Datasets:
Tasks:
Feature Extraction
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10K - 100K
Tags:
code
License:
Update README.md
Browse filesThis dataset aims to teach LLMs relative positioning (e.g. above, left from, below, etc.), which in my findings most LLMs, even SOTA where not able to produce under all circumstances.
Will be pushing a fine-tuned mixtral-7x8B with this dataset.
Dataset:
Contains Data for relative positioning on a grid(256, 256).
Assumes Origin [0, 0] is in the bottom left.
Two Objects (Object 1, Object 2) are randomly created.
Answer is there relative position to one another.
README.md
CHANGED
@@ -16,4 +16,14 @@ configs:
|
|
16 |
data_files:
|
17 |
- split: train
|
18 |
path: data/train-*
|
19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
data_files:
|
17 |
- split: train
|
18 |
path: data/train-*
|
19 |
+
license: apache-2.0
|
20 |
+
task_categories:
|
21 |
+
- feature-extraction
|
22 |
+
language:
|
23 |
+
- en
|
24 |
+
tags:
|
25 |
+
- code
|
26 |
+
pretty_name: relative-positioning
|
27 |
+
size_categories:
|
28 |
+
- 10K<n<100K
|
29 |
+
---
|