Downloads · 30 days
6
11% of all-time downloads
qstoreprice-sft
nandeshjeyalakshmi/qstoreprice-sft
qstoreprice-sft is a reinforcement learning model from nandeshjeyalakshmi. Use it for the reinforcement learning task on the model card, and read the license before you ship it in a product.
Base model: Qwen/Qwen2.5-1.5B-Instruct Training: SFT warm-start (5 epochs) + GRPO rollouts (3 episodes) Metric: WRR (Weekly Waste Recovery Rate)
All-time downloads
56
Public
Parameters
1.5B
3.2 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors3.2 GB · 98%
At a glance
- Task
- Reinforcement Learning
- Model type
- qwen2
- Access
- Public
- Created
- Apr 25, 2026
- Updated
- Apr 25, 2026
- SHA
- 7865b5f8
Base models
Model card
From the Hugging Face model README
QStorePrice AI — Trained Checkpoint
Base model: Qwen/Qwen2.5-1.5B-Instruct
Training: SFT warm-start (5 epochs) + GRPO rollouts (3 episodes)
Metric: WRR (Weekly Waste Recovery Rate)
Evaluation Results
{
"STABLE_WEEK": {
"wrr_mean": 2.1512,
"wrr_std": 0.3669,
"wrr_min": 1.8917,
"wrr_max": 2.4106,
"quality": 0.7933,
"violations_mean": 0.5,
"constitutional_pass_rate": "2/2"
},
"CRISIS_WEEK": {
"wrr_mean": 2.2058,
"wrr_std": 0.1102,
"wrr_min": 2.1279,
"wrr_max": 2.2838,
"quality": 0.7639,
"violations_mean": 5.0,
"constitutional_pass_rate": "0/2"
}
}
Project
- GitHub: https://github.com/nandeshkanagaraju/QStorePrice
- Trained on Kaggle with Unsloth 4-bit + LoRA
- Task
- Reinforcement Learning
- Type
- qwen2
- Created
- Apr 25, 2026
- Updated
- Apr 25, 2026