Instructions to use DarliAI/kissi-w2v2-lg-xls-r-300m-kinyarwanda with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use DarliAI/kissi-w2v2-lg-xls-r-300m-kinyarwanda with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="DarliAI/kissi-w2v2-lg-xls-r-300m-kinyarwanda")# Load model directly from transformers import AutoProcessor, AutoModelForCTC processor = AutoProcessor.from_pretrained("DarliAI/kissi-w2v2-lg-xls-r-300m-kinyarwanda") model = AutoModelForCTC.from_pretrained("DarliAI/kissi-w2v2-lg-xls-r-300m-kinyarwanda", device_map="auto") - Notebooks
- Google Colab
- Kaggle
w2v2-lg-xls-r-300m-kinyarwanda
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the common_voice_16_1 dataset.
It achieves the following results on the evaluation set:
- eval_loss: 0.590943
- eval_wer: 0.548028
- eval_runtime: 81.4422
- eval_samples_per_second: 19.916
- eval_steps_per_second: 2.493
- epoch: 1.7365
- step: 21600
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 800
- num_epochs: 30
- mixed_precision_training: Native AMP
Framework versions
- Transformers 4.52.4
- Pytorch 2.6.0+cu124
- Datasets 2.14.4
- Tokenizers 0.21.1
- Downloads last month
- 3
Model tree for DarliAI/kissi-w2v2-lg-xls-r-300m-kinyarwanda
Base model
facebook/wav2vec2-xls-r-300m