Models
MODELHugging Face
5/5

Masakhane m2m100_418M en-nya / nya-en

Chewa(nya)Encoder-Decoder (MT)M2M-100 418MM2M100ForConditionalGeneration
View Source

Model Info

Language
Chewa (nya)
Architecture
M2M100ForConditionalGeneration
Institution
Masakhane NLP
Last Updated
3 Aug 2026

Intended Use Cases

MT (en<->nya)

License & Access

SPDX identifierCC-BY-NC-4.0

Open with restrictions — attribution required; some commercial or derivative use limited.

Provenance

Source URL
https://huggingface.co/masakhane/m2m100_418M_en_nya_rel
Fetch Timestamp
3 Aug 2026
Submitter
Masakhane NLP

Full Record

19 of 25 fields stated

Every column the registry tracks. Values marked derivedwere filled by a rule or read from the source's text rather than stated in its metadata; the rule or the quoted sentence is shown under the value.

Registry columns (17/18)

iso code
nya
Model Name
Masakhane m2m100_418M en-nya / nya-en
Where is it hosted
HuggingFace
Link
https://huggingface.co/masakhane/m2m100_418M_en_nya_rel
Year
2022
Creator/Author(s)
Masakhane NLP
Paper Reference
https://aclanthology.org/2022.wmt-1.20/
NLP Task
Encoder-Decoder (MT)
Architecture
M2M100ForConditionalGenerationderivedHugging Face config: architectures / model_type (replaces the library name 'transformers', kept in `library`)
Base Model
M2M-100 418M
Fine-tuned?
Yes
License
CC-BY-NC-4.0
Open Access
Open
Has Model Card
Yes
SOTA Tasks
MT (en<->nya)
Verified By
Not stated
Date Indexed
2026-08-03
Comment
Real, dedicated Masakhane MT models for both translation directions (English->Chichewa and Chichewa->English), fine-tuned from M2M-100. Both directions kept as …

Readiness (LRS) (2/7)

model id
masakhane-m2m100-418m-en-nya-nya-en-nya
Model Type
Not stated
Training Data Sources
MAFAND-MT / OPUS African-language parallel data
Benchmark Used
Not stated
Benchmark Score
Not stated
Benchmark Date
Not stated
Status
Not stated

Readiness Score Breakdown

5/5
Data Coverage100%
Model Availability100%
Benchmarks100%
Documentation85%
Recency100%

Representativeness

Gender Balance
Regional Distribution
Dialect Coverage
Inclusivity
Relations Explorer link

Related Models

model

mmBERT-base / mmBERT-small

Massively multilingual encoder covering 1800+ languages including Chichewa; base and small are two genuinely different model sizes, kept together as one row.…

507 languagesChewa(nya)None (trained from scratch)Encoder (MLM)
HUHugging Face
Open3 Aug 2026
model

GlotLID

Language-identification model covering 2000+ languages including Chichewa. Excluded fork: lopatnov/glotlid (mirror).

550 languages2 copiesChewa(nya)FastTextText Classification (LID)
HUHugging Face
Open3 Aug 2026
model

Masakhane m2m100_418M nya-en

Reverse-direction counterpart of masakhane/m2m100_418M_en_nya_rel; both are genuine, separately fine-tuned Masakhane project releases.

Chewa(nya)M2M-100 418MEncoder-Decoder (MT)
HUHugging Face
Open3 Aug 2026