Parameter-efficient by design. Lemma-2.2M is best under 3M on BETU (13.78, #73/173). Lemma-1.1-250K is best under 1M on BETU (#126/173).
oscar lo
oscar128372
AI & ML interests
hello! i am a student primarily working on novel architectures and SLMs exclusively with cloud T4s.
Recent Activity
new activity about 7 hours ago
CNWPlayer/Vega-2-20M-PREVIEW:Training notes for the BETU preview row? updated a model 1 day ago
oscar128372/Lemma-250K updated a collection 1 day ago
LemmaOrganizations
models 27
oscar128372/Lemma-250K
249k • Updated • 54
oscar128372/Lemma-1.1-250K
250k • Updated • 20
oscar128372/test-250k-soup
250k • Updated • 45
oscar128372/Lemma-2.2M
2.2M • Updated • 75 • 2
oscar128372/Lemma-2.2M-init2
2.2M • Updated • 24 • 1
oscar128372/TinyChat-2M
Text Generation • Updated
oscar128372/axon-pico-3m
Text Generation • Updated • 1
oscar128372/axon-nano-6m
Text Generation • Updated • 17 • 1
oscar128372/difl-qwen2.5-3b-math
Updated
oscar128372/Axion-4B-v0.09
Text Generation • 4B • Updated • 11 • 1
datasets 15
oscar128372/SmallBase-1.1
Viewer • Updated • 240 • 34 • 1
oscar128372/SmallBase-1
Viewer • Updated • 140 • 18 • 1
oscar128372/pytorch_to_triton
Preview • Updated • 28
oscar128372/chess_spatial_reasoning_400kv1
Updated • 192 • 5
oscar128372/math
Viewer • Updated • 157 • 15
oscar128372/OoE
Viewer • Updated • 21 • 13
oscar128372/dataset
Viewer • Updated • 15 • 13
oscar128372/dpo
Viewer • Updated • 70 • 10
oscar128372/Axion-A
Viewer • Updated • 117 • 27 • 1
oscar128372/dataset2
Viewer • Updated • 71 • 23