Downloads · 30 days
15
23% of all-time downloads
Qwen2.5-1.5B-search-r1
LZYFirecn/Qwen2.5-1.5B-search-r1
Qwen2.5-1.5B-search-r1 is a reinforcement learning model from LZYFirecn. Use it for the reinforcement learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
This model is fine-tuned from Qwen2.5-1.5B using the SLIME framework with Search-R1 task.
All-time downloads
66
Public
Parameters
1.5B
67.9 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.distcp64.9 GB · 95%
At a glance
- Task
- Reinforcement Learning
- License
- apache-2.0
- Model type
- qwen2
- Access
- Public
- Created
- Jan 26, 2026
- Updated
- Jan 26, 2026
- SHA
- ed52e20b
Base models
Model card
From the Hugging Face model README
Qwen2.5-1.5B Search-R1
This model is fine-tuned from Qwen2.5-1.5B using the SLIME framework with Search-R1 task.
Training Details
- Base Model: Qwen2.5-1.5B
- Training Step: 149
- Algorithm: GRPO (Group Relative Policy Optimization)
- Task: Search-R1 (Question Answering with Search)
- Framework: SLIME + Megatron-LM
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("LZYFirecn/Qwen2.5-1.5B-search-r1")
tokenizer = AutoTokenizer.from_pretrained("LZYFirecn/Qwen2.5-1.5B-search-r1")
# Generate
inputs = tokenizer("Your question here", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0]))
- Task
- Reinforcement Learning
- Type
- qwen2
- License
- apache-2.0
- Created
- Jan 26, 2026
- Updated
- Jan 26, 2026