Update README.md
Browse files
README.md
CHANGED
|
@@ -12,11 +12,14 @@ Sampled version of [cerebras/SlimPajama-627B](https://huggingface.co/datasets/ce
|
|
| 12 |
Based on the [fact that the original data was shuffled before chunking](https://huggingface.co/datasets/cerebras/SlimPajama-627B/discussions/4), I only downloaded chunk1 and further sampled 10% of the chunk.
|
| 13 |
This should result in roughly 6B tokens, hence SlimPajama-6B.
|
| 14 |
|
|
|
|
|
|
|
|
|
|
| 15 |
#### Data source proportions for SlimPajama-627B and SlimPajama-6B
|
| 16 |
For sanity check, I recaluclated the byte proportion of the sampled version, it rougly matches the original dataset.
|
| 17 |
|
| 18 |
|
| 19 |
-
| Data source | SlimPajama-627B |
|
| 20 |
| ------------- | ---------- | --------- |
|
| 21 |
| Commoncrawl | 52.2% | 54.1% |
|
| 22 |
| C4 | 26.7% | 28.7% |
|
|
|
|
| 12 |
Based on the [fact that the original data was shuffled before chunking](https://huggingface.co/datasets/cerebras/SlimPajama-627B/discussions/4), I only downloaded chunk1 and further sampled 10% of the chunk.
|
| 13 |
This should result in roughly 6B tokens, hence SlimPajama-6B.
|
| 14 |
|
| 15 |
+
The dataset is roughly 24GBs in storage size and has 5489000 rows.
|
| 16 |
+
|
| 17 |
+
---
|
| 18 |
#### Data source proportions for SlimPajama-627B and SlimPajama-6B
|
| 19 |
For sanity check, I recaluclated the byte proportion of the sampled version, it rougly matches the original dataset.
|
| 20 |
|
| 21 |
|
| 22 |
+
| Data source | SlimPajama-627B | SlimPajama-6B |
|
| 23 |
| ------------- | ---------- | --------- |
|
| 24 |
| Commoncrawl | 52.2% | 54.1% |
|
| 25 |
| C4 | 26.7% | 28.7% |
|