Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model:
|
| 3 |
+
- Qwen/Qwen2.5-3B-Instruct
|
| 4 |
+
datasets:
|
| 5 |
+
- PRIME-RL/Eurus-2-RL-Data
|
| 6 |
+
language:
|
| 7 |
+
- en
|
| 8 |
+
pipeline_tag: text-generation
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# q1-3B-PRIME
|
| 12 |
+
|
| 13 |
+
**q1-3B-PRIME**, a small reasoning model trained with reinforcement learning.
|
| 14 |
+
|
| 15 |
+
# Benchmark Performance
|
| 16 |
+
|
| 17 |
+
Math
|
| 18 |
+
|
| 19 |
+
| Model | AIME24 | AMC23 | MATH-500 |
|
| 20 |
+
|---------|--------|-------|-------|
|
| 21 |
+
| Qwen2.5-3B-Instruct | 6.67 | 45 | - |
|
| 22 |
+
| **q1-3B-PRIME** | **26.667** | **65** | - |
|
| 23 |
+
| SmallThinker-3B-Preview| 16.667 | 57.5 | - |
|
| 24 |
+
| GPT-4o | 9.3 | 45.8 | 76.4 |
|
| 25 |
+
|
| 26 |
+
Coding
|
| 27 |
+
|
| 28 |
+
| Model | HumanEval | Leetcode |
|
| 29 |
+
|---------|--------|-------|
|
| 30 |
+
| Qwen2.5-3B-Instruct | **74.4** | - |
|
| 31 |
+
| **q1-3B-PRIME** | 71.95 | **20.55** |
|
| 32 |
+
| GPT-4o | 90.2 | - |
|