issue_comments: 1652721111
This data as json
| html_url | issue_url | id | node_id | user | created_at | updated_at | author_association | body | reactions | issue | performed_via_github_app |
|---|---|---|---|---|---|---|---|---|---|---|---|
| https://github.com/simonw/datasette/issues/2109#issuecomment-1652721111 | https://api.github.com/repos/simonw/datasette/issues/2109 | 1652721111 | IC_kwDOBm6k_c5igoXX | 9599 | 2023-07-27T00:23:28Z | 2023-07-27T00:23:37Z | OWNER | Once again I'm tempted to formalize the extra HTML context as a Here's an example of Hugging Face doing that: https://github.com/huggingface/transformers/blob/1689aea73346816b936b84932e12b774974e61a6/src/transformers/training_args.py#L622C1-L624 ```python from dataclasses import dataclass, field @dataclass class TrainingArguments: output_dir: str = field( metadata={"help": "The output directory where the model predictions and checkpoints will be written."}, ) ``` |
{
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} |
1822934563 |