Skip to content
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

Jump to bottom

Enable Global Weight Decay for VBE #2507

Closed
wants to merge 1 commit into from
Closed

Enable Global Weight Decay for VBE #2507

wants to merge 1 commit into from

Conversation

spcyppt
Copy link
Contributor

@spcyppt spcyppt commented Apr 17, 2024

Summary:
Enable Global weight decay for VBE

Usage:
set

optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]

e.g.,

tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)

Relevant diffs:
D53866750
D55660277
D55660762

Differential Revision: D56200676

@netlify Netlify
Copy link

netlify bot commented Apr 17, 2024 •
edited
Loading

✅ Deploy Preview for pytorch-fbgemm-docs ready!

Name Link
🔨 Latest commit 7ec38f6
🔍 Latest deploy log https://app.netlify.com/sites/pytorch-fbgemm-docs/deploys/66845e471c5d0c0008b93f83
😎 Deploy Preview https://deploy-preview-2507--pytorch-fbgemm-docs.netlify.app
📱 Preview on mobile
Toggle QR Code...

QR Code

Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify site configuration.

@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

spcyppt added a commit to spcyppt/FBGEMM that referenced this pull request Apr 17, 2024
@spcyppt @facebook-github-bot
Enable Global Weight Decay for VBE (pytorch#2507)
72f21ee 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Differential Revision: D56200676
spcyppt added a commit to spcyppt/FBGEMM that referenced this pull request May 2, 2024
@spcyppt @facebook-github-bot
Enable Global Weight Decay for VBE (pytorch#2507)
d7274df 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Differential Revision: D56200676
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

1 similar comment
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

spcyppt added a commit to spcyppt/FBGEMM that referenced this pull request Jul 2, 2024
@spcyppt @facebook-github-bot
Enable Global Weight Decay for TBE VBE (pytorch#2507)
3e4d361 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Reviewed By: sryap

Differential Revision: D56200676
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

spcyppt added a commit to spcyppt/FBGEMM that referenced this pull request Jul 2, 2024
@spcyppt @facebook-github-bot
Enable Global Weight Decay for TBE VBE (pytorch#2507)
2ae3a71 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Reviewed By: sryap

Differential Revision: D56200676
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

spcyppt added a commit to spcyppt/FBGEMM that referenced this pull request Jul 2, 2024
@spcyppt @facebook-github-bot
Enable Global Weight Decay for TBE VBE (pytorch#2507)
e5f2b58 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Reviewed By: sryap

Differential Revision: D56200676
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

spcyppt added a commit to spcyppt/FBGEMM that referenced this pull request Jul 2, 2024
@spcyppt @facebook-github-bot
Enable Global Weight Decay for TBE VBE (pytorch#2507)
dee4542 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Reviewed By: sryap

Differential Revision: D56200676
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

spcyppt added a commit to spcyppt/FBGEMM that referenced this pull request Jul 2, 2024
@spcyppt @facebook-github-bot
Enable Global Weight Decay for TBE VBE (pytorch#2507)
6fc9692 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Reviewed By: sryap

Differential Revision: D56200676
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

spcyppt added a commit to spcyppt/FBGEMM that referenced this pull request Jul 2, 2024
@spcyppt @facebook-github-bot
Enable Global Weight Decay for TBE VBE (pytorch#2507)
35a7c04 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Reviewed By: sryap

Differential Revision: D56200676
@spcyppt @facebook-github-bot
Enable Global Weight Decay for TBE VBE (pytorch#2507)
7ec38f6 
Summary:
Pull Request resolved: pytorch#2507

Enable Global weight decay for VBE
 ---
**Usage:**
set
```
optimizer = OptimType.EXACT_ROWWISE_ADAGRAD
weight_decay_mode = WeightDecayMode.DECOUPLE_GLOBAL

#  for VBE, pass batch_size_per_feature_per_rank
# Example: num_features = 2, num_ranks = 4
batch_size_per_feature_per_rank = [
    [1,  2, 8, 3] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 0
    [6, 10, 3, 5] # batch sizes for [Rank 0, Rank 1, Rank 2, Rank 3] in Feature 1
]
```

e.g.,
```
tbe = SplitTableBatchedEmbeddingBagsCodegen(
            embedding_specs=[
                (E, D, managed_option, ComputeDevice.CUDA) for (E, D) in zip(Es, Ds)
            ],
            optimizer=OptimType.EXACT_ROWWISE_ADAGRAD,
            learning_rate=0.1,
            eps=0.1,
            output_dtype=output_dtype,
            pooling_mode=pooling_mode,
            weight_decay_mode=WeightDecayMode.DECOUPLE_GLOBAL,
        )
output = tbe(indices, offsets, batch_size_per_feature_per_rank=batch_size_per_feature_per_rank)
```
Relevant diffs:
D53866750
D55660277
D55660762

Reviewed By: sryap

Differential Revision: D56200676
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D56200676

@facebook-github-bot
Copy link
Contributor

This pull request has been merged in 114bb0d.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
cla signed fb-exported Merged
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
2 participants
@spcyppt @facebook-github-bot