waxal-checkpoints

anyantudre/waxal-checkpoints

waxal-checkpoints is a automatic speech recognition model from anyantudre. Use it when you need speech turned into text. The card lists the license as apache-2.0.

PRIVATE checkpoint backup for the Google WAXAL ASR Challenge — pushed to free DICO disk quota; pull subfolders on demand. Each subfolder is a Wav2Vec2ForCTC fine-tune (char vocab) on WAXAL train (lin/sna/lug, joint, s…

Downloads · 30 days

Access

Public

Updated Aug 11, 2026

Repo size

44.4 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors39.2 GB · 88%

At a glance

Task: Automatic Speech Recognition
License: apache-2.0
Access: Public
Created: Jul 25, 2026
Updated: Aug 11, 2026
SHA: 2abd801f

Model card

From the Hugging Face model README

WAXAL Checkpoints (private backup)

PRIVATE checkpoint backup for the Google WAXAL ASR Challenge — pushed to free DICO disk quota; pull subfolders on demand. Each subfolder is a Wav2Vec2ForCTC fine-tune (char vocab) on WAXAL train (lin/sna/lug, joint, speaker-disjoint holdout).

subfolder	base	note
`w2vbert_multi`	facebook/w2v-bert-2.0	CHAMPION (best clean+robust): band 0.1775 / corrupt 0.2250
`xlsr_multi`	facebook/wav2vec2-xls-r-300m	XLS-R baseline (0.1862 / 0.2456)
`xlsr_aug`	facebook/wav2vec2-xls-r-300m	+ MCT augmentation (robustness: degradation +0.0484)
`xlsr_robust`	facebook/wav2vec2-xls-r-300m	+ full MCT
`xlsr_ext`	facebook/wav2vec2-xls-r-300m	+ FLEURS external
`xlsr_pseudo`	facebook/wav2vec2-xls-r-300m	+ 3-model pseudo-labels (student)

Scoring metric = 0.5·WER + 0.5·CER on the speaker-disjoint holdout / MUSAN+RIR corrupted band.

Pull one

from huggingface_hub import snapshot_download
snapshot_download("anyantudre/waxal-checkpoints", allow_patterns="w2vbert_multi/*", local_dir="ckpt")

Bases are Apache-2.0; trained on google/WaxalNLP (respect its license). No Phase-1 test data used.

Task: Automatic Speech Recognition
License: apache-2.0
Languages: ln, sn, lg
Created: Jul 25, 2026
Updated: Aug 11, 2026