/StorageDeivceModel

Primary LanguagePureBasic

StorageDeivceModel

This repo contains the codes of the exp7 in the 09-02 weekly report.

  1. Files:

    dataset_7: contains the train and test dataset.

    model: the storageNet

    TRACEPROFILE: trace after replay (for the generation of the true LDS)

    TRACETOREPLAY: trace to replay (for the generation of the prev LDS (input))

    weights: save the weights of the model

    eval.py: run the multi-step prediction

    parser & parser_script.sh: generate the ‘dataset_7’ dataset from TRACEPROFILE & TRACETOREPLAY

    train.py train storageNet model

  2. Generate the dataset:

    ./parser_script.sh

  3. train: (around 10 min on my local machine)

    python3 train.py

  4. multi-step evaluation:

    python eval.py