Skip to content

Latest commit

 

History

History
111 lines (96 loc) · 5.65 KB

README.md

File metadata and controls

111 lines (96 loc) · 5.65 KB
Raw

SSD-RN34 CPU Training

SSD-RN34 Training best known configurations with Intel® Extension for PyTorch.

Model Information

Use Case Framework Model Repo Branch/Commit/Tag Optional Patch
Training PyTorch https://github.com/weiliu89/caffe/tree/ssd - -

Pre-Requisite

  • Installation of PyTorch and Intel Extension for PyTorch

  • Installation of Build PyTorch + IPEX + TorchVision Jemalloc and TCMalloc

  • Installation of oneccl-bind-pt (if running distributed)

  • Set Jemalloc and tcmalloc Preload for better performance

    The jemalloc and tcmalloc should be built from the General setup section.

    export LD_PRELOAD="<path to the jemalloc directory>/lib/libjemalloc.so":"path_to/tcmalloc/lib/libtcmalloc.so":$LD_PRELOAD
export MALLOC_CONF="oversize_threshold:1,background_thread:true,metadata_thp:auto,dirty_decay_ms:9000000000,muzzy_decay_ms:9000000000"

  • Set IOMP preload for better performance

  pip install packaging intel-openmp
  export LD_PRELOAD=path/lib/libiomp5.so:$LD_PRELOAD
  • Set ENV to use fp16 AMX if you are using a supported platform
  export DNNL_MAX_CPU_ISA=AVX512_CORE_AMX_FP16

  • Set ENV to use multi-node distributed training (no need for single-node multi-sockets)

    In this case, we use data-parallel distributed training and every rank will hold same model replica. The NNODES is the number of ip in the HOSTFILE. To use multi-nodes distributed training you should firstly setup the passwordless login (you can refer to link) between these nodes.

    export NNODES=#your_node_number
export HOSTFILE=your_ip_list_file #one ip per line

Prepare Dataset

Download the 2017 COCO dataset using the download_dataset.sh script. Export the DATASET_DIR environment variable to specify the directory where the dataset will be downloaded. This environment variable will be used again when running quickstart scripts.

cd <MODEL_DIR=path_to_ssd-resnet34_training_cpu>
export DATASET_DIR=<directory where the dataset will be saved>
./download_dataset.sh
cd -

Download Pretrained Model

cd <MODEL_DIR=path_to_ssd-resnet34_training_cpu> export CHECKPOINT_DIR= ./download_model.sh

Training

  1. git clone https://github.com/IntelAI/models.git

  2. cd models/models_v2/pytorch/sdd-resnet34/training/cpu

  3. Create virtual environment venv and activate it:

    python3 -m venv venv
. ./venv/bin/activate

  4. Install the latest CPU versions of torch, torchvision and intel_extension_for_pytorch.

  5. Run setup scripts

./setup.sh
  1. Setup required environment paramaters
Parameter export command
TEST_MODE (THROUGHPUT OR ACCURACY) export TEST_MODE=THROUGHPUT
DISTRIBUTED (leave unset if training single node) export DISTRIBUTED=true
NNODES (leave unset if training single node) export NNODES=2
HOSTFILE (leave unset if training single node) export HOSTFILE=<your-host-file>
NUM_RANKS (leave unset if training single node) export NUM_RANKS=1
DATASET_DIR export DATASET_DIR=<path-to-coco>
PRECISION export PRECISION=fp32 <Select from: fp32, avx-fp32, bf16, or bf32>
OUTPUT_DIR export OUTPUT_DIR=<path to an output directory>
CHECKPOINT_DIR export CHECKPOINT_DIR=<path to pre-trained model>
BATCH_SIZE (optional) export BATCH_SIZE=<set a value for batch size, else it will run with default batch size>
  1. Run run_model.sh

Output

The training output looks like this:

Iteration:     80, Loss function: 9.13430500, Average Loss: 0.98246424
Train: [ 90/100]        TrainTime 10.296 (10.232)
Iteration:     90, Loss function: 8.76533031, Average Loss: 1.06040848
:::MLLOG {"namespace": "", "time_ms": 1719249164799, "event_type": "INTERVAL_END", "key": "block_stop", "value": null, "metadata": {"file": "train.py", "lineno": 575, "first_epoch_num": 1, "epoch_count": 70}}
train latency 45.67 ms
train performance 21.90 fps
Throughput: 21.897 fps

Final results of the training run can be found in results.yaml file.

results:
- key : throughput
  value: 21.897
  unit: fps
- key: latency
  value: 45.67
  unit: ms
- key: accuracy
  value: 0.20004
  unit: percentage