Qwen2.5-Coder Series: Powerful, Diverse, Practical.
Qwen2.5-Coder Series: Powerful, Diverse, Practical. GITHUB HUGGING FACE MODELSCOPE KAGGLE DEMO DISCORD Introduction#Today, we are excited to open source the “Powerfulâ€, “Diverseâ€, and “Practical†Qwen2.5-Coder series, dedicated to continuously promoting the development of Open CodeLLMs. Powerful: Qwen2.5-Coder-32B-Instruct has become the current SOTA open-source code model, matching the coding cap
ドã‚ュメントファイルã®è§£æžã¨å¤‰æ›ã«ç‰¹åŒ–ã—ãŸã‚ªãƒ¼ãƒ—ンソースツール「doclingã€ã‚’試ã—ã¦ã¿ãŸ | DevelopersIO
AWS事æ¥æœ¬éƒ¨ã‚³ãƒ³ã‚µãƒ«ãƒ†ã‚£ãƒ³ã‚°éƒ¨ã®çŸ³å·ã§ã™ã€‚最近ã€æ–‡æ›¸ãƒ•ã‚¡ã‚¤ãƒ«ã®è§£æžã¨å¤‰æ›ã«ç‰¹åŒ–ã—㟠オープンソースツール「docling〠ãŒæ³¨ç›®ã‚’集ã‚ã¦ã„ã¾ã™ã€‚ã“ã®ãƒ„ールã¯ã€æœ€å…ˆç«¯ã®AIモデルを活用ã—ã¦ã€ãƒ‰ã‚ュメントã®æ§‹é€ ã‚’ç†è§£ã—ã€æ§˜ã€…ãªå½¢å¼ã«å¤‰æ›ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚今回ã€å®Ÿéš›ã«doclingを使用ã—ã¦æ—¥æœ¬èªžã®ãƒ‰ã‚ュメントファイルã‹ã‚‰Markdownファイルã¨JSONファイルã¸å¤‰æ›ã‚’試ã—ã¦ã¿ã¾ã™ã€‚ doclingã¨ã¯ doclingã¯ã€ãƒ‰ã‚ュメントã®å¤‰æ›ã¨è§£æžã«ç‰¹åŒ–ã—ãŸã‚ªãƒ¼ãƒ—ンソースã®Pythonパッケージã§ã™ã€‚最先端ã®äººå·¥çŸ¥èƒ½ãƒ¢ãƒ‡ãƒ«ã‚’活用ã—ã€ãƒ¬ã‚¤ã‚¢ã‚¦ãƒˆè§£æžã«DocLayNetモデルã€è¡¨æ§‹é€ èªè˜ã«TableFormerモデルを使用ã—ã¦ã„ã¾ã™ã€‚一般的ãªãƒãƒ¼ãƒ‰ã‚¦ã‚§ã‚¢ã§åŠ¹çŽ‡çš„ã«å‹•ä½œã—ã€å°‘ãªã„リソースã§é«˜æ€§èƒ½ã‚’発æ®ã—ã¾ã™ã€‚JSONã‚„Markdownå½¢å¼ã¸ã®å¤‰æ›ãŒå¯èƒ½ã§ã€æ‹¡å¼µæ€§ã‚‚高ãã€æ–°æ©Ÿèƒ½ã‚„モデルã®è¿½åŠ ãŒå®¹æ˜“
Llama 3.1 Swallow – Swallow LLM
Llama 3.1 Swallow Llama 3.1 Swallowã¯Llama 3.1ã®è‹±èªžã®èƒ½åŠ›ã‚’ç¶æŒã—ãªãŒã‚‰ã€æ—¥æœ¬èªžã®èƒ½åŠ›ã‚’強化ã—ãŸå¤§è¦æ¨¡è¨€èªžãƒ¢ãƒ‡ãƒ« (8B, 70B) ã§ã™ã€‚モデルã®ãƒ‘ラメータ（é‡ã¿ï¼‰ãŒHuggingFace上ã§å…¬é–‹ã•ã‚Œã¦ã„ã¾ã™ã®ã§ã€Llama 3.1ライセンスã«å¾“ã†é™ã‚Šã€ç ”究や商æ¥ç›®çš„ãªã©ã§åˆ©ç”¨ã§ãã¾ã™ï¼ˆãŸã ã—ã€æŒ‡ç¤ºãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°æ¸ˆã¿ãƒ¢ãƒ‡ãƒ«ã®åˆ©ç”¨ã«ãŠã„ã¦ã¯ã€Llama 3.1ライセンスã«å¾“ã£ãŸã†ãˆã§ã€Gemma利用è¦ç´„ã®åˆ©ç”¨åˆ¶é™ã«æŠµè§¦ã—ã¦ã¯ã„ã‘ã¾ã›ã‚“）。Llama 3.1 Swallowã¯ç±³Meta社ã®Llama 3.1をベースã«ã€æ±äº¬ç§‘å¦å¤§å¦æƒ…å ±ç†å·¥å¦é™¢ã®å²¡å´Žç ”究室ã¨æ¨ªç”°ç ”究室ã€å›½ç«‹ç ”究開発法人産æ¥æŠ€è¡“ç·åˆç ”究所ã®ç ”究ãƒãƒ¼ãƒ ã§é–‹ç™ºã•ã‚Œã¾ã—ãŸã€‚Built with Llama. View on HuggingFace Hugging Face Llama 3.1
GitHub - cvpaperchallenge/Crux: Crux is a suite of LLM-empowered summarization and retrieval services for academic activity. Crux is developed by XCCV group of cvpaper.challenge.
Crux (Accelerator of SCiENtific DEvelopment and Research) is a GitHub repository template for research projects using Python as a developing language. The following features are pre-implemented to accelerate your development: Container: Use of Docker reduces development environment dependencies and improves code portability. Virtual environment / package management: Package management using Poetry
AIアライメントï¼AIセーフティã®4ã¤ã®å•é¡Œé ˜åŸŸï½œR. Maruyama
「AIアライメントã€ã‚„「AIセーフティã€ã«é–¢ã™ã‚‹ä¼šè©±ã§ã€ã¾ãšæœ€åˆã«å¿…ãšèº“ãã®ãŒã€ã“れらã®è¨€è‘‰ã®æ„味ã 。「AI Alignmentã€ã€ŒAI Safetyã€ã¨èžã„ã¦ä½•ã‚’イメージã™ã‚‹ã‹ã«ã¯ã€äººã«ã‚ˆã£ã¦ã‚‚ã®ã™ã”ãã°ã‚‰ã¤ããŒã‚る。相手ã®é ã«ã€ã©ã®ã‚ˆã†ãªç¯„囲ã®å•é¡ŒãŒæµ®ã‹ã‚“ã§ã„ã‚‹ã‹ã€ãれを探るã¨ã“ã‚ã‹ã‚‰ä¼šè©±ã‚’始ã‚ãªãã¦ã¯ã„ã‘ãªã„。 実際ã€ã“れらã®è¨€è‘‰ã¯ä½•ã‚’æ„味ã—ã¦ã„ã‚‹ã®ã ã‚ã†ã‹ã€‚ã“れらã®ç”¨èªžãƒ»ç”¨ä¾‹ã®å‡ºè‡ªã«ã•ã‹ã®ã¼ã£ã¦ã‚‚ã€ã‚ã¾ã‚Šå½¹ã«ç«‹ãŸãªã„ã‹ã‚‚ã—ã‚Œãªã„。ã¨ã„ã†ã®ã‚‚ã€ç¨®ã€…ã®æ€æƒ‘ã€ãƒã‚¸ã‚·ãƒ§ãƒ³å–ã‚Šã®ãŸã‚ã«ã€ã“れらã®è¨€è‘‰ã®æ„味ã¯ãã®éƒ½åº¦å¤‰ã‚ã£ã¦ããŸã‹ã‚‰ã （ãã®ã‚ãŸã‚Šã®äº‹æƒ…ã®ä¸€ç«¯ã¯ã€Center for AI Safety所長ã®Dan Hendrycksã•ã‚“ãŒå…ˆæ—¥ã®Webinarã§è©±ã—ã¦ãã‚Œã¦ã„る）。 ãã“ã§æœ¬æŠ•ç¨¿ã§ã¯ã€è¨€è‘‰ã®æ´å²ã‚’ã„ã£ãŸã‚“脇ã«ç½®ãã€ä»Šç¾åœ¨ã®è°è«–ã«é›†ä¸ã—ãŸã„。ãã®ã†ãˆã§ã€ã€ŒAIアライメントï¼AIセー
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
Densing Law of LLMs
Developing a 172B LLM with Strong Japanese Capabilities Using NVIDIA Megatron-LM | NVIDIA Technical Blog
Qwen2.5-Coder Technical Report
Scaling Laws for Pre-training Agents and World Models
GitHub - iterative/datachain: AI-data warehouse to enrich, transform and analyze unstructured data
20240730_LLM-jp_model-WG_public.pdf
Deploy AI agent with tool use - a Lightning Studio by aniket
Qwen2.5-Coder Technical Report
Pytorch Conference
GitHub - Lightning-AI/LitServe: Lightning-fast serving engine for AI models. Flexible. Easy. Enterprise-scale.
Agents in Software Engineering: Survey, Landscape, and Vision
https://scontent-nrt1-1.xx.fbcdn.net/v/t39.2365-6/453304228_1160109801904614_7143520450792086005_n.pdf?_nc_cat=108&ccb=1-7&_nc_sid=3c67a6&_nc_ohc=jkAAsVocD04Q7kNvgEBNLVS&_nc_ht=scontent-nrt1-1.xx&oh=00_AYCpoJKGUZv1vP3YNkaIKRyCNbgEqa-cxhnPA8tofNUgog&oe=66E20D87
GitHub - huggingface/nanotron: Minimalistic large language model 3D-parallelism training
Training Deep Neural Networks with Joint Quantization and Pruning of Weights and Activations
Quantizable Transformers: Removing Outliers by Helping Attention Heads Do Nothing
{{#tags}}- {{label}}
{{/tags}}