Two identical d10 models (100M params) trained to validate the hypothesis
that quality-filtered data enables more efficient training.
Amir Valizadeh
vitalune
AI & ML interests
None yet
Recent Activity
published
a model
17 days ago
vitalune/llama-3.2-1b-kichwa
updated
a model
17 days ago
vitalune/llama-3.2-1b-kichwa
upvoted
a
collection
27 days ago
Oren Data Distillation Experiment
Organizations
