Mythologic
/

mark

Token Classification

tone-restoration

Model card Files Files and versions

393 MB

Ctrl+K

Ctrl+K

1 contributor

History: 6 commits

ainouche-abderahmane's picture

ainouche-abderahmane

Model card: unify INT8 size and parity, drop baseline wording

9ef05f3 verified about 3 hours ago

Sources
Release Mark v0.1.0: 38.7M parameter multilingual diacritics (INT8/FP16/FP32/CoreML) about 15 hours ago
coreml
Fix attention mask in exported models: full parity on Romance and sparse diacritic languages about 22 hours ago
.gitattributes

1.52 kB
initial commit about 23 hours ago
Package.swift

469 Bytes
Release Mark v0.1.0: 38.7M parameter multilingual diacritics (INT8/FP16/FP32/CoreML) about 22 hours ago
README.md

12.6 kB
Model card: unify INT8 size and parity, drop baseline wording about 3 hours ago
config.json

348 Bytes
Release Mark v0.1.0: 38.7M parameter multilingual diacritics (INT8/FP16/FP32/CoreML) about 23 hours ago
languages.json

298 Bytes
Release Mark v0.1.0: 38.7M parameter multilingual diacritics (INT8/FP16/FP32/CoreML) about 23 hours ago
mark.mlpackage.zip

36.2 MB
xet

Fix attention mask in exported models: full parity on Romance and sparse diacritic languages about 22 hours ago
mark_fp16.onnx

80.1 MB
xet

Fix attention mask in exported models: full parity on Romance and sparse diacritic languages about 22 hours ago
mark_fp32.onnx

159 MB
xet

Fix attention mask in exported models: full parity on Romance and sparse diacritic languages about 22 hours ago
mark_int8.onnx

41.8 MB
xet

Fix attention mask in exported models: full parity on Romance and sparse diacritic languages about 22 hours ago
tags.json

36.2 kB
Release Mark v0.1.0: 38.7M parameter multilingual diacritics (INT8/FP16/FP32/CoreML) about 23 hours ago