Update README.md
Browse files
README.md
CHANGED
@@ -1,5 +1,52 @@
|
|
1 |
-
---
|
2 |
-
license: mit
|
3 |
-
tags:
|
4 |
-
- unsloth
|
5 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
tags:
|
4 |
+
- unsloth
|
5 |
+
- deepseek_v3
|
6 |
+
---
|
7 |
+
|
8 |
+
DeepSeek-R1 Release
|
9 |
+
__________________________________________________________________________________________
|
10 |
+
|
11 |
+
⚡ Performance on par with OpenAI-o1
|
12 |
+
|
13 |
+
📖 Fully open-source model & technical report
|
14 |
+
|
15 |
+
🏆 MIT licensed: Distill & commercialize freely!
|
16 |
+
|
17 |
+
🌐 Website & API are live now! Try DeepThink at chat.deepseek.com today!
|
18 |
+
__________________________________________________________________________________________
|
19 |
+
|
20 |
+
🔥 Bonus: Open-Source Distilled Models!
|
21 |
+
|
22 |
+
🔬 Distilled from DeepSeek-R1, 6 small models fully open-sourced
|
23 |
+
|
24 |
+
📏 32B & 70B models on par with OpenAI-o1-mini
|
25 |
+
|
26 |
+
🤝 Empowering the open-source community
|
27 |
+
|
28 |
+
🌍 Pushing the boundaries of open AI!
|
29 |
+
_____________________________________________________________________
|
30 |
+
|
31 |
+
🛠️ DeepSeek-R1: Technical Highlights
|
32 |
+
|
33 |
+
📈 Large-scale RL in post-training
|
34 |
+
|
35 |
+
🏆 Significant performance boost with minimal labeled data
|
36 |
+
|
37 |
+
🔢 Math, code, and reasoning tasks on par with OpenAI-o1
|
38 |
+
|
39 |
+
📄 More details: https://github.com/deepseek-ai/DeepSeek-R1/blob/main/DeepSeek_R1.pdf
|
40 |
+
_____________________________________________________________________
|
41 |
+
|
42 |
+
🌐 API Access & Pricing
|
43 |
+
|
44 |
+
⚙️ Use DeepSeek-R1 by setting model=deepseek-reasoner
|
45 |
+
|
46 |
+
💰 $0.14 / million input tokens (cache hit)
|
47 |
+
|
48 |
+
💰 $0.55 / million input tokens (cache miss)
|
49 |
+
|
50 |
+
💰 $2.19 / million output tokens
|
51 |
+
|
52 |
+
📖 API guide: https://api-docs.deepseek.com/guides/reasoning_model
|