mirror of
https://github.com/open-mmlab/mmclassification.git
synced 2025-06-03 21:53:55 +08:00
b4eefe4794
* implement training and evaluation on IPU * fp16 SOTA * Tput reaches 5600 * 123 * add poptorch dataloder * change ipu_replicas to ipu-replicas * add noqa to config long line(website) * remove ipu dataloder test code * del one blank line in test_builder * refine the dataloder initialization * fix a typo * refine args for dataloder * remove an annoted line * process one more conflict * adjust code structure in mmcv.ipu * adjust ipu code structure in mmcv * IPUDataloader to IPUDataLoader * align with mmcv * adjust according to mmcv * mmcv code structre fixed Co-authored-by: hudi <dihu@graphcore.ai>
10 lines
607 B
Bash
10 lines
607 B
Bash
|
|
|
|
# get SOTA accuracy 81.2 for 224 input ViT fine-tuning, reference is below:
|
|
# https://github.com/google-research/vision_transformer#available-vit-models
|
|
# cfg: vit-base-p16_ft-4xb544_in1k-224_ipu train model in fp16 precision
|
|
# 8 epoch, 2176 batch size, 16 IPUs, 4 replicas, model Tput = 5600 images, training time 0.6 hour roughly
|
|
cfg_name=vit-base-p16_ft-4xb544_in1k-224_ipu
|
|
python3 tools/train.py configs/vision_transformer/${cfg_name}.py --ipu-replicas 4 --no-validate &&
|
|
python3 tools/test.py configs/vision_transformer/${cfg_name}.py work_dirs/${cfg_name}/latest.pth --metrics accuracy --device ipu
|