4 ms·
Here's a link that opens their training data (52,000 rows) in Datasette Lite: https://lite.datasette.io/?json=https://github.com/tatsu-lab/stanford_alpaca/blob/
by simonw 4y ago
Here's a link that opens their training data (52,000 rows) in Datasette Lite: https://lite.datasette.io/?json=https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json https://lite.datasette.io/?json=https://github.com/tatsu-lab...
Means you can run SQL LIKE queries against it to try and get a feel for what's in there.
This link includes a query for running searches for terms: https://lite.datasette.io/?json=https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json#/data?sql=select+instruction%2C+input%2C+output+from+alpaca_data++where+instruction+%7C%7C+%27+%27+%7C%7C+input+%7C%7C+%27+%27+%7C%7C+output+like+%27%25%27+%7C%7C+%3Asearch+%7C%7C+%27%25%27+order+by+random%28%29&search=occam https://lite.datasette.io/?json=https://github.com/tatsu-lab...
- Jackson__ 4y agoConsidering the size of the training data, there's still a concerning amount of complete nonsense in this. Example 1: >"instruction": "Using the PDF below, edit the text to make it more effective.", >"input": "https://www.example.com/example.pdf https://www.example.com/example.pdf", >"output": "We recently launched a new product and wanted to(...) Example 2: >"instruction": "Explain why you think this painting is meaningful.", >"input": "https://i.pinimg.com/originals/67/2a/d1/672ad1ed7e8d71e3e7fd280eb63f263c.jpg https://i.pinimg.com/originals/67/2a/d1/672ad1ed7e8d71e3e7fd...", >"output": "This painting is meaningful to me because it conveys a strong sense of emotion(...) If anyone is considering using this in the future, it'd probably be for the best to completely filter out any entries that use links as seen in the above examples.