Update common used toy linear model #3275
Conversation
Co-authored-by: Jerry Zhang <[email protected]>
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3275
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 92ba186 with merge base 315e9b4 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
|
thanks for splitting the PR, this should be much easier to land |
| if model_type == "linear": | ||
| model = ToyLinearModel(k, n, high_precision_dtype).to(device) | ||
| model = ToySingleLinearModel(k, n, device=device, dtype=high_precision_dtype) | ||
| input_data = torch.randn(m, k, device=device, dtype=high_precision_dtype) |
There was a problem hiding this comment.
can this be changed to use example_inputs from model now
e.g. input_data = model.example_inputs(batch_size=m)[0]
|
@namgyu-youn please add label for the PR as well, as shown in https://github.com/pytorch/ao/actions/runs/19093634780/job/54642925650?pr=3275 |
|
this one should be |
|
@pytorchbot label "topic: not user facing" |
pytorch-bot
bot
added
the
topic: not user facing
Thanks for the info, I will keep in mind for future PRs. |
|
@pytorchbot label "topic: for developers" |
pytorch-bot
bot
added
the
topic: for developers
2 participants
Summary:
This PR is smaller chunks of #2729. Instead of implementing a linear model in each test code, we want to support its API for developers' utility and consistency. The updated toy model requires device and dtype to help clearly identify.
Test Plan:
test/sparsity/test_fast_sparse_training.py
Future Plan:
This PR only updates the toy model in test/sparsity/test_fast_sparse_training.py. For a full update, we also have to update the following tests, same as
test_fast_sparse_training.py:cc @jerryzh168