benchmarks

benchmarks copied to clipboard

3090 and following versions: Windows 10 python 3.9.5 tensorflow 2.5 CUDA 11.2.2 (path set) CuDNN 8.1 fp32 works: python tf_cnn_benchmarks.py --num_gpus=1 --batch_size=128 --model=resnet50 --variable_update=parameter_server fp16 not: python tf_cnn_benchmarks.py --num_gpus=1 --batch_size=128...

anefischer

When I run the program with "python tf_cnn_benchmarks.py --data_format=NCHW --batch_size=256 \ --model=resnet50 --optimizer=momentum --variable_update=replicated \ --nodistortions --gradient_repacking=8 --num_gpus=8 \ --num_epochs=90 --weight_decay=1e-4 --data_dir=${DATA_DIR} --use_fp16 \ --train_dir=${CKPT_DIR}". The final test accuracy is...

lljjgg

` viz_utils.visualize_boxes_and_labels_on_image_array( image_np_with_detections, detections['detection_boxes'], detections['detection_classes']+label_id_offset, category_index, use_normalized_coordinates=True, max_boxes_to_draw=5, min_score_thresh=.8, agnostic_mode=False)  `

shridharnator

not using GCE instance

2

Is there away to use the benchmark.py but not upload to GCE? Im running it on my local system in a docker container but its turning out to be a...

jjziets

libcudart.so.10.1 (and others) are not in the built docker image

6

Steps to reproduce the error: - Docker image is built with: `python perfzero/lib/setup.py --tensorflow_pip_spec=tensorflow==2.3.0` - Start image with: `docker run -it --gpus all --rm -v $(pwd):/workspace perfzero/tensorflow bash` - When...

gaborvecsei

This is useful if the base image has been created locally. Looking for a nonexistent image in a remote repo results in an error. For tf-agents we build a base...

tfboyd

@reedwm I did some performance tests of Resnet50 (also _v1.5 and v2) on Tesla T4 and V100 GPUs (1-8). I found the input pipeline made by `data_flow_ops.RecordInput + data_flow_ops.StagingArea` generally...

zhao1157

The link to estimator_benchmark.py in the PerfZero readme (see https://github.com/tensorflow/benchmarks/blob/master/perfzero/README.md#creating-tests) is broken

nsburbank

PerfZero without Docker fails with AccessDeniedException

3

``` raise Exception('"{}" failed with code:{} and stdout:\n{}'.format( Exception: "['gsutil', '-m', 'cp', '-r', '-n', 'gs://tf-performance/auth_tokens/benchmark_upload_gce.json', '/home/hotaru/tensorflow-benchmarks/perfzero/workspace']" failed with code:1 and stdout: AccessDeniedException: 403 [email protected] does not have storage.objects.list access to...

lilyanatia

How to replace AllReduce with Reduce in parameter server mode?

4

https://github.com/tensorflow/benchmarks/blob/5d03cf8e356d2ae17df440cdb612c378cbacf5ef/scripts/tf_cnn_benchmarks/variable_mgr_util.py#L575 @reedwm In parameter server mode, I managed to replace `grad = tf.add_n(grads)` by `nccl all reduce`: #nccl all reduce sum_grads = allreduce.sum_grad_and_var_all_reduce( False, grad_and_vars, 1, 'nccl', gpu_indices) #get the...

zhao1157