PocketFlow

PocketFlow copied to clipboard

InternalError (see above for traceback): Blas SGEMM launch failed : m=131072, n=16, k=16

1

I ran the command "scripts/run_local.sh nets/resnet_at_cifar10_run.py --learner channel --cp_prune_option list" There was a process in GPU:0 before I ran this command. I checked the GPU states by "Nvidia -smi", so...

ShuteLee

Channel Pruning should reCreatePruner in channel pruning

10

在DDPG训练完也就是__prune_rl()后，应该再加一个self.create_pruner()吧，如果不加这个，感觉是在RL最后一次的compress上应用新的pruning，这应该不是正解吧！！！ 感觉还是重新create_pruner()比较好一点。各位大佬看看是不是这样子？ 

Nankaiming

where should I put my self-trained model, got an error when trying:Key model/resnet_model/batch_normalization/beta not found in checkpoint

3

I'm trying to use my own model to run the channel pruning, I restore the trained model (isn't pruned) in /home/l/PocketFlow/models. and changed save_path in the abstract_learner.py, tf.app.flags.DEFINE_string('save_path', './models/resnet_model', 'model\'s...

doublelipeng

Does the PocketFlow support Inception V3 and YOLO V3 models?

1

I have two cases, one use the Inception V3 Regression model , and the other use the YOLO V3 model。 I noticed that the demos in Pocketflow are for LeNet...

mldlli

TOCO failed of channel pruning, MobileNetV2 at Cifar10

11

log of running export_chn_pruned_tflite_model.py: 2018-12-12 05:19:28.614521: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 AVX512F FMA 2018-12-12 05:19:29.483043: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Found device...

jiaxiangxu

A Little Problem in the part of Algorithmic Description (convolutional operation)

1

- Doc Link: https://pocketflow.github.io/cpr_learner/ **Describe the documentation issue** There is a little problem in the algorithm description section. When it comes to "The matrix multiplication can be decomposed along the...

my777777

This template is for miscellaneous issues not covered by the other issue categories. For questions on how to work with PocketFlow, or support for problems that are not verified bugs...

to-be-snail

Change "batch_size" to "batch_size_eval" in __build_eval function

1

**Describe the bug** The __build_eval function in learner.py uses batch_size instead of batch_eval_size for the shape of evaluation images, ie images.set_shape((FLAGS.**batch_size**, images.shape[1], images.shape[2], \ needs to be changed to images.set_shape((FLAGS.**batch_size_eval**,...

rapwag01

**Describe the bug** When I tried to use the UniformQuantTFLearner to compress the VGG_SSD model, I got nan loss in retrain stage. **To Reproduce** Steps to reproduce the behavior: 1....

howtocodewang

ChannelPruner evaluate get unstable accuarcy

1

I tried to evaluate a compressed model generated by channel prune learner, however the accuracy is unstable and lower than full precision evaluation. My resnet56_cifar10 checkpoint accuracy varies such as...

xiaomr