trained at two token budgets. Everything except the feed-forward block is identical across variants, so any measured difference is attributable to the FFN alone. Trained on belumind/en-vi-ja-curated-500k-triplets for Vietnamese→English and Japanese→English translation. A dense 2-layer MLP of hidden width Variants 2–4 split that same budget across four experts (h/4 each), so total parameters match the dense baseline. Variant 5 instead sizes experts at h/2 so that active parameters match the baseline, giving twice the total. (100M tokens). The token budget is identical within each group — that equality is what makes the comparison valid. 6 layers, dmodel 384, 6 heads, dff 1536, context 256…
Independent publisher
Samvaran Kashyap Rallabandi
samvaran
Models in Library1
Datasets in Library0
Models on Hugging Face5
Followers—
Compare the published evidence
This profile connects the 1 model and dataset record shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.
Published tasks: Translation.
Reported licenses: mit. Open a record to review its specific license and conditions.
Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.