whisper-large-v3-basque is an open-weight model from Andoni Sudupe, released under Apache License 2.0. It has 1.6B parameters. At 16-bit it needs about 3.9 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index. It draws 171 downloads a month.
This model is a fine-tuned version of openai/whisper-large-v3 on the None dataset.
Runs On
What it takes to serve whisper-large-v3-basque (1.6B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 3.2 GB | 3.9 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 1.6 GB | 1.9 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.8 GB | 1.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 7, 2026.
whisper-large-v3-basque on every accelerator the SAVRN Index prices, at every precision
Model Card
By Andoni Sudupe, published under apache-2.0, revision 51f752b4fc3d.
This model is a fine-tuned version of openai/whisper-large-v3 on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.2181 - Wer: 14.7372
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: 1e-05 - train_batch_size: 256 - eval_batch_size: 32 - seed: 42 - distributed_type: multi-GPU - num_devices: 2 - total_train_batch_size: 512 - total_eval_batch_size: 64 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - training_steps: 5000 - mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.1741 | 0.42 | 500 | 0.2910 | 23.2171 |
| 0.1388 | 0.84 | 1000 | 0.2418 | 18.3800 |
| 0.1084 | 1.26 | 1500 | 0.2245 | 16.7258 |
| 0.1041 | 1.68 | 2000 | 0.2155 | 15.6607 |
| 0.0793 | 2.1 | 2500 | 0.2108 | 15.0738 |
| 0.0794 | 2.52 | 3000 | 0.2102 | 15.1867 |
| 0.0774 | 2.94 | 3500 | 0.2072 | 14.7105 |
| 0.0607 | 3.36 | 4000 | 0.2142 | 14.7454 |
| 0.0608 | 3.78 | 4500 | 0.2117 | 14.6900 |
| 0.0506 | 4.2 | 5000 | 0.2181 | 14.7372 |
Configuration
- Architecture
- WhisperForConditionalGeneration
- Layers
- 32
- Vocabulary size
- 51,866
- Stored precision
- float16
- Model type
- whisper
Identity and Version
- Repository
- Ansu/whisper-large-v3-basque
- Publisher
- Andoni Sudupe
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 1.6B parameters
- Languages
- Not stated by the source
- Revision
- 51f752b4fc3d0e74ee23de20331c2d0de5be7104
- First published
- 2026-08-29
- Last updated
- 2026-09-30
Files and Weights
14 files, 3.2 GB in total. The weights are 2 files totalling 3.2 GB in bin, safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 3.2 GB | 43654bf0bd75 |
| training_args.bin | Weights | 6.9 KB | baeef663a1d4 |
| added_tokens.json | Configuration | 34.6 KB | — |
| config.json | Configuration | 1.3 KB | — |
| generation_config.json | Configuration | 3.9 KB | — |
| normalizer.json | Configuration | 52.7 KB | — |
| preprocessor_config.json | Configuration | 340 B | — |
| special_tokens_map.json | Configuration | 2.2 KB | — |
| README.md | Documentation | 2.1 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
| merges.txt | Tokenizer | 493.9 KB | — |
| tokenizer.json | Tokenizer | 2.5 MB | — |
| tokenizer_config.json | Tokenizer | 282.8 KB | — |
| vocab.json | Tokenizer | 835.5 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 3.2 GB
Released by Andoni Sudupe through its official repository on Hugging Face. Read the license.
Built From
- Derived from openai/whisper-large-v3
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 3.2 GB |
| 16-bit | 3.2 GB |
| 8-bit | 1.6 GB |
| 4-bit | 0.8 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About whisper-large-v3-basque
How much GPU memory does whisper-large-v3-basque need?
About 3.9 GB at 16-bit and 1 GB at 4-bit: the weights (1.6B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run whisper-large-v3-basque on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use whisper-large-v3-basque commercially?
Yes. whisper-large-v3-basque is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.