---
library_name: transformers
license: apache-2.0
datasets:
- jacktol/ATC-ASR-Dataset
- jlvdoorn/atcosim
language:
- en
metrics:
- wer
base_model:
- mistralai/Voxtral-Mini-3B-2507
tags:
- audio
- automatic-speech-recognition
- en-atc
- en
- noisy-speech-recognition
- speech-recognition
---

# Voxtral 3B fine tuned for Air Traffic Control Audio Recognition

Audio/Text model fine-tuned on Air Traffic Control (ATC) data.

While there are several Whisper based ATC transcription models, at time of publishing this is the first Voxtral based one.

## WER
| Dataset                                     | this model | untrained base | tclin/whisper-large-v3-turbo-atcosim-finetune | jacktol/whisper-large-v3-finetuned-for-ATC |
| --------------------------------------------|------------|----------------|-----------------------------------------------|--------------------------------------------|
| Typical noise (jacktol/ATC-ASR-Dataset test)|  8.0%      | 105.6%         | N/A                                           | 6.5% (reported)
| Low noise (jlvdoorn/atcosim validation)     |  1.3%      | 83.2%          | 3.7% (reported)                               | N/A

## Model Details

### Model Description

- **Developed by:** Philip Pilgerstorfer
- **Model type:** ASR/Transcription
- **Language(s) (NLP):** English (ATC) with local variations
- **License:** Apache 2.0
- **Finetuned from model:** Voxtral 3B 2507

### Model Sources [optional]

<!-- Provide the basic links for the model. -->

- **Repository:** WIP
 
## Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

### Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

[More Information Needed]

### Downstream Use [optional]

<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->

[More Information Needed]

### Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

[More Information Needed]

## Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

[More Information Needed]

### Recommendations

<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.

## How to Get Started with the Model

Use the code below to get started with the model.

[More Information Needed]

## Training Details

Checkpoint is after ca. 7 epochs, 23h training on an Nvidia 3090 Ti (24GB VRAM)

* Training and validation set of `jacktol/ATC-ASR-Dataset` (typical VHF transmission noise)
* Training set of `jlvdoorn/atcosim` (low noise environment)


### Compute Infrastructure

[More Information Needed]

#### Hardware

[More Information Needed]

#### Software

[More Information Needed]

## Citation [optional]

<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->

**BibTeX:**

[More Information Needed]

**APA:**

[More Information Needed]

## Glossary [optional]

<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->

[More Information Needed]

## More Information [optional]

[More Information Needed]

## Model Card Authors [optional]

[More Information Needed]

## Model Card Contact

[More Information Needed]