Benchmark for deeper models #64
Comments
I haven't try training from scratch with R-101 backbone. What's your results with R-101-FPN and how did you training it (command, number of GPUs) ? |
@li-js @roytseng-tw @Rizhiy I tried three experiments, 2x over 4 GPUs and 1x over 8GPUs:
Note, the codes do produce expected numbers if I do evaluation using detectron checkpoints. Evaluation command Any thoughts? |
For the Res101-FPN, it is not really training from scratch, as the ImageNet pretrained weights from Caffe is loaded. For the settings, I used 4 GPUs (GeForce GTX 1080 Ti) with python3, pytorch 0.3.1.post2 and cuda 8.0. I have two sets of results. Results: Seg AP: 0.333, Box AP: 0.369 on last step. Set 2: Results: Seg AP: 0.336, Box AP: 0.368 on last step. The results are similar to R-50-FPN. |
@li-js could you share your settings for R-50-FPN that reproduced the desired numbers? |
@fitsumreda Sure |
Thank you so much, @li-js ! |
@li-js Did you modify NUM_GPUS in the config file ? If yes, do not. I have already emphasized that in README. Maybe I should make it clearer. |
I did modify the NUM_GPUS to be 4 in my case. Thanks for pointing it out. So if I only have 4 GPUs and each GPU can only hold 1 image, what is the suggested training schedule? Since the Max_Iter and BASE_LR will be adjusted automatically, am I right to just use the cfg file here unchanged and use the following command? python3 tools/train_net_step.py --dataset coco2017 --cfg configs/e2e_mask_rcnn_R-101-FPN_2x.yaml --bs 4 And use the following for 4GPUs and each GPU can hold 2 images: Correct me if I am wrong. |
Yes, you are correct. 😃 |
Thanks, closing here. In official Detectron, the ResNeXt series backbone all use 1 images per batch due to memory constraints, yet they still have even better performance than R-101 series. Still looking forward to a benchmark on R-101-FPN/ResNext-series if anyone successfully reproduces the results. 💯 |
@roytseng-tw With your suggestions, I trained with:
without changing the config file. I got better performance AP seg 34.5, AP det 38.5, but still not matching official Detectron's AP det 40, AP seg 35.9. Any suggestions are appreciated. |
I think these numbers may be reasonable on my experience. When I trained |
Thanks for sharing the great code!
I can also get similar AP for both box and segm with R-50-FPN model, as confirmed in Issue #24.
I am wondering if there are some benchmark results for deeper models like R-101-FPN. On my side, the results for R-101-FPN is not as good as the one in Detectron. Do you guys reproduce the performance of Detectron (box ap 40, segm ap 35.9) for R-101-FPN @roytseng-tw @Rizhiy?
The text was updated successfully, but these errors were encountered: