Update README.md
Browse files
README.md
CHANGED
|
@@ -1,16 +1,16 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
language:
|
| 4 |
-
- en
|
| 5 |
-
- zh
|
| 6 |
-
library_name: transformers
|
| 7 |
-
pipeline_tag: text-generation
|
| 8 |
-
tags:
|
| 9 |
-
- llm
|
| 10 |
-
- nanbeige
|
| 11 |
-
base_model:
|
| 12 |
-
- Nanbeige/Nanbeige4-3B-Base
|
| 13 |
-
---
|
| 14 |
<div align="center">
|
| 15 |
|
| 16 |
<img src="figures/nbg.png" width="220" alt="Nanbeige Logo">
|
|
@@ -24,22 +24,17 @@ base_model:
|
|
| 24 |
|
| 25 |
# Introduction
|
| 26 |
Nanbeige4-3B-Thinking-2511 is an enhanced iteration over our previous Nanbeige4-3B-Thinking-2510.
|
| 27 |
-
Through advanced distillation techniques and reinforcement learning (RL) optimization, we have
|
| 28 |
-
|
| 29 |
-
Besides, Nanbeige4-3B-Thinking-2511 achieves state-of-the-art (SOTA) results among models smaller than 32B parameters on general tasks like Arena-Hard-V2 and BFCL-V4.
|
| 30 |
-
This marks a major milestone in delivering powerful, efficient reasoning performance at a compact scale.
|
| 31 |
|
| 32 |
* Technical Report: https://arxiv.org/pdf/2512.06266
|
| 33 |
|
| 34 |
<div align="center">
|
| 35 |
|
| 36 |
-
<img src="figures/
|
| 37 |
</div>
|
| 38 |
|
| 39 |
-
<div align="center">
|
| 40 |
|
| 41 |
-
<img src="figures/performance_2511.png">
|
| 42 |
-
</div>
|
| 43 |
|
| 44 |
|
| 45 |
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
- zh
|
| 6 |
+
library_name: transformers
|
| 7 |
+
pipeline_tag: text-generation
|
| 8 |
+
tags:
|
| 9 |
+
- llm
|
| 10 |
+
- nanbeige
|
| 11 |
+
base_model:
|
| 12 |
+
- Nanbeige/Nanbeige4-3B-Base
|
| 13 |
+
---
|
| 14 |
<div align="center">
|
| 15 |
|
| 16 |
<img src="figures/nbg.png" width="220" alt="Nanbeige Logo">
|
|
|
|
| 24 |
|
| 25 |
# Introduction
|
| 26 |
Nanbeige4-3B-Thinking-2511 is an enhanced iteration over our previous Nanbeige4-3B-Thinking-2510.
|
| 27 |
+
Through advanced knowledge distillation techniques and targeted reinforcement learning (RL) optimization, we have significantly scaled the model’s reasoning capabilities, delivering stronger and more reliable performance on diverse challenging benchmarks.
|
| 28 |
+
This version establishes new state-of-the-art (SOTA) results among open models under 32B parameters on AIME, GPQA-Diamond, Arena-Hard-V2, and BFCL-V4, which marks a major milestone in delivering powerful yet efficient reasoning capabilities at a compact scale.
|
|
|
|
|
|
|
| 29 |
|
| 30 |
* Technical Report: https://arxiv.org/pdf/2512.06266
|
| 31 |
|
| 32 |
<div align="center">
|
| 33 |
|
| 34 |
+
<img src="figures/nbg_performance.png">
|
| 35 |
</div>
|
| 36 |
|
|
|
|
| 37 |
|
|
|
|
|
|
|
| 38 |
|
| 39 |
|
| 40 |
|