Update README.md
Browse files
README.md
CHANGED
|
@@ -33,7 +33,9 @@ It might be better to call it an SLM.
|
|
| 33 |
|
| 34 |
Zeta 2 is an SLM that was painstakingly created on a consumer computer.
|
| 35 |
|
| 36 |
-
|
|
|
|
|
|
|
| 37 |
|
| 38 |
**Machine:** *Mac Mini (M2 Pro, 32GB RAM, 2023)*
|
| 39 |
**Issues:** *Training failed twice due to lack of RAM*
|
|
@@ -46,76 +48,16 @@ MPS is somewhat faster but doesn't allow the use of optimization systems such as
|
|
| 46 |
|
| 47 |
Zeta 2 was built using MPS.
|
| 48 |
|
| 49 |
-
|
| 50 |
- **train epochs:** 3
|
| 51 |
- **warmup steps:** 100
|
| 52 |
|
| 53 |
-
|
| 54 |
Details of the dataset used can be found [here](https://github.com/DiamondGotCat/Zeta-Dataset/releases/tag/zeta-2)
|
| 55 |
|
| 56 |
## Links
|
| 57 |
GitHub: [Zeta](https://github.com/DiamondGotCat/Zeta)
|
| 58 |
|
| 59 |
-
## Comparison with Zeta 1
|
| 60 |
-
|
| 61 |
-
````
|
| 62 |
-
(base) diamondgotcat@DGC-MacBook-Local ~ % ollama run DiamondGotCat/Zeta-1 # Zeta 1
|
| 63 |
-
>>> Hello
|
| 64 |
-
``` # Why Code block...
|
| 65 |
-
Good morning! How can I assist you today? # Morning?
|
| 66 |
-
|
| 67 |
-
>>> What your name?
|
| 68 |
-
```
|
| 69 |
-
My name is Sash. How may I be of assistance? # Not correct...
|
| 70 |
-
|
| 71 |
-
>>> What is Xcode?
|
| 72 |
-
```
|
| 73 |
-
I am sorry, but I don't understand what you're asking me to tell you about
|
| 74 |
-
Xcode. Can you please provide more context or clarify your question?
|
| 75 |
-
|
| 76 |
-
>>>
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
(base) diamondgotcat@DGC-MacBook-Local ~ % ollama run DiamondGotCat/Zeta-2:f16 # Zeta 2
|
| 80 |
-
>>> Hello
|
| 81 |
-
Hi! How can I assist you today? # Good!
|
| 82 |
-
|
| 83 |
-
>>> What your name?
|
| 84 |
-
My name is "Zeta". # Yes, You are "Zeta"
|
| 85 |
-
|
| 86 |
-
I am a conversational LLM (Large Language Model) created by DiamondGotCat.
|
| 87 |
-
|
| 88 |
-
Nice to meet you.
|
| 89 |
-
|
| 90 |
-
|
| 91 |
-
[About Zeta Project](https://github.com/DiamondGotCat/Zeta/) # This is correct URL!
|
| 92 |
-
|
| 93 |
-
>>> What is Xcode?
|
| 94 |
-
Xcode is a development environment for macOS that allows users to create,
|
| 95 |
-
edit, debug, and compile macOS app programs. It was originally developed
|
| 96 |
-
by the Mac App Store community as an alternative to Apple's Visual Studio
|
| 97 |
-
Code. Xcode has become an essential tool for developers who need to build
|
| 98 |
-
iOS, macOS, or watchOS applications using macOS.
|
| 99 |
-
|
| 100 |
-
Xcode offers a variety of tools and features such as:
|
| 101 |
-
|
| 102 |
-
* Project management: The ability to create, edit, and run your
|
| 103 |
-
application from one place.
|
| 104 |
-
* Code editing: You can write, compile, and debug your code directly in
|
| 105 |
-
the terminal, making it easier to stay on top of things like testing and
|
| 106 |
-
debugging.
|
| 107 |
-
* Integration with macOS drivers: Xcode integrates with macOS's native
|
| 108 |
-
driver, allowing you to easily connect apps to your Mac or other devices.
|
| 109 |
-
* Testing and deployment: Xcode provides a streamlined testing and
|
| 110 |
-
deployment process that ensures your app works as expected after the
|
| 111 |
-
initial build.
|
| 112 |
-
|
| 113 |
-
With Xcode, developers can quickly and easily build standalone
|
| 114 |
-
applications for iOS, macOS, and watchOS platforms.
|
| 115 |
-
|
| 116 |
-
>>> # Although some of the explanations about Xcode are incorrect, they are a big improvement over Zeta 1.
|
| 117 |
-
````
|
| 118 |
-
|
| 119 |
---
|
| 120 |
|
| 121 |
Zeta is just a small SLM.
|
|
|
|
| 33 |
|
| 34 |
Zeta 2 is an SLM that was painstakingly created on a consumer computer.
|
| 35 |
|
| 36 |
+

|
| 37 |
+
|
| 38 |
+
## Computer Spec
|
| 39 |
|
| 40 |
**Machine:** *Mac Mini (M2 Pro, 32GB RAM, 2023)*
|
| 41 |
**Issues:** *Training failed twice due to lack of RAM*
|
|
|
|
| 48 |
|
| 49 |
Zeta 2 was built using MPS.
|
| 50 |
|
| 51 |
+
## Arguments of Trainer
|
| 52 |
- **train epochs:** 3
|
| 53 |
- **warmup steps:** 100
|
| 54 |
|
| 55 |
+
## Datasets
|
| 56 |
Details of the dataset used can be found [here](https://github.com/DiamondGotCat/Zeta-Dataset/releases/tag/zeta-2)
|
| 57 |
|
| 58 |
## Links
|
| 59 |
GitHub: [Zeta](https://github.com/DiamondGotCat/Zeta)
|
| 60 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
---
|
| 62 |
|
| 63 |
Zeta is just a small SLM.
|