data2vec audio 960h needs about 0.6 GB of memory at Q8_0 (a 0.1 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 3 quants, from 0.1 GB (Q4_K) to 0.2 GB (F16).
Data2Vec Audio 960h – 95M‑parameter wav2vec2‑style acoustic model
Data2Vec Audio 960h is a 95M‑parameter wav2vec2‑style acoustic model. It was pretrained on 960 h of Librispeech and fine‑tuned with a CTC head. The model is available in GGUF format for use with CrispASR.
Good at
- Accurate English speech transcription
- Self‑supervised learning framework
- CTC‑based acoustic modeling
Limits
- No vision support
- English only
- Requires 16 kHz audio input
Best for
- Transcribing English audiobooks
- Speech recognition in research
- Evaluating ASR benchmarks
Who it’s for. Developers and researchers needing a lightweight, open‑source acoustic model for English speech. The model is suitable for integration into ASR pipelines that support GGUF and CrispASR.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K | 0.1 GB | 0.6 GB | Fits | Get file → | |
| Q8_0 | 0.1 GB | 0.6 GB | Fits | Get file → | |
| F16 | 0.2 GB | 0.7 GB | Fits | Get file → |
For 8 GB VRAM, use the Q4_K quantized file (0.08 GB, ~0.6 GB memory at 8 k context). For 12 GB VRAM, the Q8_0 file (0.12 GB, ~0.6 GB memory) is also suitable. The F16 file (0.2 GB, ~0.7 GB memory) requires more memory but offers full precision.
What GPU do you need for data2vec audio 960h?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is F16 (0.2 GB file, about 0.7 GB in use).
Which data2vec audio 960h quant should you choose?
Q8_0 (0.1 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
If memory is tight, the smallest file is Q4_K at 0.1 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about data2vec audio 960h
How much VRAM does data2vec audio 960h need?
About 0.6 GB at Q8_0 with an 8K context window. That is the 0.1 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run data2vec audio 960h on an 8 GB GPU?
Yes, with F16: the 0.2 GB file needs about 0.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of data2vec audio 960h GGUF?
It depends on the quant: Q4_K is 0.1 GB, Q8_0 is 0.1 GB, F16 is 0.2 GB.
Where can I download data2vec audio 960h GGUF?
The files are on Hugging Face in the cstr/data2vec-audio-960h-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does data2vec audio 960h use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run this model on 8 GB VRAM?
Yes, the Q4_K quantized variant uses 0.08 GB and requires about 0.6 GB memory at 8 k context, fitting comfortably on 8 GB VRAM.
Does this model support other languages?
No, the model is trained and fine‑tuned only on English data and supports English speech.
What license applies to the GGUF files?
The GGUF files are distributed under the Apache‑2.0 license, the same as the upstream model.