Skip to content

Conversation

@vkuzo
Copy link
Contributor

@vkuzo vkuzo commented Nov 26, 2025

Summary:

Splits NVFP4InferenceConfig to

  1. NVFP4DynamicActivationNVFP4Weight for dynamic quant
  2. NVFP4WeightOnlyConfig for weight-only quant

to match naming of other inference workflow configs in torchao.

Test Plan:

pytest test/prototype/mx_formats -s -x

Reviewers:

Subscribers:

Tasks:

Tags:

vkuzo added 13 commits November 24, 2025 09:34
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
@vkuzo
Copy link
Contributor Author

vkuzo commented Nov 26, 2025

Stack from ghstack (oldest at bottom):

@pytorch-bot
Copy link

pytorch-bot bot commented Nov 26, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3398

Note: Links to docs will display an error until the docs builds have been completed.

❗ 2 Active SEVs

There are 2 currently active SEVs. If your PR is affected, please view them below:

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@meta-cla meta-cla bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Nov 26, 2025
vkuzo added a commit that referenced this pull request Nov 26, 2025
Summary:

Splits `NVFP4InferenceConfig` to
1. `NVFP4DynamicActivationNVFP4Weight` for dynamic quant
2. `NVFP4WeightOnlyConfig` for weight-only quant

Test Plan:

```
pytest test/prototype/mx_formats -s -x
```

Reviewers:

Subscribers:

Tasks:

Tags:
ghstack-source-id: 363cc94
ghstack-comment-id: 3583258138
Pull-Request: #3398
vkuzo added 4 commits December 1, 2025 05:52
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
vkuzo added a commit that referenced this pull request Dec 1, 2025
Summary:

Splits `NVFP4InferenceConfig` to
1. `NVFP4DynamicActivationNVFP4Weight` for dynamic quant
2. `NVFP4WeightOnlyConfig` for weight-only quant

Test Plan:

```
pytest test/prototype/mx_formats -s -x
```

Reviewers:

Subscribers:

Tasks:

Tags:
ghstack-source-id: f0894bc
ghstack-comment-id: 3583258138
Pull-Request: #3398
vkuzo added 6 commits December 1, 2025 06:04
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
vkuzo added a commit that referenced this pull request Dec 1, 2025
Summary:

Splits `NVFP4InferenceConfig` to
1. `NVFP4DynamicActivationNVFP4Weight` for dynamic quant
2. `NVFP4WeightOnlyConfig` for weight-only quant

Test Plan:

```
pytest test/prototype/mx_formats -s -x
```

Reviewers:

Subscribers:

Tasks:

Tags:
ghstack-source-id: f0894bc
ghstack-comment-id: 3583258138
Pull-Request: #3398
vkuzo added 5 commits December 1, 2025 06:05
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
vkuzo added a commit that referenced this pull request Dec 1, 2025
Summary:

Splits `NVFP4InferenceConfig` to
1. `NVFP4DynamicActivationNVFP4Weight` for dynamic quant
2. `NVFP4WeightOnlyConfig` for weight-only quant

Test Plan:

```
pytest test/prototype/mx_formats -s -x
```

Reviewers:

Subscribers:

Tasks:

Tags:
ghstack-source-id: f0894bc
ghstack-comment-id: 3583258138
Pull-Request: #3398
vkuzo added 4 commits December 1, 2025 06:06
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
vkuzo added a commit that referenced this pull request Dec 1, 2025
Summary:

Splits `NVFP4InferenceConfig` to
1. `NVFP4DynamicActivationNVFP4Weight` for dynamic quant
2. `NVFP4WeightOnlyConfig` for weight-only quant

Test Plan:

```
pytest test/prototype/mx_formats -s -x
```

Reviewers:

Subscribers:

Tasks:

Tags:
ghstack-source-id: f0894bc
ghstack-comment-id: 3583258138
Pull-Request: #3398
vkuzo added 3 commits December 1, 2025 06:07
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
vkuzo added a commit that referenced this pull request Dec 1, 2025
Summary:

Splits `NVFP4InferenceConfig` to
1. `NVFP4DynamicActivationNVFP4Weight` for dynamic quant
2. `NVFP4WeightOnlyConfig` for weight-only quant

Test Plan:

```
pytest test/prototype/mx_formats -s -x
```

Reviewers:

Subscribers:

Tasks:

Tags:
ghstack-source-id: f0894bc
ghstack-comment-id: 3583258138
Pull-Request: #3398
vkuzo added 2 commits December 1, 2025 06:08
[ghstack-poisoned]
[ghstack-poisoned]
vkuzo added a commit that referenced this pull request Dec 1, 2025
Summary:

Splits `NVFP4InferenceConfig` to
1. `NVFP4DynamicActivationNVFP4Weight` for dynamic quant
2. `NVFP4WeightOnlyConfig` for weight-only quant

Test Plan:

```
pytest test/prototype/mx_formats -s -x
```

Reviewers:

Subscribers:

Tasks:

Tags:
ghstack-source-id: f0894bc
ghstack-comment-id: 3583258138
Pull-Request: #3398
[ghstack-poisoned]
vkuzo added a commit that referenced this pull request Dec 1, 2025
Summary:

Splits `NVFP4InferenceConfig` to
1. `NVFP4DynamicActivationNVFP4Weight` for dynamic quant
2. `NVFP4WeightOnlyConfig` for weight-only quant

Test Plan:

```
pytest test/prototype/mx_formats -s -x
```

Reviewers:

Subscribers:

Tasks:

Tags:
ghstack-source-id: f0894bc
ghstack-comment-id: 3583258138
Pull-Request: #3398
@vkuzo vkuzo changed the base branch from gh/vkuzo/172/head to main December 1, 2025 14:08
@vkuzo vkuzo merged commit 8d4a921 into main Dec 1, 2025
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants