Checkpoints trained with mixed RLVR from the 5.7T-token Snowball agentic SFT lineage.
Open Athena
non-profit
Verified
AI & ML interests
None defined yet.
Recent Activity
View all activity
Research artifacts from Snowball 2.7T math RL. Raw mutable-router exports may collapse; use only models marked repaired or verified frozen.
-
open-athena/Snowball-67B-A2B-Math-RL-E1-ctx8k-p8-Step5
Reinforcement Learning • 67B • Updated • 17 -
open-athena/Snowball-67B-A2B-Math-RL-E2-Unreg-Step20
Reinforcement Learning • 67B • Updated • 16 -
open-athena/Snowball-67B-A2B-Math-RL-E6-Step20-Repaired
Reinforcement Learning • 67B • Updated • 149 -
open-athena/Snowball-67B-A2B-Math-RL-E6-Repro-Step20
Reinforcement Learning • 67B • Updated • 16
Selected Snowball 67B-A2B checkpoints from mixed-task RLVR training.
-
open-athena/Snowball-67B-A2B-10T-Mixed-RLVR-Sync-Step92
67B • Updated • 1.07k -
open-athena/Snowball-67B-A2B-10T-Mixed-RLVR-Async-Step92
67B • Updated • 16 -
open-athena/Snowball-67B-A2B-10T-Mixed-RLVR2-Sync-Step116
Reinforcement Learning • 67B • Updated • 21 -
open-athena/Snowball-67B-A2B-10T-Mixed-RLVR2-Async-Step146
Reinforcement Learning • 67B • Updated • 54
Five immutable Snowball LCE base-model exports from marin-community/marin issue #8977.
-
open-athena/snowball-67b-a2b-base-262k-qk157
Text Generation • 67B • Updated • 518 -
open-athena/snowball-67b-a2b-base-262k-qk175
Text Generation • 67B • Updated • 476 -
open-athena/snowball-67b-a2b-base-262k-qk175-skew2
Text Generation • 67B • Updated • 286 -
open-athena/snowball-67b-a2b-base-262k-qk175-skew4
Text Generation • 67B • Updated • 304
Checkpoints trained with mixed RLVR from the 5.7T-token Snowball agentic SFT lineage.
Selected Snowball 67B-A2B checkpoints from mixed-task RLVR training.
-
open-athena/Snowball-67B-A2B-10T-Mixed-RLVR-Sync-Step92
67B • Updated • 1.07k -
open-athena/Snowball-67B-A2B-10T-Mixed-RLVR-Async-Step92
67B • Updated • 16 -
open-athena/Snowball-67B-A2B-10T-Mixed-RLVR2-Sync-Step116
Reinforcement Learning • 67B • Updated • 21 -
open-athena/Snowball-67B-A2B-10T-Mixed-RLVR2-Async-Step146
Reinforcement Learning • 67B • Updated • 54
Research artifacts from Snowball 2.7T math RL. Raw mutable-router exports may collapse; use only models marked repaired or verified frozen.
-
open-athena/Snowball-67B-A2B-Math-RL-E1-ctx8k-p8-Step5
Reinforcement Learning • 67B • Updated • 17 -
open-athena/Snowball-67B-A2B-Math-RL-E2-Unreg-Step20
Reinforcement Learning • 67B • Updated • 16 -
open-athena/Snowball-67B-A2B-Math-RL-E6-Step20-Repaired
Reinforcement Learning • 67B • Updated • 149 -
open-athena/Snowball-67B-A2B-Math-RL-E6-Repro-Step20
Reinforcement Learning • 67B • Updated • 16
Five immutable Snowball LCE base-model exports from marin-community/marin issue #8977.
-
open-athena/snowball-67b-a2b-base-262k-qk157
Text Generation • 67B • Updated • 518 -
open-athena/snowball-67b-a2b-base-262k-qk175
Text Generation • 67B • Updated • 476 -
open-athena/snowball-67b-a2b-base-262k-qk175-skew2
Text Generation • 67B • Updated • 286 -
open-athena/snowball-67b-a2b-base-262k-qk175-skew4
Text Generation • 67B • Updated • 304