Viewing enriched Twitter post
Microsoft released LLM2CLIP: a CLIP model with longer context window for complex text inputs 🤯 TLDR; they replaced CLIP's text encoder with various LLMs fine-tuned on captioning, better top-k accuracy on retrieval 🔥 All models with Apache 2.0 license on @huggingface 😍 https://t.co/xvwaWmZJj1
{
"score": 0.86,
"scored_at": "2025-08-09T13:46:07.553164",
"import_source": "network_archive_import",
"links_checked": true,
"checked_at": "2025-08-10T10:32:52.019729",
"media": [
{
"type": "photo",
"url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1857090075932123628/media_0.jpg?",
"filename": "media_0.jpg"
},
{
"media_url": "https://pbs.twimg.com/media/GcW00oIXMAAOm7B.jpg",
"type": "photo"
}
],
"reprocessed_at": "2025-08-12T15:26:16.608298",
"reprocessed_reason": "missing_media_array",
"original_structure": "had_both"
}{
"user": {
"created_at": "2019-12-04T16:45:25.000Z",
"default_profile_image": false,
"description": "open-sourceress at @huggingface 🧙🏻♀️ proud Mediterrenean 🍋 I work on: zero-shot vision & VLMs, large multimodal models, transformers, smol-vision",
"fast_followers_count": 0,
"favourites_count": 48932,
"followers_count": 67974,
"friends_count": 4537,
"has_custom_timelines": true,
"is_translator": false,
"listed_count": 830,
"location": "Paris",
"media_count": 3738,
"name": "merve",
"normal_followers_count": 67974,
"possibly_sensitive": false,
"profile_banner_url": "https://pbs.twimg.com/profile_banners/1202267633049100291/1640516441",
"profile_image_url_https": "https://pbs.twimg.com/profile_images/1832391601701736448/FG6P9v2-_normal.jpg",
"screen_name": "mervenoyann",
"statuses_count": 25090,
"translator_type": "none",
"url": "https://t.co/UDKp4Ec7yK",
"verified": true,
"withheld_in_countries": [],
"id_str": "1202267633049100291"
},
"id": "1857090075932123628",
"conversation_id": "1857090075932123628",
"full_text": "Microsoft released LLM2CLIP: a CLIP model with longer context window for complex text inputs 🤯\n\nTLDR; they replaced CLIP's text encoder with various LLMs fine-tuned on captioning, better top-k accuracy on retrieval 🔥 \n\nAll models with Apache 2.0 license on @huggingface 😍 https://t.co/xvwaWmZJj1",
"reply_count": 6,
"retweet_count": 113,
"favorite_count": 674,
"hashtags": [],
"symbols": [],
"user_mentions": [
{
"id_str": "778764142412984320",
"name": "Hugging Face",
"screen_name": "huggingface",
"profile": "https://twitter.com/huggingface"
}
],
"urls": [],
"media": [
{
"media_url": "https://pbs.twimg.com/media/GcW00oIXMAAOm7B.jpg",
"type": "photo"
}
],
"url": "https://twitter.com/mervenoyann/status/1857090075932123628",
"created_at": "2024-11-14T15:55:51.000Z",
"#sort_index": "1857090075932123628",
"view_count": 35711,
"quote_count": 3,
"is_quote_tweet": false,
"is_retweet": false,
"is_pinned": false,
"is_truncated": false,
"startUrl": "https://x.com/mervenoyann/status/1857090075932123628"
}