Datasets:
Download LICENSE.md from Squadstack/conversational-streaming-asr-benchmark: direct link, hf CLI and curl.
- Browser
- Download file 3.43 kB
-
https://huggingface.co/datasets/Squadstack/conversational-streaming-asr-benchmark/resolve/main/LICENSE.md
- Command line
-
hf download hf://datasets/Squadstack/conversational-streaming-asr-benchmark/LICENSE.md
-
curl -L -o LICENSE.md https://huggingface.co/datasets/Squadstack/conversational-streaming-asr-benchmark/resolve/main/LICENSE.md
Licensing
This release is not a single artefact under a single licence. Four components have genuinely different rights, and conflating them would claim rights over the audio that SquadStack does not hold.
| Component | Licence |
|---|---|
| Audio recordings | Benchmark Evaluation Licence 1.0 (below) |
| Reference transcripts and annotations | CC BY-NC 4.0, attribution to SquadStack |
| Compilation, schema and metadata | CC BY 4.0 |
| Evaluation code (harness and metrics — not in this repository; supplied on request) | Apache License 2.0 |
The access terms you accept at the gate are in TERMS_OF_USE.md.
Benchmark Evaluation Licence 1.0 — audio
Copyright © 2026 SquadStack Technologies Pvt. Ltd. All rights reserved.
The audio recordings are confidential business data of SquadStack and its clients, containing the voices of identifiable individuals. They are licensed, not sold, and only to accounts approved through the Hugging Face access gate.
You may
- Use the audio to evaluate automatic speech recognition systems, including commercially.
- Publish evaluation results, including comparative results and results that are unfavourable to SquadStack. No prior approval is required and none will be asked for.
- Analyse the audio for research into speech recognition, code-switching, speech quality or related topics, and publish that analysis.
- Retain your copy for as long as you are evaluating, subject to any takedown notice.
You may not
- Train, fine-tune, distil, or otherwise fit any model on this audio or its transcripts, at any stage, including as evaluation-driven model selection.
- Redistribute the audio, in whole or in part, in any form, to anyone — including inside your own organisation beyond the individuals covered by your approved access.
- Attempt to determine the identity of any speaker, or link a recording to any individual, account, phone number or customer record.
- Build speaker identification, speaker verification, speaker diarization-for-identification, or voice cloning or synthesis systems from this data.
- Use the audio to contact, profile, target or market to any individual appearing in it.
- Remove, obscure or work around the access gate, this licence, or the terms of use.
Attribution
Published results must cite the dataset (BibTeX in the dataset card) and state the version.
Termination
This licence terminates automatically on breach. On termination or on a takedown notice, delete all copies of the audio and confirm deletion in writing to SquadStack.
Warranty and liability
Provided "as is", without warranty of any kind. SquadStack is not liable for any claim, damage or other liability arising from the data or its use.
Governing law
Laws of India. Courts at Gautam Buddha Nagar, Uttar Pradesh.
Third-party material
Figures are set in Albert Sans and Noto Sans Devanagari, both under the SIL Open Font License 1.1. Neither font is redistributed here.
Questions
Reach out to SquadStack. If your intended use is not obviously covered above, ask — the answer is usually yes, and a written yes is worth more than an assumption.