|

|

soloni-114m-tdt-ctc-v0 is a fine tuned version of nvidia's parakeet-tdt_ctc-110m that transcribes bambara language speech. Unlike its base model, this model cannot write Punctuations and Capitalizations since these were absent from its training. The model was fine-tuned using NVIDIA NeMo and supports both TDT (Token-and-Duration Transducer) and CTC (Connectionist Temporal Classification) decoding.

This model, along with its associated resources, is part of an ongoing research effort, improvements and refinements are expected in future versions. Users should be aware that:

To fine-tune or play with the model you will need to install NVIDIA NeMo. We recommend you install it after you've installed latest PyTorch version.

Note that this model has been released for research purposes primarily.

Note that the decoding strategy for the TDT decoder use CUDA Graphs by default but not all GPUs and versions of cuda support that parameter. If you run into a RuntimeError: CUDA error: invalid argument you should set that argument to false in the decoding strategy before calling asr_model.transcribe()

This model accepts 16 kHz mono-channel audio (wav files) as input. But it is equipped with its own preprocessor doing the resampling so you may input audios at higher sampling rates.

This model provides transcribed speech as an hypothesis object with a text attribute containing the transcription string for a given speech sample.

This model uses a Hybrid FastConformer-TDT-CTC architecture. FastConformer is an optimized version of the Conformer model with 8x depthwise-separable convolutional downsampling. It possesses two independant but jointly trained decoder, one auto-regressive TDT decoder and a convolutional decoder with CTC loss. You may find more information on the details of FastConformer here: Fast-Conformer Model.

The NeMo toolkit was used for finetuning this model for 16,296 steps over parakeet-tdt_ctc-110m model.The finetuning codes and configurations can be found at RobotsMali-AI/bambara-asr.

The tokenizer for this model was trained on the text transcripts of the train set of RobotsMali/bam-asr-early using this script.

This model was fine-tuned on the bam-asr-early dataset, which consists of 37 hours of transcribed Bambara speech data. The dataset is primarily derived from Jeli-ASR dataset (~87%).

The performance of Automatic Speech Recognition models is commonly measured using Word Error Rate (WER) and and Character Error Rate (CER). Since this model has two decoders operating independently at inference time, each decoder is evaluated independently too.

The following table shows these two metrics for each decoder:

These are greedy WER numbers without external LM. By default the main decoder branch is the TDT branch, if you would like to switch to the CTC decoder simply run this block of code before calling the .transcribe method

This model is released under the CC-BY-4.0 license. By using this model, you agree to the terms of the license.

Feel free to open a discussion on Hugging Face or file an issue on GitHub for help or contributions.