[Tests] Improve unet / vae tests #1018
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
The documentation is not available anymore as the PR was closed or merged. |
|
Slow tests currently don't work on CPU because of: huggingface/accelerate#796 Think that's fine as they are only run on GPU anyways - device_map="auto" speeds up init from ~25seconds to ~3seconds so this does make a big difference for overall testing - cc @anton-l |
anton-l
approved these changes
Oct 28, 2022
1
Outdated
| @@ -19,9 +19,10 @@ | |||
|
|
|||
There was a problem hiding this comment.
Looks like accidental renaming here, tests/test_models_vae.py → 1
| @@ -397,7 +355,7 @@ def test_from_pretrained_hub(self): | |||
|
|
|||
| @slow | |||
| def test_output_pretrained_ve_mid(self): | |||
| model = UNet2DModel.from_pretrained("google/ncsnpp-celebahq-256") | |||
| model = UNet2DModel.from_pretrained("google/ncsnpp-celebahq-256", device_map="auto") | |||
There was a problem hiding this comment.
Nice to see "auto" helping the latency!
tests/models/test_models_unet_2d.py
Outdated
| ) | ||
| @require_torch_gpu | ||
| def test_compvis_sd_v1_4_fp16(self, seed, timestep, expected_slice): | ||
| pass |
There was a problem hiding this comment.
Is it meant to be skipped for now? 4 more passes in the tests below too
PhaneeshB
pushed a commit
to nod-ai/diffusers
that referenced
this pull request
Mar 1, 2023
also fix logo paths
yoonseokjin
pushed a commit
to yoonseokjin/diffusers
that referenced
this pull request
Dec 25, 2023
* improve tests * up * finish * upload * add init * up * finish vae * finish * reduce loading time with device_map * remove device_map from CPU * uP
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add aggressive unet, vae single-forward tests with 1e-4 precision.
Using parameterized tests over seeds and timesteps.
@anton-l I'd say we see whether precision holds over the next days (weekend) and then we eventually adapt next week.
@patil-suraj @pcuenca @anton-l - note that tests are now automatically suffixed with
_{0, 1, 2, 4}when usingparameterized=> this makes it slightly more unreadable/useable when running locally, e.g.: