simplew2011

inference a batch images in one time

2

how can i using efficientnet to inference a batch images in one time, not circle inference in every image.

CRNN keras训练代码

1

好像train文件没有包含keras版的训练脚本

train default dataset, acc unchange.

5

Fri Dec 11 22:08:15 2020 Fold 2, Epoch 11, lr: 0.0000849, train loss: 0.03911, valid loss: 0.00720, acc: 47.3684, auc: 0.500000. Fri Dec 11 22:08:15 2020 Fold 2, Epoch 12...

预测结果为全灰图

2

 - 默认数据集，训练5轮 - 预测发现全部是灰图，没有结果 看loss比较大，loss: 0.5252 - accuracy: 0.7814，就是没有模型收敛。 修改loss，lr，Adam(lr = 1e-5)，steps_per_epoch=2000, epochs=20，batch_size=32。 loss缓慢下降，正常收敛，预测正常 关键是batch_size和lr需要微调，多调整几次看loss有没有下降趋势，没有的话中断训练，继续修改 2022-09-13 05:10:38.208563: I tensorflow/stream_executor/cuda/cuda_dnn.cc:384] Loaded cuDNN version 8201 2000/2000 [==============================] - ETA: 0s -...

No such file or directory: '/tmp/tmp1pvg2dc5/tmpbf_w1ds6'

1

**Describe the bug** https://github.com/alibaba/data-juicer/issues/104 **To Reproduce** ``` pip install -U scalene pip install py-data-juicer git clone https://github.com/alibaba/data-juicer cd data-juicer scalene tools/process_data.py --config configs/demo/process.yaml ``` **Screenshots** ``` 2023-11-29 11:13:15 | INFO...

output sr image shape is same as input lr image，no super resolution

markdown不支持公式

use release weight to eval, Lane Detection Result is wrong. 

There are black spots in the prediction results

3


how to deploy to onnx or torchscript.