ai_book
ai_book copied to clipboard
YeonwooSung
AI book for everyone
[Tips for best training results (Ultralytics)](https://docs.ultralytics.com/ko/yolov5/tutorials/tips_for_best_training_results/)
AWS Bedrock is a serverless service that allows the users to utilize Foundation Models ## Sample Codes - [Hybrid Search Postgres Opensearch Bedrock](https://github.com/aws-samples/hybrid-search-postgres-opensearch-bedrock) - [Serverless RAG Demo](https://github.com/aws-samples/serverless-rag-demo) - [Multimodal RAG...
[paper](https://arxiv.org/abs/2310.06816), [code](https://github.com/jxmorris12/vec2text) ## Abstract How much private information do text embeddings reveal about the original text? We investigate the problem of embedding \textit{inversion}, reconstructing the full text represented in dense...
# Understanding Retrieval Augmentation for Long-Form Question Answering [paper](https://arxiv.org/abs/2310.12150) ## Summary (tl;dr) Explores retrieval-augmented language models on long-form question answering; finds that retrieval is an important component but evidence documents...
[Mistral-7b-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1), [Zephyr-7b-alpha](https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha) - Mistral-7b outperformed Llama2-13b-hf and gpt-3.5-turbo - Zephyr-7b-alpha outperformed mistral-7b, and beat Llama2-70b ## DPO vs PPO (DPO is better for finetuning?) - Zephyr-7b-alpha is a finetuned model...
[Open X-Embodiment: 로봇 학습 데이터세트 및 RT-X 모델](https://robotics-transformer-x.github.io/) - 최대 규모의 오픈 소스 데이터 셋 - 21개 기관의 협업을 통해 22개의 다른 로봇으로부터 수집한 데이터 셋 - 527개의 스킬셋(16만개의 태스크)...
[ExecuTorch](https://pytorch.org/executorch/stable/index.html) [ExecuTorch Runtim Overview](https://pytorch.org/executorch/stable/runtime-overview.html) ExecuTorch is a PyTorch platform that provides infrastructure to run PyTorch programs everywhere from AR/VR wearables to standard on-device iOS and Android mobile deployments. One of...
[DeepSpeen Ulysses](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-ulysses) - 기존 시스템보다 4배 더 긴 시퀀스 길이를 제공, 백만개 이상의 토큰이 포함된 시퀀스로 훈련 가능 - 통신이 10배 이상 감소하여 처리량이 최대 2.5배 향상. 처리량이 175 TFlops/GPU...
Contrastive search is originally proposed in "[A Contrastive Framework for Neural Text Generation"](Contrastive search is originally proposed in "A Contrastive Framework for Neural Text Generation") at NeurIPS 2022. The contrastive...
[LLaMA.cpp](https://github.com/ggerganov/llama.cpp) [whisper.cpp](https://github.com/ggerganov/whisper.cpp) [LLaMA rust crate](https://docs.rs/llama/latest/llama/) [Rustformer - LLaMA](https://github.com/rustformers/llm)