.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/tornado/platform/asyncio.py", line 211, in start
    self.asyncio_loop.run_forever()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 603, in run_forever
    self._run_once()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 1909, in _run_once
    handle._run()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/events.py", line 80, in _run
    self._context.run(self._callback, *self._args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 519, in dispatch_queue
    await self.process_one()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 508, in process_one
    await dispatch(*args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 400, in dispatch_shell
    await result
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 368, in execute_request
    await super().execute_request(stream, ident, parent)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 767, in execute_request
    reply_content = await reply_content
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 455, in do_execute
    res = shell.run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/zmqshell.py", line 577, in run_cell
    return super().run_cell(*args, **kwargs)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3006, in run_cell
    result = self._run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3061, in _run_cell
    result = runner(coro)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/async_helpers.py", line 129, in _pseudo_sync_runner
    coro.send(None)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3266, in run_cell_async
    has_raised = await self.run_ast_nodes(code_ast.body, cell_name,
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3445, in run_ast_nodes
    if await self.run_code(code, result, async_=asy):
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3505, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "/tmp/ipykernel_880/2379101916.py", line 18, in <module>
    fn(torch.randn(3))
  File "/tmp/ipykernel_880/2379101916.py", line 14, in fn
    torch._dynamo.graph_break()  # no error

Using error_on_graph_break(True) under error_on_graph_break(False) is helpful for when we want to use torch.compile flexibly (i.e. follow the fullgraph=False programming model), but there are some sections of the code that are performance-critical and we want to ensure that those sections do not contain graph breaks.

error_on_graph_break nesting behavior#

torch._dynamo.error_on_graph_break() affects the error_on_graph_break setting of nested calls as well:

def inner(x):
    x = x + 1
    torch._dynamo.graph_break()
    return x + 2

def inner2(x):
    with torch._dynamo.error_on_graph_break(False):
        return inner(x)

@torch._dynamo.error_on_graph_break(True)
@torch.compile
def fn(x):
    return inner2(x)

# no error
fn(torch.randn(3))

Graph break in user code at /tmp/ipykernel_880/1007149706.py:3
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
User code traceback:
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel_launcher.py", line 18, in <module>
    app.launch_new_instance()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/traitlets/config/application.py", line 1075, in launch_instance
    app.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelapp.py", line 739, in start
    self.io_loop.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/tornado/platform/asyncio.py", line 211, in start
    self.asyncio_loop.run_forever()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 603, in run_forever
    self._run_once()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 1909, in _run_once
    handle._run()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/events.py", line 80, in _run
    self._context.run(self._callback, *self._args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 519, in dispatch_queue
    await self.process_one()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 508, in process_one
    await dispatch(*args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 400, in dispatch_shell
    await result
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 368, in execute_request
    await super().execute_request(stream, ident, parent)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 767, in execute_request
    reply_content = await reply_content
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 455, in do_execute
    res = shell.run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/zmqshell.py", line 577, in run_cell
    return super().run_cell(*args, **kwargs)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3006, in run_cell
    result = self._run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3061, in _run_cell
    result = runner(coro)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/async_helpers.py", line 129, in _pseudo_sync_runner
    coro.send(None)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3266, in run_cell_async
    has_raised = await self.run_ast_nodes(code_ast.body, cell_name,
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3445, in run_ast_nodes
    if await self.run_code(code, result, async_=asy):
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3505, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "/tmp/ipykernel_880/1007149706.py", line 16, in <module>
    fn(torch.randn(3))
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 209, in inner
    return func(*args, **kwargs)
  File "/tmp/ipykernel_880/1007149706.py", line 13, in fn
    return inner2(x)
  File "/tmp/ipykernel_880/1007149706.py", line 8, in inner2
    return inner(x)
  File "/tmp/ipykernel_880/1007149706.py", line 3, in inner
    torch._dynamo.graph_break()

Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/1007149706.py:3
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/1007149706.py:3
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html

tensor([1.5647, 2.7757, 2.3073])

torch._dynamo.error_on_graph_break() can be used under another torch._dynamo.error_on_graph_break() region:

def inner(x):
    x = x + 1
    with torch._dynamo.error_on_graph_break(False):
        torch._dynamo.graph_break()
    return x + 2

def inner2(x):
    with torch._dynamo.error_on_graph_break(True):
        return inner(x)

@torch.compile
def fn(x):
    return inner2(x)

# no error
fn(torch.randn(3))

Graph break in user code at /tmp/ipykernel_880/1343774799.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
User code traceback:
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel_launcher.py", line 18, in <module>
    app.launch_new_instance()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/traitlets/config/application.py", line 1075, in launch_instance
    app.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelapp.py", line 739, in start
    self.io_loop.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/tornado/platform/asyncio.py", line 211, in start
    self.asyncio_loop.run_forever()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 603, in run_forever
    self._run_once()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 1909, in _run_once
    handle._run()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/events.py", line 80, in _run
    self._context.run(self._callback, *self._args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 519, in dispatch_queue
    await self.process_one()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 508, in process_one
    await dispatch(*args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 400, in dispatch_shell
    await result
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 368, in execute_request
    await super().execute_request(stream, ident, parent)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 767, in execute_request
    reply_content = await reply_content
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 455, in do_execute
    res = shell.run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/zmqshell.py", line 577, in run_cell
    return super().run_cell(*args, **kwargs)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3006, in run_cell
    result = self._run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3061, in _run_cell
    result = runner(coro)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/async_helpers.py", line 129, in _pseudo_sync_runner
    coro.send(None)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3266, in run_cell_async
    has_raised = await self.run_ast_nodes(code_ast.body, cell_name,
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3445, in run_ast_nodes
    if await self.run_code(code, result, async_=asy):
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3505, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "/tmp/ipykernel_880/1343774799.py", line 16, in <module>
    fn(torch.randn(3))
  File "/tmp/ipykernel_880/1343774799.py", line 13, in fn
    return inner2(x)
  File "/tmp/ipykernel_880/1343774799.py", line 9, in inner2
    return inner(x)
  File "/tmp/ipykernel_880/1343774799.py", line 4, in inner
    torch._dynamo.graph_break()

Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/1343774799.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
Graph break (user stack suppressed due to duplicate graph break) in user code at /tmp/ipykernel_880/1343774799.py:4
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html

tensor([2.7914, 2.5962, 1.1888])

Interaction with fullgraph#

fullgraph=True takes higher precedence than error_on_graph_break:

@torch._dynamo.error_on_graph_break(False)
def inner(x):
    x = x + 1
    torch._dynamo.graph_break()
    return x + 2

@torch.compile(fullgraph=True)
def fn(x):
    return inner(x)

try:
    fn(torch.randn(3))
except Exception as e:
    print(e)

Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html

from user code:
   File "/tmp/ipykernel_880/2331424258.py", line 9, in fn
    return inner(x)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 209, in inner
    return func(*args, **kwargs)
  File "/tmp/ipykernel_880/2331424258.py", line 3, in inner
    x = x + 1

Set TORCHDYNAMO_VERBOSE=1 for the internal stack trace (please do this especially if you're reporting a bug to PyTorch). For even more developer context, set TORCH_LOGS="+dynamo"

fullgraph=True cannot be toggled back to fullgraph=False:

@torch.compile(fullgraph=False)
def inner(x):
    x = x + 1
    torch._dynamo.graph_break()
    return x + 2

@torch.compile(fullgraph=True)
def fn(x):
    return inner(x)

try:
    fn(torch.randn(3))
except Exception as e:
    print(e)

Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html

from user code:
   File "/tmp/ipykernel_880/262151723.py", line 9, in fn
    return inner(x)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/polyfills/__init__.py", line 264, in getattr_and_trace
    return fn(*args[2:], **kwargs)
  File "/tmp/ipykernel_880/262151723.py", line 3, in inner
    x = x + 1

Set TORCHDYNAMO_VERBOSE=1 for the internal stack trace (please do this especially if you're reporting a bug to PyTorch). For even more developer context, set TORCH_LOGS="+dynamo"

@torch.compile(fullgraph=True)
def inner(x):
    x = x + 1
    torch._dynamo.graph_break()
    return x + 2

@torch.compile(fullgraph=False)
def fn(x):
    return inner(x)

try:
    fn(torch.randn(3))
except Exception as e:
    print(e)

Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html

from user code:
   File "/tmp/ipykernel_880/2119173801.py", line 3, in inner
    x = x + 1

Set TORCHDYNAMO_VERBOSE=1 for the internal stack trace (please do this especially if you're reporting a bug to PyTorch). For even more developer context, set TORCH_LOGS="+dynamo"

Graph break in user code at /tmp/ipykernel_880/2119173801.py:3
Graph Break Reason: Call to `torch._dynamo.graph_break()`
  Explanation: User-inserted graph break. Message: None
  Hint: Remove the `torch._dynamo.graph_break()` call.

  Developer debug context: Called `torch._dynamo.graph_break()` with args `[]`, kwargs `{}`

 For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0025.html
User code traceback:
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/envs/py_3.10/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel_launcher.py", line 18, in <module>
    app.launch_new_instance()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/traitlets/config/application.py", line 1075, in launch_instance
    app.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelapp.py", line 739, in start
    self.io_loop.start()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/tornado/platform/asyncio.py", line 211, in start
    self.asyncio_loop.run_forever()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 603, in run_forever
    self._run_once()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/base_events.py", line 1909, in _run_once
    handle._run()
  File "/opt/conda/envs/py_3.10/lib/python3.10/asyncio/events.py", line 80, in _run
    self._context.run(self._callback, *self._args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 519, in dispatch_queue
    await self.process_one()
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 508, in process_one
    await dispatch(*args)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 400, in dispatch_shell
    await result
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 368, in execute_request
    await super().execute_request(stream, ident, parent)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/kernelbase.py", line 767, in execute_request
    reply_content = await reply_content
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/ipkernel.py", line 455, in do_execute
    res = shell.run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/ipykernel/zmqshell.py", line 577, in run_cell
    return super().run_cell(*args, **kwargs)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3006, in run_cell
    result = self._run_cell(
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3061, in _run_cell
    result = runner(coro)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/async_helpers.py", line 129, in _pseudo_sync_runner
    coro.send(None)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3266, in run_cell_async
    has_raised = await self.run_ast_nodes(code_ast.body, cell_name,
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3445, in run_ast_nodes
    if await self.run_code(code, result, async_=asy):
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3505, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "/tmp/ipykernel_880/2119173801.py", line 12, in <module>
    fn(torch.randn(3))
  File "/tmp/ipykernel_880/2119173801.py", line 9, in fn
    return inner(x)
  File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/polyfills/__init__.py", line 264, in getattr_and_trace
    return fn(*args[2:], **kwargs)
  File "/tmp/ipykernel_880/2119173801.py", line 3, in inner
    x = x + 1

Summary of fullgraph=True/False vs error_on_graph_break#

Here is a table summarizing the differences between fullgraph=True/False and error_on_graph_break:

error_on_graph_break=True

error_on_graph_break=False (default)

fullgraph=True

Graph breaks result in errors. Only the first graph break will be reported. One graph guarantee.

fullgraph cannot be toggled to False. error_on_graph_break has no effect.

User code must be fully compatible with torch.compile. Guarantees no performance hits from graph breaks (because there are no graph breaks).

Ideal for code sensitive to graph breaks: framework/library code or cases where getting maximum performance is required. Prevents downstream user code from inadvertently allowing graph breaks.

Same as fullgraph=True and error_on_graph_break=True as error_on_graph_break has no effect when fullgraph=True.

fullgraph=False (default)

Graph breaks result in errors. Only the first graph break will be reported. No one graph guarantee.

error_on_graph_break can be toggled to False.

User code must be fully compatible with torch.compile. Guarantees no performance hits from graph breaks (because there are no graph breaks).

Ideal for user code sensitive to graph breaks. error_on_graph_break can be toggled to False to deal with sections that have graph breaks that are difficult to work around.

Will continue to compile after encountering graph breaks. All graph breaks will be reported.

error_on_graph_break can be toggled to True.

Doesn’t require many user code changes to work. Performance may be negatively impacted due to graph breaks.

Ideal for out-of-the-box use cases, on “non-weird” code, or where squeezing maximal performance is not necessary
On this page
Edit on GitHub
Show Source
PyTorch Libraries