Downloads · 30 days
2
5% of all-time downloads
wav2vec2-large-960h-lv60-self-neuronx-audio-classification
Jingya/wav2vec2-large-960h-lv60-self-neuronx-audio-classification
wav2vec2-large-960h-lv60-self-neuronx-audio-classification is a automatic speech recognition model from Jingya. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
This is model is compiled explicitly for AWS Neuronx(inferentia 2 / trainium 1) with the following codes:
All-time downloads
37
Public
Repo size
514 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.neuron514 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- apache-2.0
- Model type
- wav2vec2
- Access
- Public
- Created
- Jul 9, 2024
- Updated
- Jul 9, 2024
- SHA
- 5ad2e536
Model card
From the Hugging Face model README
This is model is compiled explicitly for AWS Neuronx(inferentia 2 / trainium 1) with the following codes:
from datasets import load_dataset
from transformers import AutoProcessor
from optimum.neuron import NeuronModelForAudioClassification, pipeline
dataset = load_dataset("hf-internal-testing/librispeech_asr_demo", "clean", split="validation")
dataset = dataset.sort("id")
sampling_rate = dataset.features["audio"].sampling_rate
model_id = "facebook/wav2vec2-large-960h-lv60-self"
processor = AutoProcessor.from_pretrained(model_id)
input_shapes = {"batch_size": 1, "audio_sequence_length": 100000}
compiler_args = {"auto_cast": "matmul", "auto_cast_type": "bf16"}
model = NeuronModelForAudioClassification.from_pretrained(
model_id,
export=True,
disable_neuron_cache=True,
**input_shapes,
**compiler_args,
)
model.save_pretrained("wav2vec2_neuron")
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- License
- apache-2.0
- Created
- Jul 9, 2024
- Updated
- Jul 9, 2024