Viewing enriched Twitter post
PaliGemma2 for instance segmentation it blows my mind how versatile this model is! - used google/paligemma2-3b-pt-224 checkpoint - trained on A100 with 40GB VRAM - trained only text decoder colab with very dirty fine-tuning code: https://t.co/HsOGDntGOL https://t.co/7BjyDN83eg
{
"links_checked": true,
"checked_at": "2025-08-10T10:33:04.764505",
"media": [
{
"type": "photo",
"url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1866175789483614525/media_0.jpg?",
"filename": "media_0.jpg"
},
{
"type": "photo",
"url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1866175789483614525/media_1.jpg?",
"filename": "media_1.jpg"
}
],
"reprocessed_at": "2025-08-12T15:27:30.324768",
"reprocessed_reason": "missing_media_array",
"original_structure": "had_media_only"
}{
"user": {
"created_at": "2014-02-20T22:26:49.000Z",
"default_profile_image": false,
"description": "Open-source Lead @roboflow. VLMs. GPU poor. Dog person. Coffee addict. Dyslexic. | GH: https://t.co/dEmzMDGq5H | HF: https://t.co/4Lx1Yw34W7",
"fast_followers_count": 0,
"favourites_count": 14741,
"followers_count": 28957,
"friends_count": 1058,
"has_custom_timelines": true,
"is_translator": false,
"listed_count": 406,
"location": "Kraków, Polska",
"media_count": 1492,
"name": "SkalskiP",
"normal_followers_count": 28957,
"possibly_sensitive": false,
"profile_banner_url": "https://pbs.twimg.com/profile_banners/2353861069/1660153235",
"profile_image_url_https": "https://pbs.twimg.com/profile_images/1557421659925237763/RzJu2YI3_normal.jpg",
"screen_name": "skalskip92",
"statuses_count": 8002,
"translator_type": "none",
"url": "https://t.co/8HaeEMx3mZ",
"verified": true,
"withheld_in_countries": [],
"id_str": "2353861069"
},
"id": "1866175789483614525",
"conversation_id": "1866175789483614525",
"full_text": "PaliGemma2 for instance segmentation\n\nit blows my mind how versatile this model is!\n\n- used google/paligemma2-3b-pt-224 checkpoint\n- trained on A100 with 40GB VRAM\n- trained only text decoder\n\ncolab with very dirty fine-tuning code:\nhttps://t.co/HsOGDntGOL https://t.co/7BjyDN83eg",
"reply_count": 6,
"retweet_count": 46,
"favorite_count": 333,
"hashtags": [],
"symbols": [],
"user_mentions": [],
"urls": [
{
"url": "https://t.co/HsOGDntGOL",
"expanded_url": "https://colab.research.google.com/drive/112qcAXIyetyU2Wht4FrfOpj_D-JVrjdv?usp=sharing",
"display_url": "colab.research.google.com/drive/112qcAXI…"
}
],
"media": [
{
"media_url": "https://pbs.twimg.com/media/GeXzqm3XkAASRNe.jpg",
"type": "photo"
},
{
"media_url": "https://pbs.twimg.com/media/GeXzqm4WoAAsP_a.jpg",
"type": "photo"
}
],
"url": "https://twitter.com/skalskip92/status/1866175789483614525",
"created_at": "2024-12-09T17:39:14.000Z",
"#sort_index": "1866175789483614525",
"view_count": 27213,
"quote_count": 3,
"is_quote_tweet": true,
"is_retweet": false,
"is_pinned": false,
"is_truncated": false,
"quoted_tweet": {
"user": {
"created_at": "2014-02-20T22:26:49.000Z",
"default_profile_image": false,
"description": "Open-source Lead @roboflow. VLMs. GPU poor. Dog person. Coffee addict. Dyslexic. | GH: https://t.co/dEmzMDGq5H | HF: https://t.co/4Lx1Yw34W7",
"fast_followers_count": 0,
"favourites_count": 14741,
"followers_count": 28957,
"friends_count": 1058,
"has_custom_timelines": true,
"is_translator": false,
"listed_count": 406,
"location": "Kraków, Polska",
"media_count": 1492,
"name": "SkalskiP",
"normal_followers_count": 28957,
"possibly_sensitive": false,
"profile_banner_url": "https://pbs.twimg.com/profile_banners/2353861069/1660153235",
"profile_image_url_https": "https://pbs.twimg.com/profile_images/1557421659925237763/RzJu2YI3_normal.jpg",
"screen_name": "skalskip92",
"statuses_count": 8002,
"translator_type": "none",
"url": "https://t.co/8HaeEMx3mZ",
"verified": true,
"withheld_in_countries": [],
"id_str": "2353861069"
},
"id": "1865034787582251340",
"conversation_id": "1865034787582251340",
"full_text": "PaliGemma2 for image to JSON data extraction\n\n- used google/paligemma2-3b-pt-336 checkpoint; I tried to make it happen with 224, but 336 performed a lot better\n- trained on A100 with 40GB VRAM\n- trained with LoRA\n\ncolab with complete fine-tuning code: https://t.co/M1lbYXQUg6 https://t.co/DHNHGePaqM",
"reply_count": 20,
"retweet_count": 95,
"favorite_count": 749,
"hashtags": [],
"symbols": [],
"user_mentions": [],
"urls": [
{
"url": "https://t.co/M1lbYXQUg6",
"expanded_url": "https://colab.research.google.com/github/roboflow-ai/notebooks/blob/main/notebooks/how-to-finetune-paligemma2-for-json-data-extraction.ipynb",
"display_url": "colab.research.google.com/github/roboflo…"
}
],
"media": [
{
"media_url": "https://pbs.twimg.com/media/GeHrWZ9XQAASpSi.png",
"type": "photo"
},
{
"media_url": "https://pbs.twimg.com/media/GeHrXqCXQAAy4rp.jpg",
"type": "photo"
}
],
"url": "https://twitter.com/skalskip92/status/1865034787582251340",
"created_at": "2024-12-06T14:05:18.000Z",
"#sort_index": "1866175789483614500",
"view_count": 110781,
"quote_count": 3,
"is_quote_tweet": false,
"is_retweet": false,
"is_pinned": false,
"is_truncated": false
},
"startUrl": "https://x.com/skalskip92/status/1866175789483614525"
}