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.
Wav2Vec2 model HPU configuration
This model only contains the GaudiConfig file for running the Wav2Vec2 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 implementationuse_fused_clip_norm: whether to use Habana's fused gradient norm clipping operatordisable_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.
Here is an audio classification example script to fine-tune a model. You can run it with Wav2Vec2 with the following command:
python run_audio_classification.py \
--model_name_or_path facebook/wav2vec2-base \
--dataset_name superb \
--dataset_config_name ks \
--output_dir /tmp/wav2vec2-base-ft-keyword-spotting \
--overwrite_output_dir \
--remove_unused_columns False \
--do_train \
--do_eval \
--learning_rate 3e-5 \
--max_length_seconds 1 \
--attention_mask False \
--warmup_ratio 0.1 \
--num_train_epochs 5 \
--per_device_train_batch_size 256 \
--per_device_eval_batch_size 256 \
--dataloader_num_workers 4 \
--seed 27 \
--use_habana \
--use_lazy_mode \
--gaudi_config_name Habana/wav2vec2 \
--throughput_warmup_steps 2 \
--bf16
Check the documentation out for more advanced usage and examples.