Removes always restore, assuming that a HOP will cleanup any leftover state from tracing fwd + bwd
This required a minor change to the autograd fn variable higher order op. If we are tracing forward DON'T add the call_function node into the main graph, since we are only tracing it for the purposes of speculation. Instead return the result directly to be passed to the backward for speculation. This was the only observable side effect on the output graph that I found.
Test plan:
test_smoke_from_test_autograd in test_autograd_function.py
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115317
Approved by: https://github.com/voznesenskym, https://github.com/jansel
Summary:
Rename _device_mesh.py to device_mesh.py, update all callsites, add documentation.
We created stubs for public class and methods in torch.distributed.device_mesh so that torch.distributed.device_mesh can be imported with or without distributed is available().
Original diff reverted: D51629761
Original PR reverted: https://github.com/pytorch/pytorch/pull/115099
Prior to landing, CI signals are all passed. Shipit added the "ci/trunk" label to the PR and DID NOT wait for it and went ahead committing. More context can be found in the reverted PR above.
Test Plan: CI.
Differential Revision: D51861018
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115193
Approved by: https://github.com/fegin
*
Context:
Joel sees that unless he manually writes to the fake tensor memo, fakification seems to produce spurious symbols! Voz (me) objects, saying that not only is directly writing to memo a bad pattern, recursively invoking fakification on tensor subclass elements in dynamo should suffice! Joel says that while he morally agrees, he has a test proving otherwise, a most perplexing situation.
Digging in, I figured out that while *we were* making fake tensors correctly, with properly cached symbols and the like, we were *also* incorrectly creating spurious symbols, leading the test to fail.
Before this PR, we would only cache source->symint. This was generally fine, but meant that you would create a symbol, then potentially throw it out due to symint cache. For example, the cache hit flow was:
make a symbol (ex: s2) -> use it to make a symint -> hit the cache (my_source-s1)
Now, in this example, you have a symbol in your val_to_var/var_to_val (s2) that is unused. This is sound, but wasteful, and furthermore, misleading.
This was causing a test added in a PR in this stack to fail, specifically, because the test was using
```
curr_var_to_val = {
str(k): v for k, v in context.fake_mode.shape_env.var_to_val.items()
}
````
To validate that no new symbols were being created (that is, that recursively creating fake tensors for subclasses was working).
The test is correct, but the implementation of caching would make (by this method of observation) cache hits look like cache misses.
So, the fix here is to move the cache up to be a general symbol cache, rather than only a cache for symints.
The initial implementation did that! But then, it ran into some interesting errors when it came to replay. When replaying symbol creation, behaviors would diverge in the new shape env! How could that be? The answer is because creating a new shape_env resulted in us replaying symbol creation... but with a cache from a different shape env! This was short circuiting symbol creation - and so, adding an extra layer to the cache for id(shape_env) fixes the problem.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115396
Approved by: https://github.com/mlazos
After auditing higher_order_ops.py, the graph checkpoints were only getting used in the event of an exception, so it is safe to remove because we restart analysis in this case now.
To make this clearer the current state is the following:
Checkpoint side effects
Capture subgraph
if graph break:
restore as usual
else:
throw away inlining translator and subgraph tracer
Restore side effects
This will change to the following after this change:
Checkpoint side effects
Capture subgraph:
if graph break:
restart analysis
else:
throw away inlining translator and subgraph tracer
Restore side effects
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115321
Approved by: https://github.com/jansel, https://github.com/zou3519
As titled, this PR removes the unnessecary getitem call from the graph that's manipulated in MapHigherOrder, where we want to get the first dim slice of original tensor for specualtion but using call_method will accidentally create a get_item call in the graph, so want to avoid it by calling unpack_var_sequence on input tensor.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115207
Approved by: https://github.com/yanboliang
ghstack dependencies: #115115, #115204, #115205
We want to remove the map_wrapper and replace it with dynamo always on. This is the first step of this plan.
In this PR, we make dynamo directly generates a map_impl nodes. This hasn't touch the eager logic yet. So the execution path after this PR looks like 1. `dynamo -> map_impl` when torch.compile is on. (Before this PR, it's `dynamo -> map_wrapper -> map_impl` and 2. `map_wrapper -> map_impl` (This PR did't touch the logic here).
The added TODO(yidi) is addressed in the following pr.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115205
Approved by: https://github.com/yanboliang
ghstack dependencies: #115115, #115204
Previously we only supported Tensor, Constants, and SymNode. We lift
that restriction (there's not really a good reason for it). HOPs like
torch.cond, torch.map already do input validation (those are the ones
that can only support Tensor, Constant, and SymNode inputs).
Test Plan:
New test for `wrap`, which is a HOP that has
manually_set_subgraph_inputs=False
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115186
Approved by: https://github.com/ydwu4, https://github.com/yanboliang
ghstack dependencies: #115185
Continuation of #112185, following the design in this [doc](https://docs.google.com/document/d/1ipSxcTzEMMOAPvxP-YJlD5JBZZmIGgh8Q34ixtOUCRo).
Summary:
* Introduce `SubclassSymbolicPolicy` containing separate dynamic dim / constraint policies for the outer and inner tensors
* Expand the automatic dynamic algorithm to recurse into inner tensors and produce one of these for a subclass instance
* Maintain legacy behavior for subclasses by recursively calling `mark_dynamic()` on inner tensors *of the same dim as outer* when `mark_dynamic(outer, ...)` is called
* Addresses this: 6a86cf00ad/torch/_dynamo/variables/builder.py (L1750)
* Add `outer_size` and `outer_stride` arguments to `__tensor_unflatten__()` so that you can find out what symbols were allocated for the outer size / stride (you are expected to return a tensor that compares equal to the outer symbols)
* Signatures now:
```python
# attrs is a list of inner tensor attributes on x; inner_tensor = getattr(x, attr)
# ctx is anything useful for rebuilding the class we want to guard on
attrs, ctx = x.__tensor_flatten__()
...
# inner_tensors is a dict of {attr -> tensor}
# ctx is taken unmodified from flattening and (eventually) guarded on
# outer_size is the expected size of the output; possibly symbolic
# outer_stride is the expected strides of the output; possibly symbolic
y = MySubclass.__tensor_unflatten__(inner_tensors, ctx, outer_size, outer_stride)
# at the __tensor_unflatten__() call-site in PT2, we assert y.shape == outer_size and y.stride() == outer_stride
# the assert simplifies symbols when there are relationships between outer and inner symbols
```
* Size info needed for `NestedTensor` at least, stride info needed for `DTensor` at least
* Punting on `outer_storage_offset` because storage_offset handling is horribly broken in PT2 right now
* ~~Add new `__tensor_mark_dynamic__()` to allow overriding the behavior of mark_dynamic on a per-subclass basis~~ (booted to future work)
* ~~Add guards for tensor subclasses by calling `__tensor_flatten__()` in the guard to test equality on `ctx`~~
* Now handled in #114469
* Next PR: add TENSOR_MATCH guards on inner tensors
Pull Request resolved: https://github.com/pytorch/pytorch/pull/114311
Approved by: https://github.com/ezyang, https://github.com/drisspg, https://github.com/voznesenskym, https://github.com/bdhirsh
I was curious what hf_T5_generate was trying to deepcopy, so I updated the errror message:
Before:
```
STATS graph_break
("'skip function deepcopy in file /home/jansel/conda/envs/pytorch/lib/python3.10/copy.py'', skipped according skipfiles.SKIP_DIRS'", 3)
...
```
After:
```
STATS graph_break
('copy.deepcopy UserDefinedObjectVariable(GenerationConfig)', 3)
...
```
Related issue: #115122
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115120
Approved by: https://github.com/oulgen
ghstack dependencies: #115095, #115046, #115057, #115119
Summary:
Rename _device_mesh.py to device_mesh.py, update all callsites, adds documentation.
Original diff reverted: D51629761
Original PR reverted: https://github.com/pytorch/pytorch/pull/114991
It was failing because failing a public module binding tests in MacOS, and this is due to the change in import order for torch/distributed/fsdp/_common_utils.py. Since this original import would still work, we remove the changes in this file.
Test Plan: CI.
Differential Revision: D51825114
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115099
Approved by: https://github.com/wanchaol, https://github.com/fegin
**Dynamo**
We don't want setattr in the graph. Setting data has interesting implications on both aliasing and on the autograd engine.
The safe recipe is:
1) Disable grad
2) Call set_()
3) Manually lower the version counter on the object to hide it from the autograd engine
This is effectively the same exact thing as setting .data, and it composes properly with aot_autograd and inductor.
**aot_autograd**
For aot_autograd, there's another snag.
Specifically, when we invoke aot_autograd, we call `fake_mode.from_tensor()`, relying on memo to get the right tensor out. For .data mutations, this doesn't work, because the memoized fake_tensor is in the state it will be in at the end of the trace, not at the beginning. This means that the .data call is already applied, and the tensor shape (as in the case of these tests) mismatches. aot_autograd produces an invalid graph, with illegal calls like `torch.ops.aten.view.default(primals_2, [0])` where primals is actually sized `([6])` on input.
The new plan here is to:
1) Record tensor fakification policy in dynamo
2) provide a fresh fake mode to all backends
3) Invoke from_tensor with the stored policy to get fresh new fake tensors in aot_autograd
Pull Request resolved: https://github.com/pytorch/pytorch/pull/113080
Approved by: https://github.com/bdhirsh
Updated version of #108885 addressing the review. In this PR:
- We add a VT.can_reconstruct utility that checks if VT.reconstruct()
does something.
- If functools.wraps(fn) is passed a `fn` that either has a source or
has .can_reconstruct() == True, then we stash the source (or the VT)
- Later on, we use the source (or VT.reconstruct) to actually
reconstruct the object in codegen.
Test Plan:
- New tests
Pull Request resolved: https://github.com/pytorch/pytorch/pull/114279
Approved by: https://github.com/voznesenskym
This should be enough to get @voznesenskym 's FSDP branch to plumb `set_()` through AOTAutograd properly and have everything properly no-op out. Main changes are:
(1) graph break on `aten::set_.source_Tensor_storage_offset` (we could support it but it isn't needed, seems safer to graph break)
(2) Functionalization: add a "proper" functionalization kernel for `aten::set_.source_Tensor`. The previous one we had was codegen'd and it was wrong (it would just clone() and call set_(), which does not do the right thing). I also manually mark on the `FunctionalTensorWrapper` when a given tensor has been mutated by a `set_()` call.
(3) AOTAutograd: I added a new field, `InputAliasInfo.mutates_storage_metadata`, so we can distinguish between "regular" metadata mutations, and metadata mutations due to `set_()` calls. This is mainly because at runtime, one requires calling `as_strided_()` to fix up metadata, while the other requires calling `set_()`.
(4) Made AOTAutograd's detection for metadata mutations / set_() mutations smarter and detect no-ops (if the storage and metadata are all the same).
I also killed `was_updated()` and `was_metadata_updated()`, and replaced them with (existing) `has_data_mutation() ` and (new) `has_data_mutation()`, which can more accurately distinguish between data-mutation vs. `set_()` calls vs. metadata-mutation
**This PR is still silently correct in one case though**, which I'd like to discuss more. In particular, this example:
```
def f(x):
x_view = x.view(-1)
x.set_(torch.ones(2))
x_view.mul_(2)
return
```
If you have an input that experiences both a data-mutation **and** a `x_old.set_(x_new)` call, there are two cases:
(a) the data mutation happened on the storage of `x_new`. This case should be handled automatically: if x_new is a graph intermediate then we will functionalize the mutation. If x_new is a different graph input, then we will perform the usual `copy_()` on that other graph input
(b) the data mutation happened on the storage of `x_old`. This is more of a pain to handle, and doesn't currently work. At runtime, the right thing to do is probably something like:
```
def functionalized_f(x):
x_view = x.view(-1)
# set_() desugars into a no-op; later usages of x will use x_output
x_output = torch.ones(2)
# functionalize the mutation on x_view
x_view_updated = x.mul(2)
x_updated = x_view_updated.view(x.shape)
# x experienced TWO TYPES of mutations; a data mutation and a metatadata mutation
# We need to return both updated tensors in our graph
return x_updated, x_output
def runtime_wrapper(x):
x_data_mutation_result, x_set_mutation_result = compiled_graph(x)
# First, perform the data mutation on x's old storage
x.copy_(x_data_mutation_result)
# Then, swap out the storage of x with the new storage
x.set_(x_set_mutation_result)
```
There are two things that make this difficult to do though:
(1) Functionalization: the functionalization rule for `set_()` will fully throw away the old `FunctionalStorageImpl` on the graph input. So if there are any mutations to that `FunctionalStorageImpl` later on in the graph, the current graph input won't know about it. Maybe we can have a given `FunctionalTensorWrapper` remember all previous storages that it had, and track mutations on all of them - although this feels pretty complicated.
(2) AOTAutograd now needs to know that we might have *two* graph outputs that correspond to a single "mutated input", which is annoying.
It's worth pointing out that this issue is probably extremely unlikely for anyone to run into - can we just detect it and error? This feels slightly easier than solving it, although not significantly easier. We would still need `FunctionalTensorWrapper` to keep track of mutations on any of its "previous" storages, so it can report this info back to AOTAutograd so we can raise an error.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/111554
Approved by: https://github.com/ezyang
ghstack dependencies: #113926
With this PR it is possible to differentiate through NumPy code modulo
the usual caveats that apply to differentiation:
- That there are no graphbreaks
- That the decomposition in `torch._numpy` is differentiable
@ev-br and I were somewhat careful to achieve the second point, but
it is not tested though and through, so YMMV
Pull Request resolved: https://github.com/pytorch/pytorch/pull/114608
Approved by: https://github.com/voznesenskym
Summary:
The primary problem we are setting out to solve here is fake tensor freshness. Before this PR, fake tensors after dynamo represented fake tensors *at the end* of trace, so subsequent retraces like aot_autograd would start off with fake tensors in the wrong (end result) state, rather than their expected fresh state. The solution here is to start a fresh fake mode, and re-fakify the tensors. The nuance comes from ensuring that symbols are uniformly created for the symbolic sizes and strides of the tensor.
This PR is the result of *a lot* of back and forth with ezyang and eellison. Initially, the first pass at this was not super different from what we have in the PR - the broad strokes were the same:
1) We cache source->symbol in shape_env
2) We pass policy objects around, stored at dynamo fakificaiton time, and reused for later fakification
3) We create a new fake mode for backends
(from https://github.com/pytorch/pytorch/pull/113605/files)
This is ugly, and has some layering violations. We detoured our decision making through a few other alternatives. Immutable/mutable fake tensor mode was the most interesting alternative, https://github.com/pytorch/pytorch/pull/113653, and was struck down on concerns of complexity in fake mode combined with it not covering all edge cases. We also detoured on what to do about tensor memoization returning back potentially different tensors than requested, and if that was an anti pattern (it is) we want to hack in with the symbol cache (we don't).
We went back to the drawing board here, but with a few concessions:
1) the cache for source->symbol must live outside of shape_env, for both lifecycle, and layering reasons
2) A good amount of work needs to be done to pipe policy around fake_mode and meta_utils correctly, to cover all the cases (ezyang did this)
cc penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx chenyang78 aakhundov kadeng
imported-using-ghimport
Test Plan: Imported from OSS
Reviewed By: huydhn, Chillee
Differential Revision: D51566250
Pulled By: voznesenskym
Pull Request resolved: https://github.com/pytorch/pytorch/pull/114526
Approved by: https://github.com/Chillee, https://github.com/huydhn
