Downloads · 30 days
0
tinker-rl-bench-scale_gsm8k_qwen3.5-4b
arvindcr4/tinker-rl-bench-scale_gsm8k_qwen3.5-4b
tinker-rl-bench-scale_gsm8k_qwen3.5-4b is a reinforcement learning model from arvindcr4. Use it for the reinforcement learning task on the model card, and read the license before you ship it in a product. It is set up for tinker. The card lists the license as mit.
GRPO experiment from TinkerRL-Bench world-class experiment suite.
Access
Public
Updated Apr 19, 2026
Repo size
—
Likes
0
Public
Repo makeup
Click a slice to open those files.
.json7.2 KB · 72%
At a glance
- Task
- Reinforcement Learning
- Library
- tinker
- License
- mit
- Access
- Public
- Created
- Apr 19, 2026
- Updated
- Apr 19, 2026
- SHA
- 0c658944
Base models
Datasets
Model card
From the Hugging Face model README
scale_gsm8k_qwen3.5-4b
GRPO experiment from TinkerRL-Bench world-class experiment suite.
Training Details
- Base model: Qwen/Qwen3.5-4B
- Method: GRPO (Group Relative Policy Optimization)
- Platform: Tinker API v0.18.1
- Task: gsm8k
- Seed: 42
- LoRA rank: 32
- Learning rate: 3e-05
- Group size: 8
- Steps: 30
Results
- First-5 avg reward: 96.2%
- Last-10 avg reward: 85.0%
- Peak reward: 100.0%
- Zero-loss steps: 57%
- Tinker Run ID: 200e4983-3e9d-5176-87d0-c55dd3c73539:train:0
Reward Trace
[
1.0,
0.8125,
1.0,
1.0,
1.0,
0.5,
0.5,
0.875,
0.625,
0.8125,
1.0,
1.0,
1.0,
0.625,
0.875,
0.25,
1.0,
0.875,
0.6875,
0.5625,
0.9375,
0.5,
0.9375,
0.5,
1.0,
1.0,
0.625,
1.0,
1.0,
1.0
]
Citation
@misc{tinker-rl-bench-2026,
title={TinkerRL-Bench: A Unified Benchmark for RL Post-Training},
author={Arvind C R and Sandhya Jeyaraj and Madhu Kumara L and Mohammad Rafi and Dhruva N Murthy and Arumugam K},
year={2026},
url={https://github.com/arvindcr4/tinker-rl-lab}
}
- Task
- Reinforcement Learning
- Library
- tinker
- License
- mit
- Created
- Apr 19, 2026
- Updated
- Apr 19, 2026