harvard-acc / DeepRecSys

http://vlsiarch.eecs.harvard.edu/research/recommendation/
MIT License
130 stars 25 forks source link

Running DLRM with synthetic inputs #5

Closed rajesh-iiith closed 2 years ago

rajesh-iiith commented 3 years ago

After I generated synthetic trace using data_generator/trace_generator.py, I am trying to run the dlrm_s_caffe2 in synthetic mode (using the command below). It keeps compiling about "'Namespace' object has no attribute 'data_size'". If I try to pass some value by --data_size, it doesn't recognize this argument.

python dlrm_s_caffe2.py --inference_only --caffe2_net_type async_dag --config_file "configs/dlrm_rm2.json" --engine "prof_dag" --nepochs 100 --data_generation "synthetic" --round_targets 10 --num_batches 512 --mini_batch_size 16 --max_mini_batch_size 16 --data_trace_file "../data_generator/syn_traces/tbl1"

Am I doing something wrong? Can you provide a sample command line to run dlrm with synthetic inputs?

alugupta commented 3 years ago

Hi @rajesh-iiith , I am curious if you have been able to run some of the provided experiments, like the operator breakdown sweep: https://github.com/harvard-acc/DeepRecSys/tree/master/experiments/operator_breakdown

This will help know whether it is a tool/versioning issue or if there is something else going on.

You can also print out the commands being issued in that sweep (link) as examples.