AMDMIGraphX
AMDMIGraphX copied to clipboard
Update RNN-T Imports
Imports causing RNN-T Errors
Codecov Report
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 92.02%. Comparing base (
46a80f2) to head (ab85e0a). Report is 161 commits behind head on develop.
Additional details and impacted files
@@ Coverage Diff @@
## develop #3474 +/- ##
===========================================
- Coverage 92.04% 92.02% -0.03%
===========================================
Files 506 509 +3
Lines 20872 21005 +133
===========================================
+ Hits 19212 19330 +118
- Misses 1660 1675 +15
| Flag | Coverage Δ | |
|---|---|---|
? |
Flags with carried forward coverage won't be shown. Click here to find out more.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
| Test | Batch | Rate new ab85e0 |
Rate old e2e921 |
Diff | Compare |
|---|---|---|---|---|---|
| torchvision-resnet50 | 64 | 3,261.65 | 3,255.97 | 0.17% | :white_check_mark: |
| torchvision-resnet50_fp16 | 64 | 6,986.08 | 6,971.30 | 0.21% | :white_check_mark: |
| torchvision-densenet121 | 32 | 2,436.46 | 2,431.20 | 0.22% | :white_check_mark: |
| torchvision-densenet121_fp16 | 32 | 4,101.67 | 4,087.45 | 0.35% | :white_check_mark: |
| torchvision-inceptionv3 | 32 | 1,637.46 | 1,635.71 | 0.11% | :white_check_mark: |
| torchvision-inceptionv3_fp16 | 32 | 2,745.39 | 2,755.82 | -0.38% | :white_check_mark: |
| cadene-inceptionv4 | 16 | 779.00 | 780.33 | -0.17% | :white_check_mark: |
| cadene-resnext64x4 | 16 | 808.75 | 806.94 | 0.22% | :white_check_mark: |
| slim-mobilenet | 64 | 7,463.95 | 7,464.88 | -0.01% | :white_check_mark: |
| slim-nasnetalarge | 64 | 208.46 | 208.51 | -0.02% | :white_check_mark: |
| slim-resnet50v2 | 64 | 3,442.99 | 3,435.02 | 0.23% | :white_check_mark: |
| bert-mrpc-onnx | 8 | 1,152.37 | 1,146.59 | 0.50% | :white_check_mark: |
| bert-mrpc-tf | 1 | 318.39 | 318.44 | -0.01% | :white_check_mark: |
| pytorch-examples-wlang-gru | 1 | 411.02 | 421.02 | -2.38% | :white_check_mark: |
| pytorch-examples-wlang-lstm | 1 | 392.48 | 374.33 | 4.85% | :high_brightness: |
| torchvision-resnet50_1 | 1 | 811.68 | 803.63 | 1.00% | :white_check_mark: |
| cadene-dpn92_1 | 1 | 400.36 | 395.85 | 1.14% | :white_check_mark: |
| cadene-resnext101_1 | 1 | 381.32 | 379.14 | 0.58% | :white_check_mark: |
| onnx-taau-downsample | 1 | 344.94 | 366.52 | -5.89% | :red_circle: |
| dlrm-criteoterabyte | 1 | 35.06 | 35.01 | 0.13% | :white_check_mark: |
| dlrm-criteoterabyte_fp16 | 1 | 58.18 | 58.19 | -0.02% | :white_check_mark: |
| agentmodel | 1 | 8,191.15 | 8,063.61 | 1.58% | :white_check_mark: |
| unet_fp16 | 2 | 57.95 | 58.95 | -1.69% | :white_check_mark: |
| resnet50v1_fp16 | 1 | 1,000.90 | 950.25 | 5.33% | :high_brightness: |
| resnet50v1_int8 | 1 | 977.83 | 971.97 | 0.60% | :white_check_mark: |
| bert_base_cased_fp16 | 64 | 1,174.54 | 1,170.73 | 0.33% | :white_check_mark: |
| bert_large_uncased_fp16 | 32 | 363.18 | 363.46 | -0.08% | :white_check_mark: |
| bert_large_fp16 | 1 | 211.47 | 213.45 | -0.93% | :white_check_mark: |
| distilgpt2_fp16 | 16 | 2,208.97 | 2,199.49 | 0.43% | :white_check_mark: |
| yolov5s | 1 | 532.70 | 539.25 | -1.22% | :white_check_mark: |
| tinyllama | 1 | 43.49 | 43.47 | 0.05% | :white_check_mark: |
| vicuna-fastchat | 1 | 168.88 | 174.86 | -3.42% | :red_circle: |
| whisper-tiny-encoder | 1 | 417.82 | 417.66 | 0.04% | :white_check_mark: |
| whisper-tiny-decoder | 1 | 419.40 | 427.92 | -1.99% | :white_check_mark: |
This build is not recommended to merge :red_circle:
:red_circle:bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output