![\"Dot](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5570735%2F8465974ecb62279e6e2cacbcbedeaeab%2F1.jpg?generation=1685113729302744&alt=media\")
Kaggle Dataset: benetech-extra-generated-data
\nI recently updated the chart generator script to include all chart types. I have uploaded 100k of each to Kaggle. Hopefully this serves as useful training data.
\nHere are some samples.
\n
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5570735%2Facaf244bc4a571f2d999c1a48faf53dd%2F0.jpg?generation=1685113763105083&alt=media\")
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5570735%2Fea4a8ff89611758bc50ddb3fa4c98ee5%2F0.jpg?generation=1685113785680400&alt=media\")
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5570735%2F2f64a79dba0bfd2db853bf1e79582e47%2F0.jpg?generation=1685113803350714&alt=media\")
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5570735%2Ffcec988a3d44e2df201c3bec50ebff3f%2F0.jpg?generation=1685113826884071&alt=media\")
Kaggle Dataset: benetech-extra-generated-data
\nI recently updated the chart generator script to include all chart types. I have uploaded 100k of each to Kaggle. Hopefully this serves as useful training data.
\nHere are some samples.
\n
Thanks for sharing, I would imagine this will be especially helpful for the dot plots!
", "rawMarkdown": "Thanks for sharing, I would imagine this will be especially helpful for the dot plots!", "votes": 3, "replies": [ { "id": 2276590, "postDate": "2023-05-27T03:42:49.103Z", "content": "I hope so :)
", "rawMarkdown": "I hope so :)" } ] }, { "id": 2279833, "postDate": "2023-05-29T17:09:32.100Z", "content": "great work!
", "rawMarkdown": "great work!", "votes": 1 }, { "id": 2277589, "postDate": "2023-05-27T22:16:41.357Z", "content": "Just updated the metadata as the <0x0A> and | markers were swapped
@brendanartley Could you please include a license so we can know whether this data can be used. If the license is not specified, all rights are reserved and so we cannot use this data for the competition. Under no license, default copyright applies.
\nDisclaimer: IANAL
", "rawMarkdown": "@brendanartley Could you please include a license so we can know whether this data can be used. **If the license is not specified, all rights are reserved and so we cannot use this data for the competition**. Under no license, default copyright applies.\n\nDisclaimer: IANAL", "votes": 2, "replies": [ { "id": 2290052, "postDate": "2023-06-06T14:12:05.793Z", "content": "Hi @asarvazyan. Yes of course! Do you know which license I should use for the Kaggle Dataset?
\nWhen try to select a license, I do not see an MIT option..
--
\nDisclaimer made me laugh haha.
", "rawMarkdown": "Hi @asarvazyan. Yes of course! Do you know which license I should use for the Kaggle Dataset?\n\nWhen try to select a license, I do not see an `MIT` option..\n\n-- \n\nDisclaimer made me laugh haha.", "votes": 2, "replies": [ { "id": 2290270, "postDate": "2023-06-06T16:55:18.350Z", "content": "Hi @brendanartley, thanks! For the competition I believe your best bet is to use GPL 3. It's basically the same as MIT but it requires that if someone modifies this data they must also open source it, which doesn't really apply.
", "rawMarkdown": "Hi @brendanartley, thanks! For the competition I believe your best bet is to use GPL 3. It's basically the same as MIT but it requires that if someone modifies this data they must also open source it, which doesn't really apply.", "votes": 1 } ] } ] }, { "id": 2280600, "postDate": "2023-05-30T07:51:18.360Z", "content": "will it improve the score?
", "rawMarkdown": "will it improve the score? ", "replies": [ { "id": 2281016, "postDate": "2023-05-30T14:16:23.743Z", "content": "Local CV classifier accuracy increased ~2% on the extracted data.
\nStill getting submission scoring error for LB though..
", "rawMarkdown": "Local CV classifier accuracy increased ~2% on the extracted data.\n\nStill getting submission scoring error for LB though..", "replies": [ { "id": 2281030, "postDate": "2023-05-30T14:29:28.687Z", "rawMarkdown": "", "isDeleted": true }, { "id": 2281048, "postDate": "2023-05-30T14:40:07.383Z", "content": "Solved: My pipeline could not handle grayscale images. Took the hint from @nbroad's donut-infer notebook.
\nI trained for 1 epoch on the 500k images in this dataset.
", "rawMarkdown": "~~Weird. My submission matches the sample submission so not really sure what the error is here..~~ Solved: My pipeline could not handle grayscale images. Took the hint from @nbroad's [donut-infer](https://www.kaggle.com/code/nbroad/donut-infer-lb-0-44-benetech) notebook.\n\nI trained for 1 epoch on the 500k images in this dataset.", "votes": 1 } ] } ] }, { "id": 2289404, "postDate": "2023-06-06T06:18:30.070Z", "rawMarkdown": "", "votes": 3, "isDeleted": true, "replies": [ { "id": 2290063, "postDate": "2023-06-06T14:18:54.703Z", "content": "Yes. Currently trying to figure out which license to specify for the dataset.
", "rawMarkdown": "Yes. Currently trying to figure out which license to specify for the dataset." } ] }, { "id": 2276611, "postDate": "2023-05-27T04:11:14.380Z", "rawMarkdown": "", "votes": 1, "isDeleted": true }, { "id": 2301259, "postDate": "2023-06-13T18:34:14.127Z", "content": "Thanks for sharing.
", "rawMarkdown": "Thanks for sharing. " } ], "comments": [ { "id": 2276581, "author_name": "Cody_Null", "author_url": "", "post_date": "2023-05-27T03:28:55.030000", "content": "Thanks for sharing, I would imagine this will be especially helpful for the dot plots!
", "votes": 3, "replies": [ { "id": 2276590, "author_name": "Bartley", "author_url": "", "post_date": "2023-05-27T03:42:49.103000", "content": "I hope so :)
", "votes": 0, "replies": [] } ] }, { "id": 2279833, "author_name": "medicine-wave", "author_url": "", "post_date": "2023-05-29T17:09:32.100000", "content": "great work!
", "votes": 1, "replies": [] }, { "id": 2277589, "author_name": "Bartley", "author_url": "", "post_date": "2023-05-27T22:16:41.357000", "content": "Just updated the metadata as the <0x0A> and | markers were swapped
@brendanartley Could you please include a license so we can know whether this data can be used. If the license is not specified, all rights are reserved and so we cannot use this data for the competition. Under no license, default copyright applies.
\nDisclaimer: IANAL
", "votes": 2, "replies": [ { "id": 2290052, "author_name": "Bartley", "author_url": "", "post_date": "2023-06-06T14:12:05.793000", "content": "Hi @asarvazyan. Yes of course! Do you know which license I should use for the Kaggle Dataset?
\nWhen try to select a license, I do not see an MIT option..
--
\nDisclaimer made me laugh haha.
", "votes": 2, "replies": [ { "id": 2290270, "author_name": "asarvazyan", "author_url": "", "post_date": "2023-06-06T16:55:18.350000", "content": "Hi @brendanartley, thanks! For the competition I believe your best bet is to use GPL 3. It's basically the same as MIT but it requires that if someone modifies this data they must also open source it, which doesn't really apply.
", "votes": 1, "replies": [] } ] } ] }, { "id": 2280600, "author_name": "HZM", "author_url": "", "post_date": "2023-05-30T07:51:18.360000", "content": "will it improve the score?
", "votes": 0, "replies": [ { "id": 2281016, "author_name": "Bartley", "author_url": "", "post_date": "2023-05-30T14:16:23.743000", "content": "Local CV classifier accuracy increased ~2% on the extracted data.
\nStill getting submission scoring error for LB though..
", "votes": 0, "replies": [ { "id": 2281030, "author_name": "", "author_url": "", "post_date": "2023-05-30T14:29:28.687000", "content": "", "votes": 0, "replies": [] }, { "id": 2281048, "author_name": "Bartley", "author_url": "", "post_date": "2023-05-30T14:40:07.383000", "content": "Solved: My pipeline could not handle grayscale images. Took the hint from @nbroad's donut-infer notebook.
\nI trained for 1 epoch on the 500k images in this dataset.
", "votes": 1, "replies": [] } ] } ] }, { "id": 2289404, "author_name": "", "author_url": "", "post_date": "2023-06-06T06:18:30.070000", "content": "", "votes": 3, "replies": [ { "id": 2290063, "author_name": "Bartley", "author_url": "", "post_date": "2023-06-06T14:18:54.703000", "content": "Yes. Currently trying to figure out which license to specify for the dataset.
", "votes": 0, "replies": [] } ] }, { "id": 2276611, "author_name": "", "author_url": "", "post_date": "2023-05-27T04:11:14.380000", "content": "", "votes": 1, "replies": [] }, { "id": 2301259, "author_name": "Bryan Arnold", "author_url": "", "post_date": "2023-06-13T18:34:14.127000", "content": "Thanks for sharing.
", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "2276195": "Kaggle Dataset: [benetech-extra-generated-data](https://www.kaggle.com/datasets/brendanartley/benetech-extra-generated-data)\n\nI recently updated the [chart generator script](https://www.kaggle.com/code/brendanartley/benetech-5-chart-types-generator) to include all chart types. I have uploaded 100k of each to Kaggle. Hopefully this serves as useful training data. \n\nHere are some samples.\n\n## Dot\n\n\n## Line\n\n\n## Scatter\n\n\n## Horizontal Bar\n\n\n## Vertical Bar\n", "2276581": "Thanks for sharing, I would imagine this will be especially helpful for the dot plots!", "2279833": "great work!", "2277589": "Just updated the metadata as the `<0x0A>` and `|` markers were swapped", "2289700": "@brendanartley Could you please include a license so we can know whether this data can be used. **If the license is not specified, all rights are reserved and so we cannot use this data for the competition**. Under no license, default copyright applies.\n\nDisclaimer: IANAL", "2280600": "will it improve the score? ", "2289404": "", "2276611": "", "2301259": "Thanks for sharing. " } }