Enable cross entropy loss for xla autocast with FP32 precision by avizon-aws · Pull Request #7992 · pytorch/xla

avizon-aws · 2024-09-11T17:48:27Z

There are many operators in XLA autocast that have been commented, but these operators are casted in the GPU, in order to maintain consistency, we need to support these operators as well. For cross_entropy_loss, it is currently commented in the xla autocast, so there will be no casting occuring, and it will execute based on its input’s dtype.

The output type is bf16, which is expected because linear layer is specified in xla autocast. loss dtype is fp32, which is correct, but there’s a catch, there was no autocasting done for the crossEntropyLoss, the reason the dtype is FP32 is because of the target’s dtype, which is FP32. There is a multiplication which happens in crossentropyloss between the generated output and the target, all the exponentiation/log etc. is done in BF16, but only because of the final multiplication, we get the result in FP32, because it casts to the higher precision (FP32). This is not the expected behavior, all the exponentiation/logs i.e. all ops related to crossentropyloss should be executed in FP32, the reason it is not happening is because crossentropyloss is not specified in xla autocast. This finding is based after detailed analysis of the HLO outputs which is attached below.

Before this change:
Exp1.

device = 'xla' # Get the XLA device (e.g., TPU or GPU)
model = torch.nn.Linear(10, 10).to(device)
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)
data = torch.randn(16, 10).to(torch.bfloat16).(to(device)
target = torch.randn(16, 10).to(device)
print(device, torch.__version__)
for epoch in range(1):
    optimizer.zero_grad()
    # debugpy.breakpoint() 
    with torch.autocast('xla'):
        output = model(data)
        loss = torch.nn.CrossEntropyLoss()(output, target)
        print(output.dtype, loss.dtype, target.dtype)
    # loss.backward()
    optimizer.step()
    print(f"Epoch {epoch}, Loss: {loss.item()}")

HLO:

ENTRY %SyncTensorsGraph.62 (p0.1: f32[], p1.2: f32[16,10], p2.3: f32[10], p3.12: f32[10,10], p4.22: f32[16,10]) -> (f32[]) {
  %p4.22 = f32[16,10]{1,0} parameter(4), frontend_attributes={neff_input_names="input4"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %convert.23 = bf16[16,10]{1,0} convert(f32[16,10]{1,0} %p4.22), metadata={op_type="xla__cast" op_name="xla__cast"}
  %p3.12 = f32[10,10]{1,0} parameter(3), frontend_attributes={neff_input_names="input3"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %custom-call.2 = f32[10,10]{1,0} custom-call(f32[10,10]{1,0} %p3.12), custom_call_target="AwsNeuronTransferWithStaticRing", api_version=API_VERSION_UNSPECIFIED, metadata={op_type="xla___op_TransferWithStaticRingTransfer" op_name="xla___op_TransferWithStaticRingTransfer"}
  %convert.20 = bf16[10,10]{1,0} convert(f32[10,10]{1,0} %custom-call.2), metadata={op_type="xla__cast" op_name="xla__cast"}
  %transpose.21 = bf16[10,10]{0,1} transpose(bf16[10,10]{1,0} %convert.20), dimensions={1,0}, metadata={op_type="aten__permute" op_name="aten__permute"}
  %dot.24 = bf16[16,10]{1,0} dot(bf16[16,10]{1,0} %convert.23, bf16[10,10]{0,1} %transpose.21), lhs_contracting_dims={1}, rhs_contracting_dims={0}, metadata={op_type="aten__addmm" op_name="aten__addmm"}
  %p2.3 = f32[10]{0} parameter(2), frontend_attributes={neff_input_names="input2"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %custom-call.3 = f32[10]{0} custom-call(f32[10]{0} %p2.3), custom_call_target="AwsNeuronTransferWithStaticRing", api_version=API_VERSION_UNSPECIFIED, metadata={op_type="xla___op_TransferWithStaticRingTransfer" op_name="xla___op_TransferWithStaticRingTransfer"}
  %convert.11 = bf16[10]{0} convert(f32[10]{0} %custom-call.3), metadata={op_type="xla__cast" op_name="xla__cast"}
  %broadcast.28 = bf16[16,10]{1,0} broadcast(bf16[10]{0} %convert.11), dimensions={1}, metadata={op_type="aten__addmm" op_name="aten__addmm"}
  %add.29 = bf16[16,10]{1,0} add(bf16[16,10]{1,0} %dot.24, bf16[16,10]{1,0} %broadcast.28), metadata={op_type="aten__addmm" op_name="aten__addmm"}
  **%constant.32 = bf16[] constant(-inf), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %reduce.37 = bf16[16]{0} reduce(bf16[16,10]{1,0} %add.29, bf16[] %constant.32), dimensions={1}, to_apply=%MaxComputation.33, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %broadcast.38 = bf16[16,10]{1,0} broadcast(bf16[16]{0} %reduce.37), dimensions={0}, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %subtract.39 = bf16[16,10]{1,0} subtract(bf16[16,10]{1,0} %add.29, bf16[16,10]{1,0} %broadcast.38), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %exponential.40 = bf16[16,10]{1,0} exponential(bf16[16,10]{1,0} %subtract.39), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %constant.41 = bf16[] constant(0), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %reduce.46 = bf16[16]{0} reduce(bf16[16,10]{1,0} %exponential.40, bf16[] %constant.41), dimensions={1}, to_apply=%AddComputation.42, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %log.47 = bf16[16]{0} log(bf16[16]{0} %reduce.46), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %broadcast.48 = bf16[16,10]{1,0} broadcast(bf16[16]{0} %log.47), dimensions={0}, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %subtract.49 = bf16[16,10]{1,0} subtract(bf16[16,10]{1,0} %subtract.39, bf16[16,10]{1,0} %broadcast.48), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %convert.50 = f32[16,10]{1,0} convert(bf16[16,10]{1,0} %subtract.49), metadata={op_type="aten__mul" op_name="aten__mul"}
  %p1.2 = f32[16,10]{1,0} parameter(1), frontend_attributes={neff_input_names="input1"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %multiply.51 = f32[16,10]{1,0} multiply(f32[16,10]{1,0} %convert.50, f32[16,10]{1,0} %p1.2), metadata={op_type="aten__mul" op_name="aten__mul"}**
  %constant.52 = f32[] constant(0), metadata={op_type="aten__sum" op_name="aten__sum"}
  %reduce.58 = f32[] reduce(f32[16,10]{1,0} %multiply.51, f32[] %constant.52), dimensions={0,1}, to_apply=%AddComputation.54, metadata={op_type="aten__sum" op_name="aten__sum"}
  %negate.59 = f32[] negate(f32[] %reduce.58), metadata={op_type="aten__neg" op_name="aten__neg"}
  %p0.1 = f32[] parameter(0), frontend_attributes={neff_input_names="input0"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %divide.60 = f32[] divide(f32[] %negate.59, f32[] %p0.1), metadata={op_type="aten__div" op_name="aten__div"}
  ROOT %tuple.61 = (f32[]) tuple(f32[] %divide.60), frontend_attributes={neff_output_names="output0"}
}

Exp2.
The target dtype if also bf16 in this case. This experiment was done to prove that the dtype of the target was the true cause of the FP32 output as shown below.

Code change from previous experiment:
target = torch.randn(16, 10).to(torch.bfloat16).to(device)

HLO

ENTRY %SyncTensorsGraph.61 (p0.1: bf16[], p1.2: bf16[16,10], p2.3: f32[10], p3.12: f32[10,10], p4.22: f32[16,10]) -> (bf16[]) {
  %p4.22 = f32[16,10]{1,0} parameter(4), frontend_attributes={neff_input_names="input4"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %convert.23 = bf16[16,10]{1,0} convert(f32[16,10]{1,0} %p4.22), metadata={op_type="xla__cast" op_name="xla__cast"}
  %p3.12 = f32[10,10]{1,0} parameter(3), frontend_attributes={neff_input_names="input3"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %custom-call.2 = f32[10,10]{1,0} custom-call(f32[10,10]{1,0} %p3.12), custom_call_target="AwsNeuronTransferWithStaticRing", api_version=API_VERSION_UNSPECIFIED, metadata={op_type="xla___op_TransferWithStaticRingTransfer" op_name="xla___op_TransferWithStaticRingTransfer"}
  %convert.20 = bf16[10,10]{1,0} convert(f32[10,10]{1,0} %custom-call.2), metadata={op_type="xla__cast" op_name="xla__cast"}
  %transpose.21 = bf16[10,10]{0,1} transpose(bf16[10,10]{1,0} %convert.20), dimensions={1,0}, metadata={op_type="aten__permute" op_name="aten__permute"}
  %dot.24 = bf16[16,10]{1,0} dot(bf16[16,10]{1,0} %convert.23, bf16[10,10]{0,1} %transpose.21), lhs_contracting_dims={1}, rhs_contracting_dims={0}, metadata={op_type="aten__addmm" op_name="aten__addmm"}
  %p2.3 = f32[10]{0} parameter(2), frontend_attributes={neff_input_names="input2"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %custom-call.3 = f32[10]{0} custom-call(f32[10]{0} %p2.3), custom_call_target="AwsNeuronTransferWithStaticRing", api_version=API_VERSION_UNSPECIFIED, metadata={op_type="xla___op_TransferWithStaticRingTransfer" op_name="xla___op_TransferWithStaticRingTransfer"}
  %convert.11 = bf16[10]{0} convert(f32[10]{0} %custom-call.3), metadata={op_type="xla__cast" op_name="xla__cast"}
  %broadcast.28 = bf16[16,10]{1,0} broadcast(bf16[10]{0} %convert.11), dimensions={1}, metadata={op_type="aten__addmm" op_name="aten__addmm"}
  %add.29 = bf16[16,10]{1,0} add(bf16[16,10]{1,0} %dot.24, bf16[16,10]{1,0} %broadcast.28), metadata={op_type="aten__addmm" op_name="aten__addmm"}
  %constant.32 = bf16[] constant(-inf), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %reduce.37 = bf16[16]{0} reduce(bf16[16,10]{1,0} %add.29, bf16[] %constant.32), dimensions={1}, to_apply=%MaxComputation.33, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %broadcast.38 = bf16[16,10]{1,0} broadcast(bf16[16]{0} %reduce.37), dimensions={0}, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %subtract.39 = bf16[16,10]{1,0} subtract(bf16[16,10]{1,0} %add.29, bf16[16,10]{1,0} %broadcast.38), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %exponential.40 = bf16[16,10]{1,0} exponential(bf16[16,10]{1,0} %subtract.39), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %constant.41 = bf16[] constant(0), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %reduce.46 = bf16[16]{0} reduce(bf16[16,10]{1,0} %exponential.40, bf16[] %constant.41), dimensions={1}, to_apply=%AddComputation.42, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %log.47 = bf16[16]{0} log(bf16[16]{0} %reduce.46), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %broadcast.48 = bf16[16,10]{1,0} broadcast(bf16[16]{0} %log.47), dimensions={0}, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %subtract.49 = bf16[16,10]{1,0} subtract(bf16[16,10]{1,0} %subtract.39, bf16[16,10]{1,0} %broadcast.48), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %p1.2 = bf16[16,10]{1,0} parameter(1), frontend_attributes={neff_input_names="input1"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %multiply.50 = bf16[16,10]{1,0} multiply(bf16[16,10]{1,0} %subtract.49, bf16[16,10]{1,0} %p1.2), metadata={op_type="aten__mul" op_name="aten__mul"}
  %constant.51 = bf16[] constant(0), metadata={op_type="aten__sum" op_name="aten__sum"}
  %reduce.57 = bf16[] reduce(bf16[16,10]{1,0} %multiply.50, bf16[] %constant.51), dimensions={0,1}, to_apply=%AddComputation.53, metadata={op_type="aten__sum" op_name="aten__sum"}
  %negate.58 = bf16[] negate(bf16[] %reduce.57), metadata={op_type="aten__neg" op_name="aten__neg"}
  %p0.1 = bf16[] parameter(0), frontend_attributes={neff_input_names="input0"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %divide.59 = bf16[] divide(bf16[] %negate.58, bf16[] %p0.1), metadata={op_type="aten__div" op_name="aten__div"}
  ROOT %tuple.60 = (bf16[]) tuple(bf16[] %divide.59), frontend_attributes={neff_output_names="output0"}
}

After uncommenting the CrossEntropyLoss in the XLA autocast as done in this PR:

Exp3:
The input and target are in FP32, so the output of the linear layer will be in BF16, and then it should be upcasted to FP32 for the Crossentropyloss as seen in the HLO.

  %p4.8 = f32[16,10]{1,0} parameter(4), frontend_attributes={neff_input_names="input4"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %convert.9 = bf16[16,10]{1,0} convert(f32[16,10]{1,0} %p4.8), metadata={op_type="xla__cast" op_name="xla__cast"}
  %p3.5 = f32[10,10]{1,0} parameter(3), frontend_attributes={neff_input_names="input3"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %convert.6 = bf16[10,10]{1,0} convert(f32[10,10]{1,0} %p3.5), metadata={op_type="xla__cast" op_name="xla__cast"}
  %transpose.7 = bf16[10,10]{0,1} transpose(bf16[10,10]{1,0} %convert.6), dimensions={1,0}, metadata={op_type="aten__permute" op_name="aten__permute"}
  %dot.10 = bf16[16,10]{1,0} dot(bf16[16,10]{1,0} %convert.9, bf16[10,10]{0,1} %transpose.7), lhs_contracting_dims={1}, rhs_contracting_dims={0}, metadata={op_type="aten__addmm" op_name="aten__addmm"}
  %p2.3 = f32[10]{0} parameter(2), frontend_attributes={neff_input_names="input2"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %convert.4 = bf16[10]{0} convert(f32[10]{0} %p2.3), metadata={op_type="xla__cast" op_name="xla__cast"}
  %broadcast.14 = bf16[16,10]{1,0} broadcast(bf16[10]{0} %convert.4), dimensions={1}, metadata={op_type="aten__addmm" op_name="aten__addmm"}
  %add.15 = bf16[16,10]{1,0} add(bf16[16,10]{1,0} %dot.10, bf16[16,10]{1,0} %broadcast.14), metadata={op_type="aten__addmm" op_name="aten__addmm"}
  %convert.16 = f32[16,10]{1,0} convert(bf16[16,10]{1,0} %add.15), metadata={op_type="xla__cast" op_name="xla__cast"}
  %constant.17 = f32[] constant(-inf), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %reduce.22 = f32[16]{0} reduce(f32[16,10]{1,0} %convert.16, f32[] %constant.17), dimensions={1}, to_apply=%MaxComputation.18, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %broadcast.23 = f32[16,10]{1,0} broadcast(f32[16]{0} %reduce.22), dimensions={0}, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %subtract.24 = f32[16,10]{1,0} subtract(f32[16,10]{1,0} %convert.16, f32[16,10]{1,0} %broadcast.23), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %exponential.25 = f32[16,10]{1,0} exponential(f32[16,10]{1,0} %subtract.24), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %constant.26 = f32[] constant(0), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %reduce.31 = f32[16]{0} reduce(f32[16,10]{1,0} %exponential.25, f32[] %constant.26), dimensions={1}, to_apply=%AddComputation.27, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %log.32 = f32[16]{0} log(f32[16]{0} %reduce.31), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %broadcast.33 = f32[16,10]{1,0} broadcast(f32[16]{0} %log.32), dimensions={0}, metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %subtract.34 = f32[16,10]{1,0} subtract(f32[16,10]{1,0} %subtract.24, f32[16,10]{1,0} %broadcast.33), metadata={op_type="aten__log_softmax" op_name="aten__log_softmax"}
  %p1.2 = f32[16,10]{1,0} parameter(1), frontend_attributes={neff_input_names="input1"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %multiply.35 = f32[16,10]{1,0} multiply(f32[16,10]{1,0} %subtract.34, f32[16,10]{1,0} %p1.2), metadata={op_type="aten__mul" op_name="aten__mul"}
  %reduce.42 = f32[] reduce(f32[16,10]{1,0} %multiply.35, f32[] %constant.26), dimensions={0,1}, to_apply=%AddComputation.38, metadata={op_type="aten__sum" op_name="aten__sum"}
  %negate.43 = f32[] negate(f32[] %reduce.42), metadata={op_type="aten__neg" op_name="aten__neg"}
  %p0.1 = f32[] parameter(0), frontend_attributes={neff_input_names="input0"}, metadata={op_type="xla__device_data" op_name="xla__device_data"}
  %divide.44 = f32[] divide(f32[] %negate.43, f32[] %p0.1), metadata={op_type="aten__div" op_name="aten__div"}
  ROOT %tuple.45 = (f32[]) tuple(f32[] %divide.44), frontend_attributes={neff_output_names="output0"}
}

The executions of the operators are as expected.

jeffhataws · 2024-09-28T16:14:46Z

@avizon-aws Need cherry-pick to master

jeffhataws · 2024-09-28T16:20:04Z

Also request cherry-pick to 2.5 here #7977 (comment)

avizon-aws · 2024-09-28T23:28:51Z

Created PR to master: #8094

#8094)

avizon-aws added 3 commits September 4, 2024 17:21

uncommented cross entropy loss for xla autocast

9c9cbf5

added unit test for cross entropy loss autocast fp32

2635e98

fixed formatting

330158e

JackCaoG requested a review from lsy323 September 12, 2024 00:36

jeffhataws requested review from amithrm and jeffhataws September 23, 2024 22:13

lsy323 approved these changes Sep 24, 2024

View reviewed changes

JackCaoG merged commit e438a5b into r2.1_aws_neuron Sep 26, 2024

JackCaoG deleted the autocast_bf16_neuron branch September 26, 2024 16:11

jeffhataws mentioned this pull request Sep 28, 2024

2.5 backport PR request list #7977

Closed

avizon-aws added a commit that referenced this pull request Sep 28, 2024

Enable cross entropy loss for xla autocast with FP32 precision (#7992)

b21d2c0

lsy323 pushed a commit that referenced this pull request Sep 30, 2024

Enable cross entropy loss for xla autocast with FP32 precision (#7992) (

940bee4

#8094)

ManfeiBai pushed a commit that referenced this pull request Oct 1, 2024

Enable cross entropy loss for xla autocast with FP32 precision (#7992) (

5041ceb

#8094)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Enable cross entropy loss for xla autocast with FP32 precision#7992

Enable cross entropy loss for xla autocast with FP32 precision#7992
JackCaoG merged 3 commits intor2.1_aws_neuronfrom
autocast_bf16_neuron

avizon-aws commented Sep 11, 2024

Uh oh!

jeffhataws commented Sep 28, 2024

Uh oh!

jeffhataws commented Sep 28, 2024

Uh oh!

avizon-aws commented Sep 28, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

avizon-aws commented Sep 11, 2024

Uh oh!

jeffhataws commented Sep 28, 2024

Uh oh!

jeffhataws commented Sep 28, 2024

Uh oh!

avizon-aws commented Sep 28, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants