Data Science for Everyone

3). Scaling Transformer to 1 Billion Tokens - presents LongNet, a Transformer variant that can scale sequence length to more than 1 billion tokens, with no loss in shorter sequences. (paper | tweet)

2). CodeGen2.5 - releases a new code LLM trained on 1.5T tokens; the 7B model is on par with >15B code-generation models and it’s optimized for fast sampling. (paper | tweet)

1). LLMs as Effective Text Rankers - proposes a prompting technique that enables open-source LLMs to perform state-of-the-art text ranking on standard benchmarks. (paper | tweet)

https://www.linkedin.com/posts/behroozazarkhalili_pdf-file-of-the-jupyter-notebook-activity-7083807011800326144-eLf0?utm_source=share&utm_medium=member_android این یک چت‌بات هست که اخیراً نوشتم و می‌تونه به خیلی از سوالات شما در ماشین‌لرنینگ و دیپ‌لرنینگ جواب بده... البته باید openai API key داشته‌ باشید...🙏🌺

3). AudioPaLM - fuses text-based and speech-based LMs, PaLM-2 and AudioLM, into a multimodal architecture that supports speech understanding and generation; outperforms existing systems for speech translation tasks with zero-shot speech-to-text translation capabilities. (paper | tweet)

2). LMFlow - an extensible and lightweight toolkit that simplifies finetuning and inference of general large foundation models; supports continuous pretraining, instruction tuning, parameter-efficient finetuning, alignment tuning, and large model inference. (paper | tweet)

1). Textbooks Are All You Need - introduces a new 1.3B parameter LLM called phi-1; it’s significantly smaller in size and trained for 4 days using a selection of textbook-quality data and synthetic textbooks and exercises with GPT-3.5; achieves promising results on the HumanEval benchmark. (paper | tweet)

این هم فایل مربوط به مبحثSVM درس cs229 استنفورد و Andrew Ng...

2). FinGPT - an open-source LLM for the finance sector; it takes a data-centric approach, providing researchers & practitioners with accessible resources to develop FinLLMs. (paper | tweet)

1). Voicebox - an all-in-one generative speech model; it can synthesize speech across 6 languages; it can perform noise removal, content editing, style conversion, and more; it's 20x faster than current models and outperforms single-purpose models through in-context learning. (paper | tweet)

http://metrics.aman.ai سلام. یک رفرنس خیلی خوب برای یادگیری متریک‌های classification و regression. #classification #regression