Bronya-Rand
commited on
Commit
•
4556835
1
Parent(s):
a653b20
Update README.md
Browse files
README.md
CHANGED
@@ -3,8 +3,8 @@ license: apache-2.0
|
|
3 |
---
|
4 |
|
5 |
<p align="center">
|
6 |
-
|
7 |
-
<p align="center" style="font-size: 26px"><b>
|
8 |
<p align="center" style="font-size: 14px">Model Size: 7B</p>
|
9 |
</p>
|
10 |
|
@@ -14,17 +14,15 @@ license: apache-2.0
|
|
14 |
|
15 |
<p align="center" style="font-size: 20px">A <b>experimental</b> (and beginner) model merge using Intel's Neural Chat 7B</p>
|
16 |
|
17 |
-
> The Intel logo here is a parody of the real Intel logo with a bit of AI focus in mind. I am no way associated with [Intel](https://huggingface.co/Intel) Thought it be cute to add. Intel, Intel Core, and the Intel Core logo are registered trademarks of Intel Corporation.
|
18 |
-
|
19 |
## Model Details
|
20 |
|
21 |
Trained on: **Intel Xeon E5-2693v3 | NVIDIA RTX 2080 Ti | 128 GB DDR4 *(yes I'm poor :( )***
|
22 |
|
23 |
-
The
|
24 |
|
25 |
-
The
|
26 |
|
27 |
-
The models used to create the
|
28 |
|
29 |
- Intel's Neural Chat 7B V3.1 ([Intel/neural-chat-7b-v3-1](https://huggingface.co/Intel/neural-chat-7b-v3-1))
|
30 |
- Teknium's Airoboros Mistral 2.2 7B ([teknium/airoboros-mistral2.2-7b](https://huggingface.co/teknium/airoboros-mistral2.2-7b))
|
@@ -32,7 +30,7 @@ The models used to create the Intel Core i7-1000AI are as follows:
|
|
32 |
|
33 |
## Prompt
|
34 |
|
35 |
-
The
|
36 |
|
37 |
### For Intel
|
38 |
```
|
@@ -49,18 +47,15 @@ USER: <prompt>
|
|
49 |
ASSISTANT:
|
50 |
```
|
51 |
|
52 |
-
## Why brand it Intel Core iX and not some normal name?
|
53 |
-
I wanted to be different + just add a bit of a thanks to Intel using something they can recognize so people know it's based off their work.
|
54 |
-
|
55 |
## Benchmarks?
|
56 |
I have no idea how to do them. You are welcome to make your own.
|
57 |
|
58 |
## Ethical Considerations and Limitations
|
59 |
-
The intended use-case for the
|
60 |
|
61 |
-
The
|
62 |
|
63 |
-
The
|
64 |
|
65 |
### Disclaimer
|
66 |
The license on this model does not constitute legal advice. I am not responsible for the actions of third parties (services/users/etc.) who use this model and distribute it for others. Please cosult an attorney before using this model for commercial purposes.
|
|
|
3 |
---
|
4 |
|
5 |
<p align="center">
|
6 |
+
<!-- <img src="./assets/Core1000AIIMG.png"/> -->
|
7 |
+
<p align="center" style="font-size: 26px"><b>Inairtra-7B</b></p>
|
8 |
<p align="center" style="font-size: 14px">Model Size: 7B</p>
|
9 |
</p>
|
10 |
|
|
|
14 |
|
15 |
<p align="center" style="font-size: 20px">A <b>experimental</b> (and beginner) model merge using Intel's Neural Chat 7B</p>
|
16 |
|
|
|
|
|
17 |
## Model Details
|
18 |
|
19 |
Trained on: **Intel Xeon E5-2693v3 | NVIDIA RTX 2080 Ti | 128 GB DDR4 *(yes I'm poor :( )***
|
20 |
|
21 |
+
The Inairtra-7B LLM is a LLM made by Bronya Rand (bronya_rand / Bronya-Rand) as a beginning learning model to merging models using [MergeKit](https://github.com/cg123/mergekit) and GGUF quantization. This model is based off Intel's [Neural Chat 7B V3.1](https://huggingface.co/Intel/neural-chat-7b-v3-1) as the base model along with three additional Mistral models.
|
22 |
|
23 |
+
The Inairtra-7B architecture is based off: [**Mistral**](https://huggingface.co/mistralai/Mistral-7B-v0.1)
|
24 |
|
25 |
+
The models used to create the Inairtra-7B are as follows:
|
26 |
|
27 |
- Intel's Neural Chat 7B V3.1 ([Intel/neural-chat-7b-v3-1](https://huggingface.co/Intel/neural-chat-7b-v3-1))
|
28 |
- Teknium's Airoboros Mistral 2.2 7B ([teknium/airoboros-mistral2.2-7b](https://huggingface.co/teknium/airoboros-mistral2.2-7b))
|
|
|
30 |
|
31 |
## Prompt
|
32 |
|
33 |
+
The Inairtra-7B *should* (but unsure) support the same prompts as featured in Intel's Neural Chat, Airoboros Mistral and Synatra.
|
34 |
|
35 |
### For Intel
|
36 |
```
|
|
|
47 |
ASSISTANT:
|
48 |
```
|
49 |
|
|
|
|
|
|
|
50 |
## Benchmarks?
|
51 |
I have no idea how to do them. You are welcome to make your own.
|
52 |
|
53 |
## Ethical Considerations and Limitations
|
54 |
+
The intended use-case for the Inairtra-7B LLM is for fictional writing/roleplay solely for personal entertainment purposes. Any other sort of usage outside of this is out of scope of my intentions and the LLM itself.
|
55 |
|
56 |
+
The Inairtra-7B LLM has been merged with models which are uncensored/unfiltered. The LLM can produce content, including but not limited to, content that may be NSFW for those under the age of eighteen, content that may be illegal in certain states/countries, offensive content, etc.
|
57 |
|
58 |
+
The Inairtra-7B LLM is not designed to produce the most accurate information. It may produce incorrect data like all other AI models.
|
59 |
|
60 |
### Disclaimer
|
61 |
The license on this model does not constitute legal advice. I am not responsible for the actions of third parties (services/users/etc.) who use this model and distribute it for others. Please cosult an attorney before using this model for commercial purposes.
|