Fix halting N300 tests. #170
Open
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
LPanosTT
commented
Jan 7, 2025
LPanosTT
force-pushed
the
lpanos/n300_op_by_op_fix
branch
2 times, most recently
from
ec6f06c to
dfdca7a
Compare
| @@ -14,8 +14,8 @@ jobs: | |||
| fail-fast: false | |||
| matrix: | |||
| build: [ | |||
| {runs-on: n150, name: "run1", test_names: "stable_diffusion, Qwen, MobileNetV2, clip, flan_t5, mlpmixer, resnet, vilt, albert, codegen, glpn_kitti, mnist, resnet50, RMBG, unet_carvana, mgp-str-base, musicgen_small, segformer, torchvision, yolos"}, | |||
| {runs-on: n150, name: "run2", test_names: "t5, whisper, autoencoder_conv, deit, gpt2, mobilenet_ssd, roberta, timm, xglm, autoencoder_linear, detr, beit, distilbert, hand_landmark, openpose, segment_anything, unet, yolov3, bert, dpr, hardnet, opt, speecht5_tts, unet_brain, yolov5, bloom, falcon, llama, perceiver_io, squeeze_bert, gpt_neo"}, | |||
| {runs-on: n150, n300, name: "run1", test_names: "stable_diffusion, Qwen, MobileNetV2, clip, flan_t5, mlpmixer, resnet, vilt, albert, codegen, glpn_kitti, mnist, resnet50, RMBG, unet_carvana, mgp-str-base, musicgen_small, segformer, torchvision, yolos"}, | |||
There was a problem hiding this comment.
Should we skip the nightlies until we figure out what's causing the hangs?
LPanosTT
force-pushed
the
lpanos/n300_op_by_op_fix
branch
2 times, most recently
from
63c7d79 to
1f2459c
Compare
LPanosTT
force-pushed
the
lpanos/n300_op_by_op_fix
branch
from
1a3acde to
2a34e08
Compare
add _extract_outputs function to ModelTester that subclasses may need to implement. This function should return a tuple of torch.Tensors Trim unused code in tests/utils.py move verify_against_golden to verify.py replace self.model with self.framework_model and self.compiled_model Use fp32 for all models Enable n300 model tests. Use to create new processes with python multiprocessing rather than the default make sure to terminate process if an error is raised before we get to it n300 run
LPanosTT
force-pushed
the
lpanos/n300_op_by_op_fix
branch
from
2a34e08 to
004151d
Compare
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.
Open the subprocesses which compile and run ops in the op-by-op flow by spawning rather than forking the main process.
If an error is thrown during the compilation, catch it, terminate the process, then raise it again.