HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks

whisper-small-waxal-ful

waxal-benchmarking/whisper-small-waxal-ful

kind
original
quant
bf16 (16-bit)
family
misc
modality
automatic-speech-recognition
root model
openai/whisper-small
provider
waxal-benchmarking
license
apache-2.0
language
ful
context
—
library
transformers
vRAM (GB)
—
size on disk
—
downloads
412
likes
0
created
—
last modified
2026-06-05T22:41:46+00:00 (2026)
huggingface
https://huggingface.co/waxal-benchmarking/whisper-small-waxal-ful
tags
transformerssafetensorswhisperautomatic-speech-recognitionafrican-languageswaxalwaxalnetfuldataset:waxal-benchmarking/waxalarxiv:2606.02375base_model:openai/whisper-smallbase_model:finetune:openai/whisper-smalllicense:apache-2.0endpoints_compatibleregion:us

Benchmark scores (6)

benchmarkcategoryscorecontextconfevidence
WERspeech-asr44.82low| 1.1753 | 0.8439 | 500 | 0.5772 | 0.4482 | 0.1283 |
WERspeech-asr42.6**Test WER**medium| **Test WER** | 42.6% |
WERspeech-asr42.6**WER**medium| **WER** | 42.6% |
CERspeech-asr12.83low| 1.1753 | 0.8439 | 500 | 0.5772 | 0.4482 | 0.1283 |
CERspeech-asr12.0**Test CER**medium| **Test CER** | 12.0% |
CERspeech-asr12.0**CER**medium| **CER** | 12.0% |

Discussions (0)

none scraped.

Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.