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.

Sign up for GitHub

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

[Debug Ascend CI] only for test CI #907

Draft
wants to merge 13 commits into
base: main
Choose a base branch
from
Draft

[Debug Ascend CI] only for test CI #907

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
13 commits
Select commit Hold shift + click to select a range
3a679ff
change tests model list to debug ascend ci
DoorKickers Jul 19, 2024
cab07d8
fix conv2d memory format bug
DoorKickers Jul 22, 2024
fa445a8
fix ascend memory format config on conv2d
DoorKickers Jul 22, 2024
3c045e2
another fix ascend memory format config on conv2d
DoorKickers Jul 22, 2024
6fb9fd2
revert to a older DIOPI commit
DoorKickers Jul 22, 2024
05c2a47
update some one-iter-test model
DoorKickers Jul 23, 2024
7f84a56
revert model to check CI bug
DoorKickers Jul 23, 2024
0b06497
add mobile v2
DoorKickers Jul 24, 2024
c9ad988
update branch to latest
DoorKickers Jul 25, 2024
8460a38
Merge remote-tracking branch 'refs/remotes/upstream/main' into zhangl…
DoorKickers Jul 25, 2024
00cf16c
change sh to bash to debug
DoorKickers Jul 25, 2024
ffefc1f
test camb bash
DoorKickers Jul 25, 2024
f284df0
remain pin memory to original
DoorKickers Jul 25, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions dipu/scripts/ci/ci_run_one_iter.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -123,8 +123,10 @@ def process_one_iter(log_file, clear_log, model_info: dict) -> None:
cmd_run_one_iter = f"srun --job-name={job_name} --partition={partition} --gres={gpu_requests} --time=40 python {train_path}"
cmd_cp_one_iter = ""
else:
cmd_run_one_iter = f"srun --job-name={job_name} --partition={partition} --gres={gpu_requests} --time=40 sh SMART/tools/one_iter_tool/run_one_iter.sh {train_path} {config_path} {work_dir} {opt_arg}"
cmd_cp_one_iter = f"srun --job-name={job_name} --partition={partition} --gres={gpu_requests} --time=30 sh SMART/tools/one_iter_tool/compare_one_iter.sh {package_name} {atol} {rtol} {metric}"
cmd_run_one_iter = f"bash SMART/tools/one_iter_tool/run_one_iter.sh {train_path} {config_path} {work_dir} {opt_arg}"
cmd_cp_one_iter = f"bash SMART/tools/one_iter_tool/compare_one_iter.sh {package_name} {atol} {rtol} {metric}"
# cmd_run_one_iter = f"srun --job-name={job_name} --partition={partition} --gres={gpu_requests} --time=40 bash SMART/tools/one_iter_tool/run_one_iter.sh {train_path} {config_path} {work_dir} {opt_arg}"
# cmd_cp_one_iter = f"srun --job-name={job_name} --partition={partition} --gres={gpu_requests} --time=30 bash SMART/tools/one_iter_tool/compare_one_iter.sh {package_name} {atol} {rtol} {metric}"
elif device == "ascend":
if "infer" in p2 and "infer" in p3:
cmd_run_one_iter = f"python {train_path}"
Expand Down
30 changes: 18 additions & 12 deletions dipu/scripts/ci/test_one_iter_traditional_model_list.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,18 +1,18 @@
camb:
# # transformers
- model_cfg: "transformers examples/pytorch/question-answering/run_bert_qa.py workdirs_bert"
# - model_cfg: "transformers examples/pytorch/question-answering/run_bert_qa.py workdirs_bert"

# # mmpretrain
- model_cfg: "mmpretrain resnet/resnet50_8xb32_in1k.py workdirs_resnet"
- model_cfg: "mmpretrain swin_transformer/swin-base_16xb64_in1k.py workdirs_swin_transformer"
- model_cfg: "mmpretrain vision_transformer/vit-base-p16_32xb128-mae_in1k.py workdirs_vision_transformer"
# - model_cfg: "mmpretrain swin_transformer/swin-base_16xb64_in1k.py workdirs_swin_transformer"
# - model_cfg: "mmpretrain vision_transformer/vit-base-p16_32xb128-mae_in1k.py workdirs_vision_transformer"
- model_cfg: "mmpretrain mobilenet_v2/mobilenet-v2_8xb32_in1k.py workdirs_mobilenetv2 --no-pin-memory"
precision: {atol: 0.015, metric: 0.015, rtol: 0.01}
- model_cfg: "mmpretrain mobilenet_v3/mobilenet-v3-large_8xb128_in1k.py workdirs_mobilenetv3"
- model_cfg: "mmpretrain efficientnet/efficientnet-b2_8xb32_in1k.py workdirs_efficientnet"
- model_cfg: "mmpretrain convnext/convnext-small_32xb128_in1k.py workdirs_convnext"
- model_cfg: "mmpretrain shufflenet_v2/shufflenet-v2-1x_16xb64_in1k_256.py workdirs_shufflenetv2"
precision: {atol: 0.015, metric: 0.015, rtol: 0.01}
# - model_cfg: "mmpretrain convnext/convnext-small_32xb128_in1k.py workdirs_convnext"
# - model_cfg: "mmpretrain shufflenet_v2/shufflenet-v2-1x_16xb64_in1k_256.py workdirs_shufflenetv2"
# precision: {atol: 0.015, metric: 0.015, rtol: 0.01}

# # mmdetection
- model_cfg: "mmdetection yolo/yolov3_d53_8xb8-320-273e_coco.py workdirs_yolov3"
Expand Down Expand Up @@ -97,13 +97,19 @@ cuda:
# - model_cfg: "mmagic stable_diffusion/stable-diffusion_ddim_denoisingunet_infer.py workdirs_stable_diffusion"

ascend:
# mmsegmentation
# - model_cfg: "mmsegmentation unet/unet-s5-d16_fcn_4xb4-160k_cityscapes-512x1024.py workdirs_unet"
# fallback_op_list: "nll_loss2d_forward,nll_loss2d_backward,native_batch_norm,topk,convolution_overrideable,convolution_backward_overrideable,native_batch_norm_backward"
# - model_cfg: "mmdetection detr/detr_r50_8xb2-150e_coco.py workdirs_detr"
# fallback_op_list: "fill_.Scalar,baddbmm.out,where.self,linear_backward,linear,uniform_,any.all_out,_foreach_addcdiv_.ScalarList,native_batch_norm_backward,convolution_overrideable"
# mmpretrain
# - model_cfg: "mmpretrain resnet/resnet50_8xb32_in1k.py workdirs_resnet"
- model_cfg: "mmpretrain convnext/convnext-small_32xb128_in1k.py workdirs_convnext"
- model_cfg: "mmpretrain efficientnet/efficientnet-b2_8xb32_in1k.py workdirs_efficientnet"
- model_cfg: "mmpretrain mobilenet_v2/mobilenet-v2_8xb32_in1k.py workdirs_mobilenetv2"
precision: {atol: 0.015, metric: 0.015, rtol: 0.01}
- model_cfg: "mmpretrain mobilenet_v3/mobilenet-v3-large_8xb128_in1k.py workdirs_mobilenetv3"
- model_cfg: "mmpretrain resnet/resnet50_8xb32_in1k.py workdirs_resnet"
- model_cfg: "mmpretrain vision_transformer/vit-base-p16_32xb128-mae_in1k.py workdirs_vision_transformer"

# mmsegmentation
- model_cfg: "mmsegmentation deeplabv3/deeplabv3_r50-d8_4xb2-40k_cityscapes-512x1024.py workdirs_deeplabv3"
- model_cfg: "mmsegmentation deeplabv3plus/deeplabv3plus_r50-d8_4xb2-40k_cityscapes-512x1024.py workdirs_deeplabv3plus"
- model_cfg: "mmsegmentation unet/unet-s5-d16_fcn_4xb4-160k_cityscapes-512x1024.py workdirs_unet"

kunlunxin:
# mmpretrain
Expand Down
Loading