Commit
•
a139750
1
Parent(s):
7a19e2b
Update README.md
Browse files
README.md
CHANGED
|
@@ -116,8 +116,10 @@ In the course of this study, the [SlimOrca](https://huggingface.co/datasets/Open
|
|
| 116 |
|
| 117 |
|
| 118 |
Subsequently, two distinct subsets were crafted, comprising 102,000 and 1,000 samples, denoted as:
|
|
|
|
| 119 |
- [prince-canuma/SmallOrca](https://huggingface.co/datasets/prince-canuma/SmallOrca)
|
| 120 |
- [prince-canuma/TinyOrca](https://huggingface.co/datasets/prince-canuma/TinyOrca)
|
|
|
|
| 121 |
Although experimentation was conducted with both datasets, optimal results were achieved through fine-tuning on a modest set of 200 samples.
|
| 122 |
Notably, the investigation revealed that augmenting the training data beyond this threshold predominantly enhanced the model's proficiency in generating Chain-of-Thought responses.
|
| 123 |
However, it is imperative to note that the preference for Chain-of-Thought responses may not be universally applicable. Particularly in scenarios like the RAG setup,
|
|
|
|
| 116 |
|
| 117 |
|
| 118 |
Subsequently, two distinct subsets were crafted, comprising 102,000 and 1,000 samples, denoted as:
|
| 119 |
+
|
| 120 |
- [prince-canuma/SmallOrca](https://huggingface.co/datasets/prince-canuma/SmallOrca)
|
| 121 |
- [prince-canuma/TinyOrca](https://huggingface.co/datasets/prince-canuma/TinyOrca)
|
| 122 |
+
|
| 123 |
Although experimentation was conducted with both datasets, optimal results were achieved through fine-tuning on a modest set of 200 samples.
|
| 124 |
Notably, the investigation revealed that augmenting the training data beyond this threshold predominantly enhanced the model's proficiency in generating Chain-of-Thought responses.
|
| 125 |
However, it is imperative to note that the preference for Chain-of-Thought responses may not be universally applicable. Particularly in scenarios like the RAG setup,
|