Downloads · 30 days
0
JWei05/DAPO-gemma4-26b-a4b-PT-DeepScaleR-gemma26b-hard-seed42-teacher-step180
DAPO-gemma4-26b-a4b-PT-DeepScaleR-gemma26b-hard-seed42-teacher-step180 is a machine learning model from JWei05. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as gemma.
Weights-only HF export of globalstep180/actor/huggingface from the run DAPO-gemma4-26b-a4b-pt-DeepScaleR-gemma26b-hard-seed42 (W&B rl-distill/DAPO, id g4ds26b-26b-a4b-hard-s42-v1). Best in-distribution validation so f…
Downloads · 30 days
0
Access
Public
Updated Sep 4, 2026
Repo size
53.1 GB
Likes
0
Public
Click a slice to open those files.
.safetensors53.1 GB · 100%
From the Hugging Face model README
Weights-only HF export of global_step_180/actor/huggingface from the run
DAPO-gemma4-26b-a4b-pt-DeepScaleR-gemma26b-hard-seed42 (W&B rl-distill/DAPO, id g4ds26b-26b-a4b-hard-s42-v1).
Best in-distribution validation so far: mean@16 0.4329 at step 180 (300 hard questions x16, val-core/math/acc/mean@16).
This repo is frozen (never pruned or squashed) so it can be pinned as a distillation teacher; the rolling
run repo JWei05/DAPO-gemma4-26b-a4b-PT-DeepScaleR-gemma26b-hard-seed42 keeps only its newest 3 snapshots.
Layout matches the other teacher exports: step_000180/ holds the model.