HF Model & Benchmark Matrix

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

MERaLiON-3-10B

MERaLiON/MERaLiON-3-10B

kind
original
quant
bf16 (16-bit)
family
misc
modality
automatic-speech-recognition
root model
openai/whisper-large-v3
provider
MERaLiON
license
other
language
en, zh, ms, ta, id, th, vi
context
—
library
transformers
vRAM (GB)
—
size on disk
—
downloads
530
likes
8
created
—
last modified
2026-06-17T06:04:35+00:00 (2026)
huggingface
https://huggingface.co/MERaLiON/MERaLiON-3-10B
tags
transformerssafetensorsmeralion3automatic-speech-recognitionmeralionmeralion-3custom_codeenzhmstaidthvidataset:MERaLiON/Multitask-National-Speech-Corpus-v1arxiv:2412.09818arxiv:2501.01034arxiv:2511.04914arxiv:2508.07273arxiv:2409.06635arxiv:2501.08335base_model:google/gemma-2-9b-itbase_model:finetune:google/gemma-2-9b-itlicense:other

Benchmark scores (4)

benchmarkcategoryscorecontextconfevidence
WER FLEURSspeech-asr99.69fleursmedium| fleurs | en | eng | 99.69 | 58.27 | <u>**100.00**</u> | 73.00 | 78.00 |
WER Common Voicespeech-asr76.84Age (commonvoice-en, ta, th, vi, zh)medium| Age (commonvoice-en, ta, th, vi, zh) | 76.84 | 61.77 | 70.38 | <u>**77.00**</u> | 68.90 |
WER Common Voicespeech-asr64.86Commonvoicemedium| Commonvoice | en | eng | 64.86 | 63.10 | 64.20 | <u>**68.00**</u> | 65.00 |
WER Common Voicespeech-asr45.2commonvoicemedium| commonvoice | id | eng | <u>**97.10**</u> | 45.20 | 96.80 | 86.00 | 46.00 |

Discussions (0)

none scraped.

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