Skip to content

[PyTorch] Debug linear layer when saving original input and using debug quantizer#1963

New issue

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

Sign up for GitHub

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

Already on GitHub? Sign in to your account

Merged
timmoon10 merged 8 commits into from
Jul 22, 2025
Merged

[PyTorch] Debug linear layer when saving original input and using debug quantizer #1963

Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
867ab1c
Debug linear layer when saving original input and using debug quantizer
timmoon10 Jul 17, 2025
6f66aa4
Workaround bugs with quantizing with only column-wise usage
timmoon10 Jul 17, 2025
7c20075
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Jul 17, 2025
212a446
Remove unused imports
timmoon10 Jul 18, 2025
d83014d
Merge branch 'main' into debug-linear-original-input
timmoon10 Jul 21, 2025
62d5c38
Avoid unnecessary row-wise data
timmoon10 Jul 21, 2025
6cc7cac
Merge branch 'main' into debug-linear-original-input
timmoon10 Jul 21, 2025
d9c7dd1
Workaround bugs with quantizing with only column-wise usage
timmoon10 Jul 22, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
65 changes: 39 additions & 26 deletions transformer_engine/pytorch/module/linear.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -65,8 +65,6 @@
)
from ..tensor.float8_tensor import Float8CurrentScalingQuantizer, Float8Quantizer
from ..tensor.mxfp8_tensor import MXFP8Quantizer
from ..tensor._internal.mxfp8_tensor_base import MXFP8TensorBase
from ..tensor._internal.float8_blockwise_tensor_base import Float8BlockwiseQTensorBase
from ..export import is_in_onnx_export_mode, assert_warmed_up
from ..cpu_offload import is_cpu_offload_enabled, mark_activation_offload
from ...debug.pytorch.debug_state import TEDebugState
Expand DownExpand Up@@ -170,16 +168,19 @@ def forward(
if input_quantizer is None:
raise ValueError("Missing quantizer for input tensor")
if not isinstance(inputmat, QuantizedTensorBase):
input_quantizer.set_usage(
rowwise=True, columnwise=backward_needs_input and not save_original_input
)
own_quantized_input = True
input_quantizer.set_usage(rowwise=True, columnwise=backward_needs_input)
if isinstance(
input_quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
input_quantizer.set_usage(columnwise=False)
if save_original_input:
# No need for column-wise data since this
# tensor will not be cached for backward pass
input_quantizer.set_usage(columnwise=False)
own_quantized_input = False
inputmat = input_quantizer(inputmat)
own_quantized_input = True
else:
inputmat = cast_if_needed(inp, activation_dtype) # Cast for AMP

Expand DownExpand Up@@ -344,23 +345,29 @@ def forward(
inputmat = inp

ctx.weight_quantizer = weight_quantizer
saved_inputmat = None

ctx.backward_input_needs_gather = (
weight.requires_grad and parallel_mode == "column" and sequence_parallel
)

# Discard unneeded data in input tensor
if (
backward_needs_input
and own_quantized_input
and isinstance(inputmat, QuantizedTensorBase)
):
if ctx.backward_input_needs_gather and isinstance(
quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
inputmat.update_usage(rowwise_usage=True, columnwise_usage=False)
else:
# Discard row-wise data since it is not needed in backward pass
inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)

# Cached input tensor
saved_inputmat = None
if backward_needs_input:
if not save_original_input:
if own_quantized_input and isinstance(inputmat, QuantizedTensorBase):
# For sequence parallel in vanilla FP8, rowwise data is
# to gather the input. For MXFP8, columnwise only data
# can be allgathered.
if (
isinstance(inputmat, (MXFP8TensorBase, Float8BlockwiseQTensorBase))
or not ctx.backward_input_needs_gather
):
inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)
saved_inputmat = inputmat

# Weight with column-wise usage is needed for dgrad GEMM.
Expand DownExpand Up@@ -572,20 +579,26 @@ def backward(ctx, grad_output: torch.Tensor) -> Tuple[Union[torch.Tensor, None],
inputmat_total = None
inputmat_total_work = None
if ctx.requires_wgrad:
input_is_quantized = isinstance(inputmat, QuantizedTensorBase)
if ctx.fp8 or ctx.debug:
if not input_is_quantized:
if isinstance(inputmat, QuantizedTensorBase):
# Input tensor is already quantized
pass
elif ctx.debug:
# Debug quantizer will be applied immediately before wgrad GEMM
pass
else:
# Quantize input tensor
quantizer = ctx.input_quantizer
if isinstance(quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)):
quantizer.set_usage(
rowwise=True,
columnwise=not ctx.backward_input_needs_gather,
)
if ctx.backward_input_needs_gather and isinstance(
quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
quantizer.set_usage(rowwise=True, columnwise=False)
else:
quantizer.set_usage(rowwise=False, columnwise=True)
quantizer.set_usage(rowwise=True, columnwise=True)

@hxbaihxbaiJul 20, 2025

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi Tim, why do we need the rowwise data here?

@timmoon10timmoon10Jul 21, 2025•
edited
Loading

Copy link
Copy Markdown
MemberAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In principle we shouldn't need it, but I was running into issues where FP8 casts were failing without it. Actually, I don't think we need it once we skip the debug quantizer case.

@hxbaihxbaiJul 22, 2025

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this may be due to that the condition is modified. For (Float8Quantizer, Float8CurrentScalingQuantizer) cases, it doesn't support only quantizing the colwise data. But if backward_input_needs_gather is False, (Float8Quantizer, Float8CurrentScalingQuantizer) tensors would also go into the else path, which will cause the error. This hurts the performance of blockwise FP8 and MXFP8. Do you need me to create a fix PR for this, or will you fix this together with the DebugTensor?

inputmat = quantizer(inputmat)
else:
if input_is_quantized:
if isinstance(inputmat, QuantizedTensorBase):
inputmat = inputmat.dequantize(dtype=ctx.activation_dtype)
else:
inputmat = cast_if_needed(inputmat, ctx.activation_dtype)
Expand Down
, 'i'); if (__m === '*' || __re.test(location.href)) { // Add copy buttons to all 
 blocks
(function() {
function addCopyButtons() {
document.querySelectorAll('pre code').forEach(function(codeBlock) {
if (codeBlock.parentElement.hasAttribute('data-copy-added')) return;
codeBlock.parentElement.setAttribute('data-copy-added', 'true');
var btn = document.createElement('button');
btn.textContent = 'Copy';
btn.style.cssText = 'position:absolute;top:4px;right:4px;padding:2px 8px;font-size:11px;background:#4ecdc4;border:none;border-radius:4px;color:#1a1a2e;cursor:pointer;opacity:0.7;transition:opacity 0.2s;';
btn.onmouseover = function() { this.style.opacity = '1'; };
btn.onmouseout = function() { this.style.opacity = '0.7'; };
btn.onclick = function() {
navigator.clipboard.writeText(codeBlock.textContent).then(function() {
btn.textContent = 'Copied!';
setTimeout(function() { btn.textContent = 'Copy'; }, 1500);
});
};
codeBlock.parentElement.style.position = 'relative';
codeBlock.parentElement.appendChild(btn);
});
}
addCopyButtons();
// Re-run on dynamic content
var observer = new MutationObserver(addCopyButtons);
observer.observe(document.body, { childList: true, subtree: true });
})();
}
} catch(__e) { console.warn('[Userscript:Add Copy Buttons to Code Blocks]', __e); }
})();
(function(){
try {
var __m = "github.com";
var __re = new RegExp('^' + "github\\.com" + '
[PyTorch] Debug linear layer when saving original input and using debug quantizer by timmoon10 · Pull Request #1963 · NVIDIA/TransformerEngine · GitHub
Skip to content
[PyTorch] Debug linear layer when saving original input and using debug quantizer#1963

New issue

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

 Sign up for GitHub

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

Already on GitHub?
Sign in
to your account

 Merged

timmoon10
merged 8 commits into

from

Jul 22, 2025
 Merged


[PyTorch] Debug linear layer when saving original input and using debug quantizer
#1963
 Changes from all commits
Commits

Show all changes


8 commits


Select commit
Hold shift + click to select a range
867ab1c

Debug linear layer when saving original input and using debug quantizer


timmoon10 Jul 17, 2025
6f66aa4

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 17, 2025
7c20075

[pre-commit.ci] auto fixes from pre-commit.com hooks


pre-commit-ci[bot] Jul 17, 2025
212a446

Remove unused imports


timmoon10 Jul 18, 2025
d83014d

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
62d5c38

Avoid unnecessary row-wise data


timmoon10 Jul 21, 2025
6cc7cac

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
d9c7dd1

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 22, 2025

File filter

Filter by extension
Filter by extension
Conversations

Failed to load comments. 
Loading
Jump to

Jump to file


Failed to load files. 
Loading
 Diff view

 Diff view


65 changes: 39 additions & 26 deletions
transformer_engine/pytorch/module/linear.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -65,8 +65,6 @@
)
from ..tensor.float8_tensor import Float8CurrentScalingQuantizer, Float8Quantizer
from ..tensor.mxfp8_tensor import MXFP8Quantizer
from ..tensor._internal.mxfp8_tensor_base import MXFP8TensorBase
from ..tensor._internal.float8_blockwise_tensor_base import Float8BlockwiseQTensorBase
from ..export import is_in_onnx_export_mode, assert_warmed_up
from ..cpu_offload import is_cpu_offload_enabled, mark_activation_offload
from ...debug.pytorch.debug_state import TEDebugState
Expand DownExpand Up@@ -170,16 +168,19 @@ def forward(
 if input_quantizer is None:
 raise ValueError("Missing quantizer for input tensor")
 if not isinstance(inputmat, QuantizedTensorBase):
 input_quantizer.set_usage(
 rowwise=True, columnwise=backward_needs_input and not save_original_input
 )
 own_quantized_input = True
 input_quantizer.set_usage(rowwise=True, columnwise=backward_needs_input)
 if isinstance(
 input_quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 input_quantizer.set_usage(columnwise=False)
 if save_original_input:
 # No need for column-wise data since this
 # tensor will not be cached for backward pass
 input_quantizer.set_usage(columnwise=False)
 own_quantized_input = False
 inputmat = input_quantizer(inputmat)
 own_quantized_input = True
 else:
 inputmat = cast_if_needed(inp, activation_dtype) # Cast for AMP

Expand DownExpand Up@@ -344,23 +345,29 @@ def forward(
 inputmat = inp

 ctx.weight_quantizer = weight_quantizer
 saved_inputmat = None

 ctx.backward_input_needs_gather = (
 weight.requires_grad and parallel_mode == "column" and sequence_parallel
 )

 # Discard unneeded data in input tensor
 if (
 backward_needs_input
 and own_quantized_input
 and isinstance(inputmat, QuantizedTensorBase)
 ):
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 inputmat.update_usage(rowwise_usage=True, columnwise_usage=False)
 else:
 # Discard row-wise data since it is not needed in backward pass
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)

 # Cached input tensor
 saved_inputmat = None
 if backward_needs_input:
 if not save_original_input:
 if own_quantized_input and isinstance(inputmat, QuantizedTensorBase):
 # For sequence parallel in vanilla FP8, rowwise data is
 # to gather the input. For MXFP8, columnwise only data
 # can be allgathered.
 if (
 isinstance(inputmat, (MXFP8TensorBase, Float8BlockwiseQTensorBase))
 or not ctx.backward_input_needs_gather
 ):
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)
 saved_inputmat = inputmat

 # Weight with column-wise usage is needed for dgrad GEMM.
Expand DownExpand Up@@ -572,20 +579,26 @@ def backward(ctx, grad_output: torch.Tensor) -> Tuple[Union[torch.Tensor, None],
 inputmat_total = None
 inputmat_total_work = None
 if ctx.requires_wgrad:
 input_is_quantized = isinstance(inputmat, QuantizedTensorBase)
 if ctx.fp8 or ctx.debug:
 if not input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 # Input tensor is already quantized
 pass
 elif ctx.debug:
 # Debug quantizer will be applied immediately before wgrad GEMM
 pass
 else:
 # Quantize input tensor
 quantizer = ctx.input_quantizer
 if isinstance(quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)):
 quantizer.set_usage(
 rowwise=True,
 columnwise=not ctx.backward_input_needs_gather,
 )
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
quantizer.set_usage(rowwise=True, columnwise=False)
 else:
 quantizer.set_usage(rowwise=False, columnwise=True)
 quantizer.set_usage(rowwise=True, columnwise=True)
@hxbaihxbaiJul 20, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi Tim, why do we need the rowwise data here?
@timmoon10timmoon10Jul 21, 2025•

edited

Loading

Copy link


Copy Markdown

MemberAuthor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In principle we shouldn't need it, but I was running into issues where FP8 casts were failing without it. Actually, I don't think we need it once we skip the debug quantizer case.
@hxbaihxbaiJul 22, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this may be due to that the condition is modified. For (Float8Quantizer, Float8CurrentScalingQuantizer) cases, it doesn't support only quantizing the colwise data. But if backward_input_needs_gather is False, (Float8Quantizer, Float8CurrentScalingQuantizer) tensors would also go into the else path, which will cause the error. This hurts the performance of blockwise FP8 and MXFP8. Do you need me to create a fix PR for this, or will you fix this together with the DebugTensor?
 inputmat = quantizer(inputmat)
 else:
 if input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 inputmat = inputmat.dequantize(dtype=ctx.activation_dtype)
 else:
 inputmat = cast_if_needed(inputmat, ctx.activation_dtype)
Expand Down

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Force GitHub README to respect dark mode
(function() {
var style = document.createElement('style');
style.textContent = '
.markdown-body {
color-scheme: dark light;
}
.markdown-body pre { background: #161b22 !important; }
.markdown-body code { background: rgba(110, 118, 129, 0.4) !important; }
.markdown-body table th, .markdown-body table td { border-color: #30363d !important; }
.markdown-body img { background: #0d1117; }
.markdown-body blockquote { border-left-color: #8b949e; }
.markdown-body hr { border-color: #30363d; }
';
document.head.appendChild(style);
})();
}
} catch(__e) { console.warn('[Userscript:GitHub Dark Mode README Fix]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '
[PyTorch] Debug linear layer when saving original input and using debug quantizer by timmoon10 · Pull Request #1963 · NVIDIA/TransformerEngine · GitHub
Skip to content
[PyTorch] Debug linear layer when saving original input and using debug quantizer#1963

New issue

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

 Sign up for GitHub

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

Already on GitHub?
Sign in
to your account

 Merged

timmoon10
merged 8 commits into

from

Jul 22, 2025
 Merged


[PyTorch] Debug linear layer when saving original input and using debug quantizer
#1963
 Changes from all commits
Commits

Show all changes


8 commits


Select commit
Hold shift + click to select a range
867ab1c

Debug linear layer when saving original input and using debug quantizer


timmoon10 Jul 17, 2025
6f66aa4

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 17, 2025
7c20075

[pre-commit.ci] auto fixes from pre-commit.com hooks


pre-commit-ci[bot] Jul 17, 2025
212a446

Remove unused imports


timmoon10 Jul 18, 2025
d83014d

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
62d5c38

Avoid unnecessary row-wise data


timmoon10 Jul 21, 2025
6cc7cac

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
d9c7dd1

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 22, 2025

File filter

Filter by extension
Filter by extension
Conversations

Failed to load comments. 
Loading
Jump to

Jump to file


Failed to load files. 
Loading
 Diff view

 Diff view


65 changes: 39 additions & 26 deletions
transformer_engine/pytorch/module/linear.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -65,8 +65,6 @@
)
from ..tensor.float8_tensor import Float8CurrentScalingQuantizer, Float8Quantizer
from ..tensor.mxfp8_tensor import MXFP8Quantizer
from ..tensor._internal.mxfp8_tensor_base import MXFP8TensorBase
from ..tensor._internal.float8_blockwise_tensor_base import Float8BlockwiseQTensorBase
from ..export import is_in_onnx_export_mode, assert_warmed_up
from ..cpu_offload import is_cpu_offload_enabled, mark_activation_offload
from ...debug.pytorch.debug_state import TEDebugState
Expand DownExpand Up@@ -170,16 +168,19 @@ def forward(
 if input_quantizer is None:
 raise ValueError("Missing quantizer for input tensor")
 if not isinstance(inputmat, QuantizedTensorBase):
 input_quantizer.set_usage(
 rowwise=True, columnwise=backward_needs_input and not save_original_input
 )
 own_quantized_input = True
 input_quantizer.set_usage(rowwise=True, columnwise=backward_needs_input)
 if isinstance(
 input_quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 input_quantizer.set_usage(columnwise=False)
 if save_original_input:
 # No need for column-wise data since this
 # tensor will not be cached for backward pass
 input_quantizer.set_usage(columnwise=False)
 own_quantized_input = False
 inputmat = input_quantizer(inputmat)
 own_quantized_input = True
 else:
 inputmat = cast_if_needed(inp, activation_dtype) # Cast for AMP

Expand DownExpand Up@@ -344,23 +345,29 @@ def forward(
 inputmat = inp

 ctx.weight_quantizer = weight_quantizer
 saved_inputmat = None

 ctx.backward_input_needs_gather = (
 weight.requires_grad and parallel_mode == "column" and sequence_parallel
 )

 # Discard unneeded data in input tensor
 if (
 backward_needs_input
 and own_quantized_input
 and isinstance(inputmat, QuantizedTensorBase)
 ):
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 inputmat.update_usage(rowwise_usage=True, columnwise_usage=False)
 else:
 # Discard row-wise data since it is not needed in backward pass
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)

 # Cached input tensor
 saved_inputmat = None
 if backward_needs_input:
 if not save_original_input:
 if own_quantized_input and isinstance(inputmat, QuantizedTensorBase):
 # For sequence parallel in vanilla FP8, rowwise data is
 # to gather the input. For MXFP8, columnwise only data
 # can be allgathered.
 if (
 isinstance(inputmat, (MXFP8TensorBase, Float8BlockwiseQTensorBase))
 or not ctx.backward_input_needs_gather
 ):
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)
 saved_inputmat = inputmat

 # Weight with column-wise usage is needed for dgrad GEMM.
Expand DownExpand Up@@ -572,20 +579,26 @@ def backward(ctx, grad_output: torch.Tensor) -> Tuple[Union[torch.Tensor, None],
 inputmat_total = None
 inputmat_total_work = None
 if ctx.requires_wgrad:
 input_is_quantized = isinstance(inputmat, QuantizedTensorBase)
 if ctx.fp8 or ctx.debug:
 if not input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 # Input tensor is already quantized
 pass
 elif ctx.debug:
 # Debug quantizer will be applied immediately before wgrad GEMM
 pass
 else:
 # Quantize input tensor
 quantizer = ctx.input_quantizer
 if isinstance(quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)):
 quantizer.set_usage(
 rowwise=True,
 columnwise=not ctx.backward_input_needs_gather,
 )
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
quantizer.set_usage(rowwise=True, columnwise=False)
 else:
 quantizer.set_usage(rowwise=False, columnwise=True)
 quantizer.set_usage(rowwise=True, columnwise=True)
@hxbaihxbaiJul 20, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi Tim, why do we need the rowwise data here?
@timmoon10timmoon10Jul 21, 2025•

edited

Loading

Copy link


Copy Markdown

MemberAuthor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In principle we shouldn't need it, but I was running into issues where FP8 casts were failing without it. Actually, I don't think we need it once we skip the debug quantizer case.
@hxbaihxbaiJul 22, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this may be due to that the condition is modified. For (Float8Quantizer, Float8CurrentScalingQuantizer) cases, it doesn't support only quantizing the colwise data. But if backward_input_needs_gather is False, (Float8Quantizer, Float8CurrentScalingQuantizer) tensors would also go into the else path, which will cause the error. This hurts the performance of blockwise FP8 and MXFP8. Do you need me to create a fix PR for this, or will you fix this together with the DebugTensor?
 inputmat = quantizer(inputmat)
 else:
 if input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 inputmat = inputmat.dequantize(dtype=ctx.activation_dtype)
 else:
 inputmat = cast_if_needed(inputmat, ctx.activation_dtype)
Expand Down

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Highlight search terms from Google/DuckDuckGo/Bing referrer
(function() {
var ref = document.referrer;
var terms = [];
if (ref.includes('google.com') || ref.includes('duckduckgo.com') || ref.includes('bing.com')) {
var url = new URL(ref);
var q = url.searchParams.get('q') || url.searchParams.get('p');
if (q) {
terms = q.split(/\s+/).filter(function(t) { return t.length > 2; });
}
}
if (terms.length === 0) return;
var style = document.createElement('style');
style.textContent = '.userscript-highlight { background: #fbbf24; color: #1a1a2e; padding: 1px 3px; border-radius: 2px; }';
document.head.appendChild(style);
function highlight(node) {
if (node.nodeType === 3) { // text node
var text = node.textContent;
var found = false;
terms.forEach(function(term) {
var regex = new RegExp('(' + term.replace(/[.*+?^${}()|[\]\\]/g, '\\') + ')', 'gi');
if (regex.test(text)) {
found = true;
var frag = document.createDocumentFragment();
var parts = text.split(regex);
parts.forEach(function(part, i) {
if (i % 2 === 0) {
frag.appendChild(document.createTextNode(part));
} else {
var span = document.createElement('span');
span.className = 'userscript-highlight';
span.textContent = part;
frag.appendChild(span);
}
});
node.parentNode.replaceChild(frag, node);
}
});
} else if (node.nodeType === 1 && node.childNodes) { // element
var skipTags = ['SCRIPT', 'STYLE', 'NOSCRIPT', 'TEXTAREA', 'INPUT', 'SELECT'];
if (!skipTags.includes(node.tagName)) {
Array.from(node.childNodes).forEach(highlight);
}
}
}
highlight(document.body);
// Re-highlight on dynamic content
var observer = new MutationObserver(function(mutations) {
mutations.forEach(function(m) {
m.addedNodes.forEach(function(node) {
if (node.nodeType === 1 || node.nodeType === 3) highlight(node);
});
});
});
observer.observe(document.body, { childList: true, subtree: true });
})();
}
} catch(__e) { console.warn('[Userscript:Highlight Search Terms]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '
[PyTorch] Debug linear layer when saving original input and using debug quantizer by timmoon10 · Pull Request #1963 · NVIDIA/TransformerEngine · GitHub
Skip to content
[PyTorch] Debug linear layer when saving original input and using debug quantizer#1963

New issue

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

 Sign up for GitHub

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

Already on GitHub?
Sign in
to your account

 Merged

timmoon10
merged 8 commits into

from

Jul 22, 2025
 Merged


[PyTorch] Debug linear layer when saving original input and using debug quantizer
#1963
 Changes from all commits
Commits

Show all changes


8 commits


Select commit
Hold shift + click to select a range
867ab1c

Debug linear layer when saving original input and using debug quantizer


timmoon10 Jul 17, 2025
6f66aa4

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 17, 2025
7c20075

[pre-commit.ci] auto fixes from pre-commit.com hooks


pre-commit-ci[bot] Jul 17, 2025
212a446

Remove unused imports


timmoon10 Jul 18, 2025
d83014d

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
62d5c38

Avoid unnecessary row-wise data


timmoon10 Jul 21, 2025
6cc7cac

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
d9c7dd1

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 22, 2025

File filter

Filter by extension
Filter by extension
Conversations

Failed to load comments. 
Loading
Jump to

Jump to file


Failed to load files. 
Loading
 Diff view

 Diff view


65 changes: 39 additions & 26 deletions
transformer_engine/pytorch/module/linear.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -65,8 +65,6 @@
)
from ..tensor.float8_tensor import Float8CurrentScalingQuantizer, Float8Quantizer
from ..tensor.mxfp8_tensor import MXFP8Quantizer
from ..tensor._internal.mxfp8_tensor_base import MXFP8TensorBase
from ..tensor._internal.float8_blockwise_tensor_base import Float8BlockwiseQTensorBase
from ..export import is_in_onnx_export_mode, assert_warmed_up
from ..cpu_offload import is_cpu_offload_enabled, mark_activation_offload
from ...debug.pytorch.debug_state import TEDebugState
Expand DownExpand Up@@ -170,16 +168,19 @@ def forward(
 if input_quantizer is None:
 raise ValueError("Missing quantizer for input tensor")
 if not isinstance(inputmat, QuantizedTensorBase):
 input_quantizer.set_usage(
 rowwise=True, columnwise=backward_needs_input and not save_original_input
 )
 own_quantized_input = True
 input_quantizer.set_usage(rowwise=True, columnwise=backward_needs_input)
 if isinstance(
 input_quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 input_quantizer.set_usage(columnwise=False)
 if save_original_input:
 # No need for column-wise data since this
 # tensor will not be cached for backward pass
 input_quantizer.set_usage(columnwise=False)
 own_quantized_input = False
 inputmat = input_quantizer(inputmat)
 own_quantized_input = True
 else:
 inputmat = cast_if_needed(inp, activation_dtype) # Cast for AMP

Expand DownExpand Up@@ -344,23 +345,29 @@ def forward(
 inputmat = inp

 ctx.weight_quantizer = weight_quantizer
 saved_inputmat = None

 ctx.backward_input_needs_gather = (
 weight.requires_grad and parallel_mode == "column" and sequence_parallel
 )

 # Discard unneeded data in input tensor
 if (
 backward_needs_input
 and own_quantized_input
 and isinstance(inputmat, QuantizedTensorBase)
 ):
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 inputmat.update_usage(rowwise_usage=True, columnwise_usage=False)
 else:
 # Discard row-wise data since it is not needed in backward pass
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)

 # Cached input tensor
 saved_inputmat = None
 if backward_needs_input:
 if not save_original_input:
 if own_quantized_input and isinstance(inputmat, QuantizedTensorBase):
 # For sequence parallel in vanilla FP8, rowwise data is
 # to gather the input. For MXFP8, columnwise only data
 # can be allgathered.
 if (
 isinstance(inputmat, (MXFP8TensorBase, Float8BlockwiseQTensorBase))
 or not ctx.backward_input_needs_gather
 ):
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)
 saved_inputmat = inputmat

 # Weight with column-wise usage is needed for dgrad GEMM.
Expand DownExpand Up@@ -572,20 +579,26 @@ def backward(ctx, grad_output: torch.Tensor) -> Tuple[Union[torch.Tensor, None],
 inputmat_total = None
 inputmat_total_work = None
 if ctx.requires_wgrad:
 input_is_quantized = isinstance(inputmat, QuantizedTensorBase)
 if ctx.fp8 or ctx.debug:
 if not input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 # Input tensor is already quantized
 pass
 elif ctx.debug:
 # Debug quantizer will be applied immediately before wgrad GEMM
 pass
 else:
 # Quantize input tensor
 quantizer = ctx.input_quantizer
 if isinstance(quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)):
 quantizer.set_usage(
 rowwise=True,
 columnwise=not ctx.backward_input_needs_gather,
 )
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
quantizer.set_usage(rowwise=True, columnwise=False)
 else:
 quantizer.set_usage(rowwise=False, columnwise=True)
 quantizer.set_usage(rowwise=True, columnwise=True)
@hxbaihxbaiJul 20, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi Tim, why do we need the rowwise data here?
@timmoon10timmoon10Jul 21, 2025•

edited

Loading

Copy link


Copy Markdown

MemberAuthor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In principle we shouldn't need it, but I was running into issues where FP8 casts were failing without it. Actually, I don't think we need it once we skip the debug quantizer case.
@hxbaihxbaiJul 22, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this may be due to that the condition is modified. For (Float8Quantizer, Float8CurrentScalingQuantizer) cases, it doesn't support only quantizing the colwise data. But if backward_input_needs_gather is False, (Float8Quantizer, Float8CurrentScalingQuantizer) tensors would also go into the else path, which will cause the error. This hurts the performance of blockwise FP8 and MXFP8. Do you need me to create a fix PR for this, or will you fix this together with the DebugTensor?
 inputmat = quantizer(inputmat)
 else:
 if input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 inputmat = inputmat.dequantize(dtype=ctx.activation_dtype)
 else:
 inputmat = cast_if_needed(inputmat, ctx.activation_dtype)
Expand Down

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Strip utm_, fbclid, gclid, etc. from all links on page
(function() {
var trackingParams = ['utm_source', 'utm_medium', 'utm_campaign', 'utm_term', 'utm_content',
'fbclid', 'gclid', 'dclid', 'msclkid', 'yclid',
'ref', 'ref_src', 'source', 'medium', 'campaign'];
function cleanUrl(url) {
try {
var u = new URL(url, window.location.origin);
var changed = false;
trackingParams.forEach(function(p) {
if (u.searchParams.has(p)) {
u.searchParams.delete(p);
changed = true;
}
});
return changed ? u.toString() : url;
} catch (e) {
return url;
}
}
function cleanLinks() {
document.querySelectorAll('a[href]').forEach(function(a) {
var clean = cleanUrl(a.href);
if (clean !== a.href) a.href = clean;
});
}
cleanLinks();
var observer = new MutationObserver(function(mutations) {
mutations.forEach(function(m) {
m.addedNodes.forEach(function(node) {
if (node.nodeType === 1) {
if (node.tagName === 'A') cleanLinks();
node.querySelectorAll('a[href]').forEach(function(a) {
var clean = cleanUrl(a.href);
if (clean !== a.href) a.href = clean;
});
}
});
});
});
observer.observe(document.body, { childList: true, subtree: true });
})();
}
} catch(__e) { console.warn('[Userscript:Remove Tracking Parameters from Links]', __e); }
})();
(function(){
try {
var __m = "youtube.com";
var __re = new RegExp('^' + "youtube\\.com" + '
[PyTorch] Debug linear layer when saving original input and using debug quantizer by timmoon10 · Pull Request #1963 · NVIDIA/TransformerEngine · GitHub
Skip to content
[PyTorch] Debug linear layer when saving original input and using debug quantizer#1963

New issue

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

 Sign up for GitHub

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

Already on GitHub?
Sign in
to your account

 Merged

timmoon10
merged 8 commits into

from

Jul 22, 2025
 Merged


[PyTorch] Debug linear layer when saving original input and using debug quantizer
#1963
 Changes from all commits
Commits

Show all changes


8 commits


Select commit
Hold shift + click to select a range
867ab1c

Debug linear layer when saving original input and using debug quantizer


timmoon10 Jul 17, 2025
6f66aa4

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 17, 2025
7c20075

[pre-commit.ci] auto fixes from pre-commit.com hooks


pre-commit-ci[bot] Jul 17, 2025
212a446

Remove unused imports


timmoon10 Jul 18, 2025
d83014d

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
62d5c38

Avoid unnecessary row-wise data


timmoon10 Jul 21, 2025
6cc7cac

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
d9c7dd1

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 22, 2025

File filter

Filter by extension
Filter by extension
Conversations

Failed to load comments. 
Loading
Jump to

Jump to file


Failed to load files. 
Loading
 Diff view

 Diff view


65 changes: 39 additions & 26 deletions
transformer_engine/pytorch/module/linear.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -65,8 +65,6 @@
)
from ..tensor.float8_tensor import Float8CurrentScalingQuantizer, Float8Quantizer
from ..tensor.mxfp8_tensor import MXFP8Quantizer
from ..tensor._internal.mxfp8_tensor_base import MXFP8TensorBase
from ..tensor._internal.float8_blockwise_tensor_base import Float8BlockwiseQTensorBase
from ..export import is_in_onnx_export_mode, assert_warmed_up
from ..cpu_offload import is_cpu_offload_enabled, mark_activation_offload
from ...debug.pytorch.debug_state import TEDebugState
Expand DownExpand Up@@ -170,16 +168,19 @@ def forward(
 if input_quantizer is None:
 raise ValueError("Missing quantizer for input tensor")
 if not isinstance(inputmat, QuantizedTensorBase):
 input_quantizer.set_usage(
 rowwise=True, columnwise=backward_needs_input and not save_original_input
 )
 own_quantized_input = True
 input_quantizer.set_usage(rowwise=True, columnwise=backward_needs_input)
 if isinstance(
 input_quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 input_quantizer.set_usage(columnwise=False)
 if save_original_input:
 # No need for column-wise data since this
 # tensor will not be cached for backward pass
 input_quantizer.set_usage(columnwise=False)
 own_quantized_input = False
 inputmat = input_quantizer(inputmat)
 own_quantized_input = True
 else:
 inputmat = cast_if_needed(inp, activation_dtype) # Cast for AMP

Expand DownExpand Up@@ -344,23 +345,29 @@ def forward(
 inputmat = inp

 ctx.weight_quantizer = weight_quantizer
 saved_inputmat = None

 ctx.backward_input_needs_gather = (
 weight.requires_grad and parallel_mode == "column" and sequence_parallel
 )

 # Discard unneeded data in input tensor
 if (
 backward_needs_input
 and own_quantized_input
 and isinstance(inputmat, QuantizedTensorBase)
 ):
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 inputmat.update_usage(rowwise_usage=True, columnwise_usage=False)
 else:
 # Discard row-wise data since it is not needed in backward pass
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)

 # Cached input tensor
 saved_inputmat = None
 if backward_needs_input:
 if not save_original_input:
 if own_quantized_input and isinstance(inputmat, QuantizedTensorBase):
 # For sequence parallel in vanilla FP8, rowwise data is
 # to gather the input. For MXFP8, columnwise only data
 # can be allgathered.
 if (
 isinstance(inputmat, (MXFP8TensorBase, Float8BlockwiseQTensorBase))
 or not ctx.backward_input_needs_gather
 ):
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)
 saved_inputmat = inputmat

 # Weight with column-wise usage is needed for dgrad GEMM.
Expand DownExpand Up@@ -572,20 +579,26 @@ def backward(ctx, grad_output: torch.Tensor) -> Tuple[Union[torch.Tensor, None],
 inputmat_total = None
 inputmat_total_work = None
 if ctx.requires_wgrad:
 input_is_quantized = isinstance(inputmat, QuantizedTensorBase)
 if ctx.fp8 or ctx.debug:
 if not input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 # Input tensor is already quantized
 pass
 elif ctx.debug:
 # Debug quantizer will be applied immediately before wgrad GEMM
 pass
 else:
 # Quantize input tensor
 quantizer = ctx.input_quantizer
 if isinstance(quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)):
 quantizer.set_usage(
 rowwise=True,
 columnwise=not ctx.backward_input_needs_gather,
 )
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
quantizer.set_usage(rowwise=True, columnwise=False)
 else:
 quantizer.set_usage(rowwise=False, columnwise=True)
 quantizer.set_usage(rowwise=True, columnwise=True)
@hxbaihxbaiJul 20, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi Tim, why do we need the rowwise data here?
@timmoon10timmoon10Jul 21, 2025•

edited

Loading

Copy link


Copy Markdown

MemberAuthor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In principle we shouldn't need it, but I was running into issues where FP8 casts were failing without it. Actually, I don't think we need it once we skip the debug quantizer case.
@hxbaihxbaiJul 22, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this may be due to that the condition is modified. For (Float8Quantizer, Float8CurrentScalingQuantizer) cases, it doesn't support only quantizing the colwise data. But if backward_input_needs_gather is False, (Float8Quantizer, Float8CurrentScalingQuantizer) tensors would also go into the else path, which will cause the error. This hurts the performance of blockwise FP8 and MXFP8. Do you need me to create a fix PR for this, or will you fix this together with the DebugTensor?
 inputmat = quantizer(inputmat)
 else:
 if input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 inputmat = inputmat.dequantize(dtype=ctx.activation_dtype)
 else:
 inputmat = cast_if_needed(inputmat, ctx.activation_dtype)
Expand Down

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Auto-enable theater mode on YouTube
(function() {
function tryTheater() {
var btn = document.querySelector('button[aria-label="Theater mode"], ytd-player #player button[title="Theater mode"]');
if (btn && !btn.classList.contains('activated')) {
btn.click();
}
}
// Try immediately
tryTheater();
// Try after navigation (SPA)
var lastUrl = location.href;
setInterval(function() {
if (location.href !== lastUrl) {
lastUrl = location.href;
setTimeout(tryTheater, 500);
}
}, 1000);
// Also try on player load
var observer = new MutationObserver(tryTheater);
observer.observe(document.body, { childList: true, subtree: true });
})();
}
} catch(__e) { console.warn('[Userscript:YouTube Theater Mode Default]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '
[PyTorch] Debug linear layer when saving original input and using debug quantizer by timmoon10 · Pull Request #1963 · NVIDIA/TransformerEngine · GitHub
Skip to content
[PyTorch] Debug linear layer when saving original input and using debug quantizer#1963

New issue

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

 Sign up for GitHub

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

Already on GitHub?
Sign in
to your account

 Merged

timmoon10
merged 8 commits into

from

Jul 22, 2025
 Merged


[PyTorch] Debug linear layer when saving original input and using debug quantizer
#1963
 Changes from all commits
Commits

Show all changes


8 commits


Select commit
Hold shift + click to select a range
867ab1c

Debug linear layer when saving original input and using debug quantizer


timmoon10 Jul 17, 2025
6f66aa4

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 17, 2025
7c20075

[pre-commit.ci] auto fixes from pre-commit.com hooks


pre-commit-ci[bot] Jul 17, 2025
212a446

Remove unused imports


timmoon10 Jul 18, 2025
d83014d

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
62d5c38

Avoid unnecessary row-wise data


timmoon10 Jul 21, 2025
6cc7cac

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
d9c7dd1

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 22, 2025

File filter

Filter by extension
Filter by extension
Conversations

Failed to load comments. 
Loading
Jump to

Jump to file


Failed to load files. 
Loading
 Diff view

 Diff view


65 changes: 39 additions & 26 deletions
transformer_engine/pytorch/module/linear.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -65,8 +65,6 @@
)
from ..tensor.float8_tensor import Float8CurrentScalingQuantizer, Float8Quantizer
from ..tensor.mxfp8_tensor import MXFP8Quantizer
from ..tensor._internal.mxfp8_tensor_base import MXFP8TensorBase
from ..tensor._internal.float8_blockwise_tensor_base import Float8BlockwiseQTensorBase
from ..export import is_in_onnx_export_mode, assert_warmed_up
from ..cpu_offload import is_cpu_offload_enabled, mark_activation_offload
from ...debug.pytorch.debug_state import TEDebugState
Expand DownExpand Up@@ -170,16 +168,19 @@ def forward(
 if input_quantizer is None:
 raise ValueError("Missing quantizer for input tensor")
 if not isinstance(inputmat, QuantizedTensorBase):
 input_quantizer.set_usage(
 rowwise=True, columnwise=backward_needs_input and not save_original_input
 )
 own_quantized_input = True
 input_quantizer.set_usage(rowwise=True, columnwise=backward_needs_input)
 if isinstance(
 input_quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 input_quantizer.set_usage(columnwise=False)
 if save_original_input:
 # No need for column-wise data since this
 # tensor will not be cached for backward pass
 input_quantizer.set_usage(columnwise=False)
 own_quantized_input = False
 inputmat = input_quantizer(inputmat)
 own_quantized_input = True
 else:
 inputmat = cast_if_needed(inp, activation_dtype) # Cast for AMP

Expand DownExpand Up@@ -344,23 +345,29 @@ def forward(
 inputmat = inp

 ctx.weight_quantizer = weight_quantizer
 saved_inputmat = None

 ctx.backward_input_needs_gather = (
 weight.requires_grad and parallel_mode == "column" and sequence_parallel
 )

 # Discard unneeded data in input tensor
 if (
 backward_needs_input
 and own_quantized_input
 and isinstance(inputmat, QuantizedTensorBase)
 ):
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 inputmat.update_usage(rowwise_usage=True, columnwise_usage=False)
 else:
 # Discard row-wise data since it is not needed in backward pass
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)

 # Cached input tensor
 saved_inputmat = None
 if backward_needs_input:
 if not save_original_input:
 if own_quantized_input and isinstance(inputmat, QuantizedTensorBase):
 # For sequence parallel in vanilla FP8, rowwise data is
 # to gather the input. For MXFP8, columnwise only data
 # can be allgathered.
 if (
 isinstance(inputmat, (MXFP8TensorBase, Float8BlockwiseQTensorBase))
 or not ctx.backward_input_needs_gather
 ):
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)
 saved_inputmat = inputmat

 # Weight with column-wise usage is needed for dgrad GEMM.
Expand DownExpand Up@@ -572,20 +579,26 @@ def backward(ctx, grad_output: torch.Tensor) -> Tuple[Union[torch.Tensor, None],
 inputmat_total = None
 inputmat_total_work = None
 if ctx.requires_wgrad:
 input_is_quantized = isinstance(inputmat, QuantizedTensorBase)
 if ctx.fp8 or ctx.debug:
 if not input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 # Input tensor is already quantized
 pass
 elif ctx.debug:
 # Debug quantizer will be applied immediately before wgrad GEMM
 pass
 else:
 # Quantize input tensor
 quantizer = ctx.input_quantizer
 if isinstance(quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)):
 quantizer.set_usage(
 rowwise=True,
 columnwise=not ctx.backward_input_needs_gather,
 )
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
quantizer.set_usage(rowwise=True, columnwise=False)
 else:
 quantizer.set_usage(rowwise=False, columnwise=True)
 quantizer.set_usage(rowwise=True, columnwise=True)
@hxbaihxbaiJul 20, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi Tim, why do we need the rowwise data here?
@timmoon10timmoon10Jul 21, 2025•

edited

Loading

Copy link


Copy Markdown

MemberAuthor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In principle we shouldn't need it, but I was running into issues where FP8 casts were failing without it. Actually, I don't think we need it once we skip the debug quantizer case.
@hxbaihxbaiJul 22, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this may be due to that the condition is modified. For (Float8Quantizer, Float8CurrentScalingQuantizer) cases, it doesn't support only quantizing the colwise data. But if backward_input_needs_gather is False, (Float8Quantizer, Float8CurrentScalingQuantizer) tensors would also go into the else path, which will cause the error. This hurts the performance of blockwise FP8 and MXFP8. Do you need me to create a fix PR for this, or will you fix this together with the DebugTensor?
 inputmat = quantizer(inputmat)
 else:
 if input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 inputmat = inputmat.dequantize(dtype=ctx.activation_dtype)
 else:
 inputmat = cast_if_needed(inputmat, ctx.activation_dtype)
Expand Down

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Remove or un-stick sticky/fixed headers that block content
(function() {
function unstick() {
document.querySelectorAll('header, nav, [role="banner"], .header, .navbar, .sticky, .fixed-top, [style*="position: fixed"], [style*="position:sticky"]').forEach(function(el) {
if (el.style.position === 'fixed' || el.style.position === 'sticky' || getComputedStyle(el).position === 'fixed' || getComputedStyle(el).position === 'sticky') {
el.style.position = 'static';
el.style.top = 'auto';
el.style.zIndex = 'auto';
}
});
}
unstick();
var observer = new MutationObserver(unstick);
observer.observe(document.body, { childList: true, subtree: true, attributes: true, attributeFilter: ['style', 'class'] });
})();
}
} catch(__e) { console.warn('[Userscript:Kill Sticky Headers]', __e); }
})();
})();
[PyTorch] Debug linear layer when saving original input and using debug quantizer by timmoon10 · Pull Request #1963 · NVIDIA/TransformerEngine · GitHub
Skip to content
[PyTorch] Debug linear layer when saving original input and using debug quantizer#1963

New issue

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

 Sign up for GitHub

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

Already on GitHub?
Sign in
to your account

 Merged

timmoon10
merged 8 commits into

from

Jul 22, 2025
 Merged


[PyTorch] Debug linear layer when saving original input and using debug quantizer
#1963
 Changes from all commits
Commits

Show all changes


8 commits


Select commit
Hold shift + click to select a range
867ab1c

Debug linear layer when saving original input and using debug quantizer


timmoon10 Jul 17, 2025
6f66aa4

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 17, 2025
7c20075

[pre-commit.ci] auto fixes from pre-commit.com hooks


pre-commit-ci[bot] Jul 17, 2025
212a446

Remove unused imports


timmoon10 Jul 18, 2025
d83014d

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
62d5c38

Avoid unnecessary row-wise data


timmoon10 Jul 21, 2025
6cc7cac

Merge branch 'main' into debug-linear-original-input


timmoon10 Jul 21, 2025
d9c7dd1

Workaround bugs with quantizing with only column-wise usage


timmoon10 Jul 22, 2025

File filter

Filter by extension
Filter by extension
Conversations

Failed to load comments. 
Loading
Jump to

Jump to file


Failed to load files. 
Loading
 Diff view

 Diff view


65 changes: 39 additions & 26 deletions
transformer_engine/pytorch/module/linear.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -65,8 +65,6 @@
)
from ..tensor.float8_tensor import Float8CurrentScalingQuantizer, Float8Quantizer
from ..tensor.mxfp8_tensor import MXFP8Quantizer
from ..tensor._internal.mxfp8_tensor_base import MXFP8TensorBase
from ..tensor._internal.float8_blockwise_tensor_base import Float8BlockwiseQTensorBase
from ..export import is_in_onnx_export_mode, assert_warmed_up
from ..cpu_offload import is_cpu_offload_enabled, mark_activation_offload
from ...debug.pytorch.debug_state import TEDebugState
Expand DownExpand Up@@ -170,16 +168,19 @@ def forward(
 if input_quantizer is None:
 raise ValueError("Missing quantizer for input tensor")
 if not isinstance(inputmat, QuantizedTensorBase):
 input_quantizer.set_usage(
 rowwise=True, columnwise=backward_needs_input and not save_original_input
 )
 own_quantized_input = True
 input_quantizer.set_usage(rowwise=True, columnwise=backward_needs_input)
 if isinstance(
 input_quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 input_quantizer.set_usage(columnwise=False)
 if save_original_input:
 # No need for column-wise data since this
 # tensor will not be cached for backward pass
 input_quantizer.set_usage(columnwise=False)
 own_quantized_input = False
 inputmat = input_quantizer(inputmat)
 own_quantized_input = True
 else:
 inputmat = cast_if_needed(inp, activation_dtype) # Cast for AMP

Expand DownExpand Up@@ -344,23 +345,29 @@ def forward(
 inputmat = inp

 ctx.weight_quantizer = weight_quantizer
 saved_inputmat = None

 ctx.backward_input_needs_gather = (
 weight.requires_grad and parallel_mode == "column" and sequence_parallel
 )

 # Discard unneeded data in input tensor
 if (
 backward_needs_input
 and own_quantized_input
 and isinstance(inputmat, QuantizedTensorBase)
 ):
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
 ):
 # All-gather is not supported with FP8 column-wise data
 inputmat.update_usage(rowwise_usage=True, columnwise_usage=False)
 else:
 # Discard row-wise data since it is not needed in backward pass
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)

 # Cached input tensor
 saved_inputmat = None
 if backward_needs_input:
 if not save_original_input:
 if own_quantized_input and isinstance(inputmat, QuantizedTensorBase):
 # For sequence parallel in vanilla FP8, rowwise data is
 # to gather the input. For MXFP8, columnwise only data
 # can be allgathered.
 if (
 isinstance(inputmat, (MXFP8TensorBase, Float8BlockwiseQTensorBase))
 or not ctx.backward_input_needs_gather
 ):
 inputmat.update_usage(rowwise_usage=False, columnwise_usage=True)
 saved_inputmat = inputmat

 # Weight with column-wise usage is needed for dgrad GEMM.
Expand DownExpand Up@@ -572,20 +579,26 @@ def backward(ctx, grad_output: torch.Tensor) -> Tuple[Union[torch.Tensor, None],
 inputmat_total = None
 inputmat_total_work = None
 if ctx.requires_wgrad:
 input_is_quantized = isinstance(inputmat, QuantizedTensorBase)
 if ctx.fp8 or ctx.debug:
 if not input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 # Input tensor is already quantized
 pass
 elif ctx.debug:
 # Debug quantizer will be applied immediately before wgrad GEMM
 pass
 else:
 # Quantize input tensor
 quantizer = ctx.input_quantizer
 if isinstance(quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)):
 quantizer.set_usage(
 rowwise=True,
 columnwise=not ctx.backward_input_needs_gather,
 )
 if ctx.backward_input_needs_gather and isinstance(
 quantizer, (Float8Quantizer, Float8CurrentScalingQuantizer)
):
# All-gather is not supported with FP8 column-wise data
quantizer.set_usage(rowwise=True, columnwise=False)
 else:
 quantizer.set_usage(rowwise=False, columnwise=True)
 quantizer.set_usage(rowwise=True, columnwise=True)
@hxbaihxbaiJul 20, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi Tim, why do we need the rowwise data here?
@timmoon10timmoon10Jul 21, 2025•

edited

Loading

Copy link


Copy Markdown

MemberAuthor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In principle we shouldn't need it, but I was running into issues where FP8 casts were failing without it. Actually, I don't think we need it once we skip the debug quantizer case.
@hxbaihxbaiJul 22, 2025

Copy link


Copy Markdown

Contributor
Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this may be due to that the condition is modified. For (Float8Quantizer, Float8CurrentScalingQuantizer) cases, it doesn't support only quantizing the colwise data. But if backward_input_needs_gather is False, (Float8Quantizer, Float8CurrentScalingQuantizer) tensors would also go into the else path, which will cause the error. This hurts the performance of blockwise FP8 and MXFP8. Do you need me to create a fix PR for this, or will you fix this together with the DebugTensor?
 inputmat = quantizer(inputmat)
 else:
 if input_is_quantized:
 if isinstance(inputmat, QuantizedTensorBase):
 inputmat = inputmat.dequantize(dtype=ctx.activation_dtype)
 else:
 inputmat = cast_if_needed(inputmat, ctx.activation_dtype)
Expand Down