Datasets:
Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: odc-by
|
3 |
+
task_categories:
|
4 |
+
- text-generation
|
5 |
+
language:
|
6 |
+
- en
|
7 |
+
tags:
|
8 |
+
- language-modeling
|
9 |
+
- casual-lm
|
10 |
+
- llm
|
11 |
+
pretty_name: Dolma
|
12 |
+
size_categories:
|
13 |
+
- 100B<n<1T
|
14 |
+
---
|
15 |
+
|
16 |
+
Tokenized (Llama 2) verison of [NousResearch/dolma-v1_7-30B](https://huggingface.co/datasets/NousResearch/dolma-v1_7-30B) as a [Nanotron](https://github.com/huggingface/nanotron) dataset split into 10 GB chunks.
|
17 |
+
|
18 |
+
To download:
|
19 |
+
```shell
|
20 |
+
huggingface-cli download --repo-type dataset --local-dir dolma-v1_7-30B-tokenized-llama2-nanoset --local-dir-use-symlinks False NousResearch/dolma-v1_7-30B-tokenized-llama2-nanoset
|
21 |
+
```
|
22 |
+
|
23 |
+
To recombine:
|
24 |
+
```shell
|
25 |
+
cat dolma-v1_7-30B-tokenized-llama2-nanoset_input_ids.npy.* > dolma-v1_7-30B-tokenized-llama2-nanoset_input_ids.npy
|
26 |
+
rm -rf dolma-v1_7-305B-tokenized-llama3-nanoset
|
27 |
+
```
|
28 |
+
|
29 |
+
Can also be used directly with numpy, for example
|
30 |
+
```python
|
31 |
+
import numpy as np
|
32 |
+
|
33 |
+
dataset_buffer_mmap = np.memmap("dolma-v1_7-30B-tokenized-llama2-nanoset_input_ids.npy",
|
34 |
+
mode="r", order="C", dtype=np.int32)
|
35 |
+
dataset_buffer = memoryview(dataset_buffer_mmap)
|
36 |
+
dataset_number_of_tokens = int(len(dataset_buffer))
|
37 |
+
```
|