README.md

metadata

license: apache-2.0

Optimum Habana is the interface between the Hugging Face Transformers and Diffusers libraries and Habana's Gaudi processor (HPU). It provides a set of tools enabling easy and fast model loading, training and inference on single- and multi-HPU settings for different downstream tasks. Learn more about how to take advantage of the power of Habana HPUs to train and deploy Transformers and Diffusers models at hf.co/hardware/habana.

Swin Transformer model HPU configuration

This model only contains the GaudiConfig file for running the Swin Transformer model on Habana's Gaudi processors (HPU).

This model contains no model weights, only a GaudiConfig.

This enables to specify:

• use_fused_adam: whether to use Habana's custom AdamW implementation
• use_fused_clip_norm: whether to use Habana's fused gradient norm clipping operator
• disable_autocast: whether to disable autocast; this parameter takes precedence over --bf16 flag and is temporary as some scripts produce nan values. In those cases this parameter is already present in huggingface topology Habana gaudi_config.json.

Usage

The model is instantiated the same way as in the Transformers library. The only difference is that there are a few new training arguments specific to HPUs. This model is supported only in mixed precision training with bf16 type.

Here is an image classification example script to fine-tune a model. You can run it with Swin with the following command:

python run_image_classification.py \
    --model_name_or_path microsoft/swin-base-patch4-window7-224-in22k \
    --dataset_name cifar10 \
    --output_dir /tmp/outputs/ \
    --remove_unused_columns False \
    --do_train \
    --do_eval \
    --learning_rate 3e-5 \
    --num_train_epochs 5 \
    --per_device_train_batch_size 64 \
    --per_device_eval_batch_size 64 \
    --evaluation_strategy epoch \
    --save_strategy epoch \
    --load_best_model_at_end True \
    --save_total_limit 3 \
    --seed 1337 \
    --use_habana \
    --use_lazy_mode \
    --gaudi_config_name Habana/swin \
    --throughput_warmup_steps 2 \
    --ignore_mismatched_sizes \
    --bf16

Check the documentation out for more advanced usage and examples.