File size: 1,534 Bytes
5b800f1
 
 
 
9a0afd2
 
5b800f1
 
ff99345
 
 
 
cfd497b
5b800f1
 
 
 
a362924
 
d7898ad
a362924
 
ff99345
5b800f1
 
49acb3d
5b800f1
 
 
 
 
 
86967a1
 
 
5b800f1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
---
license: apache-2.0
language:
- en
datasets:
- appvoid/no-prompt-15k
pipeline_tag: text-generation
---
![palmer](https://huggingface.co/appvoid/no-prompt-1.3b/resolve/main/_ccd1a5dd-2ddc-4d5a-8163-fd6d1b39f5f4.jpeg?download=true)
# no-prompt
### a sheared-llama-1.3b fine-tuning
This model uses an 1.3 billion parameters model as base to be further fine-tuned on the same data as palmer. It works pretty good and even surpasses sota model on `hellaswag`.

### evaluation
|Model|	ARC_C|	HellaSwag|	PIQA|	Winogrande|
|------|-----|-----------|------|-------------|
|tinyllama-2t| 0.2807| 0.5463| 0.7067| 0.5683|
|palmer-001  | 0.2807| 0.5524| 0.7106| 0.5896|
|sheared-1.3b| 0.2910| 0.5935| 0.7339| 0.5809|
|no-prompt-1.3b| 0.3157| **0.6022**| 0.7334| 0.5864|
|falcon-rw-1b-instruct-openorca (sota) | **0.3362**|   0.5997|  **0.7394**|  **0.6148**|

This model was trained on less than 25% of the dataset yet achieves competitive performance to current sota on open llm leaderboard.

### training
Training took ~5 P100 gpu hours. It was trained on 15,000 gpt-4 shuffled samples. no-prompt was fine-tuned using lower learning rates ensuring it keeps as much general knowledge as possible.

### prompt
```
no prompt
```

### limitations
Hallucinations are frequent, just as any transformer model this size.

<a href="https://ko-fi.com/appvoid" target="_blank"><img src="https://cdn.buymeacoffee.com/buttons/v2/default-yellow.png" alt="Buy Me A Coffee" style="height: 48px !important;width: 180px !important; filter: invert(70%);" ></a>