Quantization calibration refactor #6893
Code refactor
60832443
Modify code to tackle OOM when calibrating on larget dataset
7d70d7e9
Fix mismatch issue when setting keepdims on ReduceMin/ReduceMax
8758f09f
Add COCO val 2017 annotation
8d0e7a18
Fix mismatch issue when setting keepdims on ReduceMin/ReduceMax
a5e0c314
Fix bug of "No module named:onnxruntime.quantization.CalTableFlatBuff…
721f8303
Check and install flatbuffers module
6c521e28
Add script to donwload coco dataset image and refactor example
f3dfee78
Fix bug of "No module
4cd2c7ad
Add CalTableFaltBuffers as module
ba45e6d4
Remove annotation, user can download by themselves.
4e081ae4
Uncommet code
08c11fe3
Add back instances_val2017.json
ac098744
Make sure flatbuffers installed when ORT is installed
d9c5dabb
Refactor code to call coco api
690a2da8
Enable FP16 for example
e71f146c
yufenglee
approved these changes
on 2021-03-19
chilo-ms
merged
8c3b59a0
into master 5 years ago
chilo-ms
deleted the quantization_calibration_refactor branch 5 years ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub