Downloads · 30 days
0
logan7000/mllm-mmr1-gt-internvl35-8b-mmupt-full
mllm-mmr1-gt-internvl35-8b-mmupt-full is a machine learning model from logan7000. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers.
GRPO with ground-truth rewards, MMR1-Math-RL-Data-v0, 481 steps (mmupt recipe: beta=0.01, 10 generations/prompt, cap 2048, T=0.7).
Downloads · 30 days
0
Access
Public
Updated Sep 2, 2026
Repo size
239 GB
Likes
0
Public
Click a slice to open those files.
.pt205 GB · 86%
From the Hugging Face model README
GRPO with ground-truth rewards, MMR1-Math-RL-Data-v0, 481 steps (mmupt recipe: beta=0.01, 10 generations/prompt, cap 2048, T=0.7).
| folder | step | note |
|---|---|---|
best/ | 480 | best in-loop val (eval_reward = 0.7237) |
endpoint/ | 481 | end of training |
In-loop validation rose monotonically (0.652 -> 0.688 half-mean); best fell on
the second-to-last checkpoint. training/ holds metrics, train.log, and the
Slurm joblog. Step-count fingerprint: 481 = mmupt recipe (the older recipe runs
722 steps).