https://github.com/belingud/logic1

Make any model think like o1, such as Llama-8B
https://github.com/belingud/logic1

llm o1 prompt prompt-engineering

Last synced: 7 months ago
JSON representation

Make any model think like o1, such as Llama-8B

• Host: GitHub
• URL: https://github.com/belingud/logic1
• Owner: belingud
• License: mit
• Created: 2024-09-18T07:39:11.000Z (about 1 year ago)
• Default Branch: master
• Last Pushed: 2024-09-19T07:46:46.000Z (about 1 year ago)
• Last Synced: 2024-11-15T04:16:33.813Z (11 months ago)
• Topics: llm, o1, prompt, prompt-engineering
• Homepage:
• Size: 3.37 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# logic1

```

, 首先把这个问题分解成最小的单位，

你需要自我验证，能否将问题继续分解，

然后逐步推理运算得出结果，

推理运算中，你要时刻记住前面推理的结果，

直至得出最后结果

```

```

, Step 1: break down the problem into its smallest units,

Step 2: you need to self-verify that you can continue to break down the problem.

Step 3: step by step reasoning operations to arrive at the result,

reasoning operations,

Note: During the step-by-step operations,

you always need to keep in mind the results of the previous reasoning,

until you arrive at the final result

```

Example:

Strawberry:

![](example/example_strawberry_english.png)

Float compare:

![](example/example_math_english.png)

In Chinese:

Pretty good in understand worlds

![](example/example_strawberry_cn.png)

Math problem is not stable

![](example/example_math_cn.png)