Commit 860efcf0 authored by Yuxin Wu's avatar Yuxin Wu

[FasterRCNN] using the wrong schedule for the published result

parent 17261566
...@@ -47,7 +47,7 @@ To evaluate the performance (pretrained models can be downloaded in [model zoo]( ...@@ -47,7 +47,7 @@ To evaluate the performance (pretrained models can be downloaded in [model zoo](
Mean Average Precision @IoU=0.50:0.95: Mean Average Precision @IoU=0.50:0.95:
+ trainval35k/minival, FASTRCNN_BATCH=256: 34.2. Takes 49h on 8 TitanX. + trainval35k/minival, FASTRCNN_BATCH=256: 34.2. Takes 49h on 8 TitanX.
+ trainval35k/minival, FASTRCNN_BATCH=64: 32.7. Takes 25h on 8 TitanX. + trainval35k/minival, FASTRCNN_BATCH=64: 32.7. Takes 31h on 8 TitanX.
The hyperparameters are not carefully tuned. You can probably get better performance by e.g. training longer. The hyperparameters are not carefully tuned. You can probably get better performance by e.g. training longer.
......
...@@ -320,12 +320,12 @@ if __name__ == '__main__': ...@@ -320,12 +320,12 @@ if __name__ == '__main__':
'learning_rate', 'learning_rate',
[(warmup_epoch * factor, 1e-2), [(warmup_epoch * factor, 1e-2),
(150000 * factor // stepnum, 1e-3), (150000 * factor // stepnum, 1e-3),
(210000 * factor // stepnum, 1e-4)]), (230000 * factor // stepnum, 1e-4)]),
EvalCallback(), EvalCallback(),
GPUUtilizationTracker(), GPUUtilizationTracker(),
], ],
steps_per_epoch=stepnum, steps_per_epoch=stepnum,
max_epoch=230000 * factor // stepnum, max_epoch=280000 * factor // stepnum,
session_init=get_model_loader(args.load) if args.load else None, session_init=get_model_loader(args.load) if args.load else None,
) )
trainer = SyncMultiGPUTrainerReplicated(get_nr_gpu()) trainer = SyncMultiGPUTrainerReplicated(get_nr_gpu())
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment