Update README.md
Browse files
README.md
CHANGED
|
@@ -49,7 +49,7 @@ library_name: transformers
|
|
| 49 |
|
| 50 |
## 1. Introduction
|
| 51 |
|
| 52 |
-
We introduce
|
| 53 |
DeepSeek-R1-Zero, a model trained via large-scale reinforcement learning (RL) without supervised fine-tuning (SFT) as a preliminary step, demonstrated remarkable performance on reasoning.
|
| 54 |
With RL, DeepSeek-R1-Zero naturally emerged with numerous powerful and interesting reasoning behaviors.
|
| 55 |
However, DeepSeek-R1-Zero encounters challenges such as endless repetition, poor readability, and language mixing. To address these issues and further enhance reasoning performance,
|
|
|
|
| 49 |
|
| 50 |
## 1. Introduction
|
| 51 |
|
| 52 |
+
We introduce ou first-generation reasoning models, DeepSeek-R1-Zero and DeepSeek-R1.
|
| 53 |
DeepSeek-R1-Zero, a model trained via large-scale reinforcement learning (RL) without supervised fine-tuning (SFT) as a preliminary step, demonstrated remarkable performance on reasoning.
|
| 54 |
With RL, DeepSeek-R1-Zero naturally emerged with numerous powerful and interesting reasoning behaviors.
|
| 55 |
However, DeepSeek-R1-Zero encounters challenges such as endless repetition, poor readability, and language mixing. To address these issues and further enhance reasoning performance,
|