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.

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

bf16 verify tests #3707

Open
wants to merge 11 commits into
base: develop
Choose a base branch
from
Open

bf16 verify tests #3707

wants to merge 11 commits into from

Conversation

richagadgil
Copy link
Contributor

No description provided.

@richagadgil richagadgil requested a review from causten as a code owner December 12, 2024 01:26
@richagadgil richagadgil self-assigned this Dec 13, 2024
@migraphx-bot
Copy link
Collaborator

Test Batch Rate new
17dacd
Rate old
810b3f
Diff Compare
torchvision-resnet50 64 3,225.22 3,227.48 -0.07%
torchvision-resnet50_fp16 64 6,939.36 6,935.80 0.05%
torchvision-densenet121 32 2,445.51 2,443.57 0.08%
torchvision-densenet121_fp16 32 4,174.30 4,190.37 -0.38%
torchvision-inceptionv3 32 1,617.88 1,618.09 -0.01%
torchvision-inceptionv3_fp16 32 2,701.36 2,703.46 -0.08%
cadene-inceptionv4 16 754.06 754.54 -0.06%
cadene-resnext64x4 16 799.64 799.94 -0.04%
slim-mobilenet 64 7,037.14 7,038.75 -0.02%
slim-nasnetalarge 64 203.72 203.74 -0.01%
slim-resnet50v2 64 3,417.20 3,418.15 -0.03%
bert-mrpc-onnx 8 1,131.98 1,130.84 0.10%
bert-mrpc-tf 1 479.04 483.04 -0.83%
pytorch-examples-wlang-gru 1 482.07 471.20 2.31%
pytorch-examples-wlang-lstm 1 439.16 446.88 -1.73%
torchvision-resnet50_1 1 800.85 802.10 -0.16%
cadene-dpn92_1 1 425.08 426.83 -0.41%
cadene-resnext101_1 1 384.57 383.89 0.17%
onnx-taau-downsample 1 372.62 372.62 0.00%
dlrm-criteoterabyte 1 32.54 32.54 0.02%
dlrm-criteoterabyte_fp16 1 51.03 51.07 -0.07%
agentmodel 1 8,611.71 8,676.04 -0.74%
unet_fp16 2 57.94 57.90 0.07%
resnet50v1_fp16 1 1,012.13 1,041.98 -2.86%
resnet50v1_int8 1 795.94 779.61 2.09%
bert_base_cased_fp16 64 1,172.37 1,172.53 -0.01%
bert_large_uncased_fp16 32 362.48 362.45 0.01%
bert_large_fp16 1 202.12 200.21 0.95%
distilgpt2_fp16 16 2,217.06 2,216.14 0.04%
yolov5s 1 514.48 518.19 -0.72%
tinyllama 1 43.61 43.59 0.03%
vicuna-fastchat 1 175.12 174.63 0.28%
whisper-tiny-encoder 1 404.82 406.05 -0.30%
whisper-tiny-decoder 1 402.63 398.24 1.10%

This build is OK for merge ✅

@migraphx-bot
Copy link
Collaborator


     ✅ bert-mrpc-onnx: PASSED: MIGraphX meets tolerance

     ✅ bert-mrpc-tf: PASSED: MIGraphX meets tolerance

     ✅ pytorch-examples-wlang-gru: PASSED: MIGraphX meets tolerance

     ✅ pytorch-examples-wlang-lstm: PASSED: MIGraphX meets tolerance

     ✅ torchvision-resnet50_1: PASSED: MIGraphX meets tolerance

     ✅ cadene-dpn92_1: PASSED: MIGraphX meets tolerance

     ✅ cadene-resnext101_1: PASSED: MIGraphX meets tolerance

     ✅ dlrm-criteoterabyte: PASSED: MIGraphX meets tolerance

     ✅ agentmodel: PASSED: MIGraphX meets tolerance

     ✅ unet: PASSED: MIGraphX meets tolerance

     ✅ resnet50v1: PASSED: MIGraphX meets tolerance

     ✅ bert_base_cased_fp16: PASSED: MIGraphX meets tolerance

🔴bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output


     ✅ bert_large: PASSED: MIGraphX meets tolerance

     ✅ yolov5s: PASSED: MIGraphX meets tolerance

     ✅ tinyllama: PASSED: MIGraphX meets tolerance

     ✅ vicuna-fastchat: PASSED: MIGraphX meets tolerance

     ✅ whisper-tiny-encoder: PASSED: MIGraphX meets tolerance

     ✅ whisper-tiny-decoder: PASSED: MIGraphX meets tolerance

     ✅ distilgpt2_fp16: PASSED: MIGraphX meets tolerance

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants