Shadowgraph
TrendingBuzzResearch
Chrome extension
Sign in

mdozmorov/MachineLearning_notes

Machine learning and deep learning resources

574

367 commits

updated Oct 13, 2026

See the code

README

https://github.com/ed-donner/llm_engineering?tab=readme-ov-file# Machine- and Deep Learning resources

License: MIT PR's Welcome

Machine and AI/deep learning and data analysis resources. Please, contribute and get in touch! See MDmisc notes for other programming and genomics-related notes.

Table of content

Cheatsheets

Awesome Deep Learning

Keras, Tensorflow

PyTorch

JAX

JAX is a combination of Automatic Differentiation and XLA (Accelerated Linear ALgebra). XLA is a compiler developed by Google to work on TPU units. Jax has Numpy as its higher layer of abstraction, and works the same way on CPU, GPU, and TPU (much faster).

  • awesome-jax - JAX - A curated list of resources

  • JAX - Jupyter (Colab) notebooks introducing JAX basic (jit, vmap, pmap, grad, and other) and advanced concepts, by @yvrjsharma

Graph Neural Networks

Transformers

DL Books

DL Courses & Tutorials

DL Videos

DL Papers

  • Model Context Protocol - analysis of MCP ecosystem, architecture (host, client, server), adoption (Table 1), servers (Table 2), security issues (Table 3) and recommendations.
    Paper Hou, Xinyi, Yanjie Zhao, Shenao Wang, and Haoyu Wang. “Model Context Protocol (MCP): Landscape, Security Threats, and Future Research Directions.” ACM Transactions on Software Engineering and Methodology 35, no. 10 (2026): 1–37. https://doi.org/10.1145/3796519.
  • Paper2Agent - converting papers into agents by constructing a model context protocol (MCP) server (Paper2Skill, Paper2MCP). Implemented with Claude Code. Outperforms Biomni. AlphaGenome agent for variant analysis, 22 MCP tools. Scanpy agent for single-cell analysis, 7 tools. TISSUE (Transcript Imputation with Spatial Single-cell Uncertainty Estimation) agent. Tested on 100 papers. Agents can be hosted at https://huggingface.co/spaces. GitHub explaining how parts play together. Online paper2agent.ai.
    Paper Miao, Jiacheng, Joe R. Davis, Yaohui Zhang, Jonathan K. Pritchard, and James Zou. “Reimagining Research Papers as Interactive and Reliable AI Agents.” Nature, ahead of print, September 16, 2026. https://doi.org/10.1038/s41586-026-11044-y.
  • HNet - Hierarchical network (U-net-inspired, tokenization-free architecture for end-to-end sequencing modeling. Raw data processed by encoder (SSM, Mamba-2 layers, downsampled, passed through the main network (Transformer layers, can itself be an H-net), and upsampled by a decoder (a smoothing module and an upsampler). Novel dynamic chunking (a routing module for boundary prediction and a downsampler module retaining those boundaries). Dynamically predicts chunking boundaries as spots where the embeddings of nearby tokens diverge, and those boundaries are selected for processing. Careful sequence of projection and normalization, adjustment of optimization parameters for each layer, regularization, other techniques for hierarchical sequence modeling and training. Outperforms MambaByte, LlamaByte, SpaceByte and its modifications (HellaSwag, high computational efficiency. HuggingFace.
    Paper Hwang, Sukjun, Brandon Wang, and Albert Gu. “Dynamic Chunking for End-to-End Hierarchical Sequence Modeling.” arXiv:2507.07955. Preprint, arXiv, July 15, 2025. https://doi.org/10.48550/arXiv.2507.07955.
  • Vision Transformer - pure transformer applied to sequences of 16x16 image patches. Architecture experiments. GitHub references other papers about visual transformers.
    Paper Dosovitskiy, Alexey, Lucas Beyer, Alexander Kolesnikov, et al. “An Image Is Worth 16x16 Words: Transformers for Image Recognition at Scale.” arXiv:2010.11929. Preprint, arXiv, June 3, 2021. https://doi.org/10.48550/arXiv.2010.11929.

  • annotated_deep_learning_paper_implementations - a collection of simple PyTorch implementations of neural networks and related algorithms, with explanations. Jupyter notebooks. Website rendering

  • best_AI_papers_2022 - A curated list of the latest breakthroughs in AI (in 2022) by release date with a clear video explanation, link to a more in-depth article, and code.

  • DeepMind Research - implementations and illustrative code to accompany DeepMind publications. Jupyter notebooks and data, list of projects

  • Lee, Benjamin D, Anthony Gitter, Casey S Greene, Sebastian Raschka, Finlay Maguire, Alexander J Titus, Michael D Kessler, et al. “Ten Quick Tips for Deep Learning in Biology.” ArXiv 29 May 2021 - 1. Use appropriate method; 2. Establish baseline; 3. Train reproducibly; 4. Know your data; 5. Select sensible architecture; 6. Optimize hyperparameters; 7. Mitigate overfitting; 8. Maximize interpretability; 9. Avoid over-interpretation; 10. Prioritize research ethics. Summary in Figure 1. References. Latest version

  • Sebastian Ruder, “An Overview of Gradient Descent Optimization Algorithms.” June 15, 2017 - Gradient descent optimization algorithm review, by . Definitions, intuitive progression of algorithm improvements. Gradient descent variants: Batch, Stochastic, Mini-batch. Gradient descent algorithme: Momentum, Nesterov accelerated gradient, Adagrad, Adadelta, RMSprop, Adam, AdaMax, Nadam. Visualizattion. Parallel implementations.

  • eugeneyan/applied-ml - Papers & tech blogs by companies sharing their work on data science & machine learning in production.

  • 2020: A Year Full of Amazing AI papers- A Review - A curated list of the latest breakthroughs in AI by release date with a clear video explanation, link to a more in-depth article, and code

  • awesome-deepbio - A curated list of awesome deep learning publications in the field of computational biology

  • Deep Learning Papers Reading Roadmap - Deep Learning papers reading roadmap for anyone who are eager to learn this amazing tech!

  • Papers with code - Systematic collection of machine- and deep learning papers with code, State-of-the-art, Methods, Datasets

  • Deep_learning_examples - Examples of using deep learning in Bioinformatics. Deep Learning in Bioinformatics

  • deep_learning_papers - A place to collect papers that are related to deep learning and computational biology, by Harold Pimentel

  • Deep-Learning-Papers-Reading-Roadmap - Deep Learning papers reading roadmap for anyone who are eager to learn this amazing tech!

  • deeplearning-biology - A list of papers on deep learning implementations in biology

  • Machine-learning-for-proteins - List of papers about machine learning for proteins

  • Key Papers in Deep Reinforcement Learning, Twitter

  • LeCun, Bengio, and Hinton, “Deep Learning.” - Classical deep learning review. Areas of application, historical development, principles of supervised learning, stochastic gradient descent (Figure 1 - illustration of forward and backpropagation, with equations), convolutional neural networks for image recognition and in other areas, language processing, recurrent neural networks, LSTMs

  • Vincent et al., “Extracting and Composing Robust Features with Denoising Autoencoders.” - Denoising autoencoder paper, statistical formulations

  • Schmidhuber, “Deep Learning in Neural Networks.” - Deep overview of deep learning history. Year-by-year description of types of DL, approaches, algorithmic (backpropagation) improvements, problems, and solutions

  • Scientific-LLM-Survey - Scientific Large Language Models: A Survey on Biological & Chemical Domains

  • TextGrad - text (prompt) optimization by backpropagating LLM-generated feedback. Each AI system is transformed into a computational graph where components exchange text, code, and images through complex (not necessarily differentiable) functions. TextGrad generates feedback in the form of text to guide the modifications of unstructured variables in a system - textual gradient descent analogy. Tested on LeetCode Hard, Google-proof Question Answering (GPQA), Massive Multitask Language Understanding (MMLU), MathVista, ScienceQA, GSM8k datasets. Outperforms Reflexion, OPRO, DSPy.

    Paper Yuksekgonul, Mert, Federico Bianchi, Joseph Boen, Sheng Liu, Pan Lu, Zhi Huang, Carlos Guestrin, and James Zou. “Optimizing Generative AI by Backpropagating Language Model Feedback.” Nature 639, no. 8055 (March 20, 2025): 609–16. https://doi.org/10.1038/s41586-025-08661-4.

DL Genomics

  • dnaHNet - the HNet architecture specifically tuned for DNA end-to-end tokenizer-free modeling. Differentiable dynamic chunking mechanism, balancing compression with predictive accuracy. Outperforms StripedHyena2, Evo2. Evaluated on protein variant efects (MaveDB), Gene essentiality (DEG), Genomic structure interpretation (NCBI). Computationally efficient, scales well.
    Paper Shah, Arnav, Junzhe Li, Parsa Idehpour, et al. “dnaHNet: A Scalable and Hierarchical Foundation Model for Genomic Sequence Learning.” Version 3. Preprint, arXiv, 2026. https://doi.org/10.48550/ARXIV.2602.10603.
  • VoxTell - text-guided 3D image segmentation (CT, MRI, PET). nnU-Net framework, ResEncL encoder for vision, Qwen3-Embedding-4B encoder for text, cross-modal integration via a two-layer MLP. Repeated image-text integration through the decoding hierarchy. Loss function promoting this integration. Tested on 158 public datasets with over 62K volumetric scans (Appendix B). Outperforms TotalSegmentator, BioMedParse, Text3dSAM, SegVol, SAT on zero-shot segmentation performance and cross-modality and unseen concept generalization, Mask2Former and MaskFormed on ablation study.
    Paper Rokuss, Maximilian, Moritz Langenberg, Yannick Kirchhoff, et al. “VoxTell: Free-Text Promptable Universal 3D Medical Image Segmentation.” Version 1. Preprint, arXiv, 2025. https://doi.org/10.48550/ARXIV.2511.11450.
  • EVO 2 - generative model trained on 9.3 trillion DNA bases across all domains of life (OpenGenome2 dataset with non-redundant nucleotide sequence data). StripedHyena 2 architecture, convolutional multi-hybrid architecture, operators arranged in a striped pattern. IM token context window with single-nucleotide resolution. 7B and 40B parameters versions. Predicts the functional impact of genetic variation on protein, RNA, organismal fitness, regulatory motifs. Zero-shot performance better than other models (AlphaMissense, ESM-1b, GPN-MSA). Sparse autoencoders to interpret what the model learns. Inference mode to generate epigenomic structure. EvoMech - mechanical interpretability visualizer, predicted features for various genomes. Evo designer - DNA generation from nucleotides and specific prompts or score a sequence's entropy. Models on Huggingface.
    Paper Brixi, Garyk, Matthew G. Durrant, Jerome Ku, et al. “Genome Modeling and Design across All Domains of Life with Evo 2.” Preprint, Genomics, February 21, 2025. https://doi.org/10.1101/2025.02.18.638918.
  • AlphaGenome - DNA deep learning model with 1Mb input to predict base-level genomic tracks (gene expression, transcription initiation, chromatin accessibility, histone modifications, TF binding, Hi-C contact maps, splice sites). 1Mb input partitioned into 131kb overlapping sequences, U-net-style design (encoder-transformer-decoder, rotary position embedding). Trained on ENCODE, GTEx, 4D Nucleome, FANTOM5. Outperforms nearly all models (Enformer, Borzoi, Pangolin, SpliceAI, DeltaSplice, others). Implemented using JAX and Haiku, pseudocode for main methods steps. API, Google Deepmind blog.
    Paper Avsec, Žiga, Natasha Latysheva, Jun Cheng, et al. “Advancing Regulatory Variant Effect Prediction with AlphaGenome.” Nature 649, no. 8099 (2026): 1206–18. https://doi.org/10.1038/s41586-025-10014-0.
  • Chai-2 model for antibody development. Compared with AlphaProteo, RFDiffusion. Tested on IL7R, InsulinR, PD-L1, PDGFb, TNFa, experimental validation. Target-specific and cross-reactive proteins.
Paper Chai Discovery, Jacques Boitreaud, Jack Dent, Matthew McPartlon, Joshua Meier, Vinicius Reis, Alex Rogozhonikov, and Kevin Wu. “Chai-1: Decoding the Molecular Interactions of Life.” BioRxiv, 2024, 2024–10. https://doi.org/10.1101/2024.10.10.615955
  • Evo - foundational model for sequence prediction (mutational effect, activity of regulatory DNA, gene essentiality) and generation (CRISPR-Cas complexes, transposons). Striped Hyena-based architecture using 29 Hyena layers, 3 rotary position embedding, over 131K single-nucleotide token window. Pretrained on the OpenGenome dataset with more than 80K bacterial and archaeal genomes, millions of phage and plasmid sequences, viruses excluded. Compared Transformer, Mamba, Hyena, StripedHyena architectures.
    Paper Nguyen, Eric, Michael Poli, Matthew G Durrant, Brian Kang, Dhruva Katrekar, David B Li, Liam J Bartie, et al. “Sequence Modeling and Design from Molecular to Genome Scale with Evo,” 2024. https://www.science.org/doi/10.1126/science.ado9336
  • Review or genome Language Models (gLMs) using the transformer architecture. Intro about deep learning in genomics, references. DNA tokenization approaches (one-hot, k-mer (overlapping and not), byte-pair encoding). Masked Language Modeling (MLM), Autoregressive Language Modeling (ALM), State Space Models (SSM). The Hyena layer and the Mamba layer. Description of individual models: SATORI (self-attention-based model to detect regulatory element interactions), Enformer, Borzoi, C.Origami, DNABERT (-XL, -2), the Nucleotide transformer, GENA-LM, GPN, HyenaDNA, Evo (Striped Hyena), summarized in Table 1. Benchmarking datasets, the BEND benchmarking paper shows gLMs do not consistently outperform supervised baselines. Limitations - not accounting for long-range interactions, cell type specificity, high computing cost. Interpretability approaches - attention scores, attention flow, rollout, SHAP. Excellent supplementary texts describing the attention mechanisms, each model (SSM, Hyena), add-and-norm layers, skip connections, fully connected layers, encoder-decoder interplay. Supplementary Table S1 - an in-depth summary of all models. Supplementary figures - schematic representation.
    Paper Consens, Micaela E., Cameron Dufault, Michael Wainberg, Duncan Forster, Mehran Karimzadeh, Hani Goodarzi, Fabian J. Theis, Alan Moses, and Bo Wang. “Transformers and Genome Language Models.” Nature Machine Intelligence 7, no. 3 (March 13, 2025): 346–62. https://doi.org/10.1038/s42256-025-01007-9.

  • genomicsnotebook - Genomics Data Analysis with Jupyter Notebooks on Azure.

  • LLM-for-genomics-training - Tutorial on large language models for genomics

  • Machine Learning for Genomics - ML4GLand is a community for that develops and maintains tools (primarily in Python) for genomics sequence based machine learning.

  • SEQUOIA - a linearized transformer model for gene expression prediction from pathology slides. Uses UNI (foundational model for slides), compared with ResNet50. Compared with tRNAsformer, HE2RNA. Trained on 7584 tumor samples across 16 cancer types (TCGA), validated on independent cohorts (CPTAC, Tempus). BRCA shows best performance, the model predicts the risk of breast cancer recurrence. About 15K out of 20K genes can be predicted, well-known signature genes are predicted best. Detected 272 genes significantly associated with recurrence. Predicts spatially-specific gene expression. Python, GitHub.

    Paper Pizurica, Marija. “Digital Profiling of Gene Expression from Histology Images with Linearized Attention.” Nature Communications, 2024.
  • MAMBA original paper. Selective state space models (SSMs), integrated into a simple architecture without attention or MLP blocks. Performance improves with longer sequence length, up to million tokens. Detailed theory. Compared with H3, Gated MLP architectures. Outperforms other models on language tasks, DNA modeling, audio modeling and generation, fast.
    Paper Gu, Albert, and Tri Dao. “Mamba: Linear-Time Sequence Modeling with Selective State Spaces.” arXiv, May 31, 2024. https://doi.org/10.48550/arXiv.2312.00752.
  • HyenaDNA - a genomic foundation model with context lengths of up to 1M tokens at the single nucleotide-level. A Hyena operator (long convolutions and element-wise gate layers). Pretrained on the human reference genome to predict next token, masked tokens. Scales sub-quadratically. Tested on GenomicBenchmarks dataset (classification, prediction), compared with DNABERT, Nucleotide Transformer.
    Paper Nguyen, Eric, Michael Poli, Marjan Faizi, Armin W Thomas, Callum Birch Sykes, Michael Wornow, Aman Patel, et al. “HyenaDNA: Long-Range Genomic Sequence Modeling at Single Nucleotide Resolution,” https://arxiv.org/abs/2306.15794
  • Hyena original paper. Replaces attention with long convolutions and data-controlled gating. The Hyena operator definition as a recurrence of long convolution and multiplicative element-wise gating of the (projected) input, Figure 1.1. Excels on long sequences (8192 tokens), compared with FlashAttention. Theory, prior work on state-space models.
    Paper Poli, Michael, Stefano Massaroli, Eric Nguyen, Daniel Y. Fu, Tri Dao, Stephen Baccus, Yoshua Bengio, Stefano Ermon, and Christopher Ré. “Hyena Hierarchy: Towards Larger Convolutional Language Models.” arXiv, April 19, 2023. https://doi.org/10.48550/arXiv.2302.10866.

  • Avsec, Žiga, Vikram Agarwal, Daniel Visentin, Joseph R. Ledsam, Agnieszka Grabska-Barwinska, Kyle R. Taylor, Yannis Assael, John Jumper, Pushmeet Kohli, and David R. Kelley. “Effective Gene Expression Prediction from Sequence by Integrating Long-Range Interactions.” Nature Methods, (October 2021) - Enformer - a deep learning model (transformers) to predict epigenetic and gene expression profiles (128bp resolution) from human and mouse cell types using only the DNA sequence as input, incorporating information from up to 100 kb on either side of the target locus (200kb of input DNA sequence). Transformers, allow to increase the receptive field up to 100kb, in contrast to 20kb for Basenji2 or ExPecto. Significant increase in performance. Predictions using CAGE data. Improves variant effect prediction on eQTL data. Excellent transformer description.

  • Foundation models in bioinformatics, language, vision, graph, multimodel. Specific models, tools, briefly described, organized in categories (Figures, tables). Genes, proteomics, drug discovery, cell clustering, multi-omics integration.

    Paper Guo, Fei, Renchu Guan, Yaohang Li, et al. “Foundation Models in Bioinformatics.” National Science Review 12, no. 4 (2025): nwaf028. https://doi.org/10.1093/nsr/nwaf028.
  • Deep learning for genomics. Network architectures - CNN, RNN, autoencoders, hybrid architectures, transformer-based LLMs. Model interpretation - saliency maps, attention scores. Transfer, multitask, multi-view learning. Applications - prediction of gene expression, promoters, enhancers, splicing, DNA and RNA binding sites, mutation effect, subcellular locations. Proteins, structural classification, prediction, secondary, tertiary structure.
    Paper Yue, Tianwei, Yuanxin Wang, Longxiang Zhang, et al. “Deep Learning for Genomics: A Concise Overview.” arXiv:1802.00810. Preprint, arXiv, October 4, 2023. https://doi.org/10.48550/arXiv.1802.00810.

DL Tools

  • Interactive_Tools - Interactive Tools for Machine Learning, Deep Learning and Math. Play with deep neural network in browser

  • ivy - The Unified Machine Learning Framework supporting JAX, TensorFlow, PyTorch, MXNet, and Numpy. Python module. Documentation

  • keras - Deep Learning for humans http://keras.io/

  • MXNet-Gluon-Style-Transfer - neural artistic style transfer using MXNet. PyTorch and Torch implementations available

  • openai.com - GPT-3 Access Without the Wait (API access to GPT-3)

  • OpenCV - Open Source Computer Vision library. GitHub, opencv-python - CPU-only OpenCV packages for Python. Documentation. Video - 3h OpenCV crash course

  • pathology_learning - Using traditional machine learning and deep learning methods to predict stuff from TCGA pathology slides

  • ruta - Unsupervised Deep Architechtures in R, autoencoders. Requires Keras and TensorFlow. Book

  • tensor2tensor - Library of deep learning models and datasets designed to make deep learning more accessible and accelerate ML research

  • Janggu - deep learning interface to genomic data (FASTA, BAM, BigWig, BED, GFF). Numpy-like Bioseq and Cover objects accessable by Keras. Includes model evaluation and interpretation features. Pypi, Docs, Janggu - Deep learning for genomics

  • maui - Multi-omics Autoencoder Integration. Latent factors from different data types (stacked variational autoencoders), and their clustering, testing for association with survival. Tested vs. latent factors extracted using Multifactor Analysis (MFA) and iCluster+, on TCGA colorectal cancer RNA-seq, SNPs, CNVs. Evaluation of Colorectal Cancer Subtypes and Cell Lines Using Deep Learning

  • Ludwig is a toolbox built on top of TensorFlow that allows to train and test deep learning models without the need to write code. GitHub

  • Mask_RCNN - Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow

  • PennAI - AI-Driven Data Science, entry-level machine learning interface for non-experts. A System for Accessible Artificial Intelligence

Auto ML

DL models

  • PyTorch model zoos

  • Keras model zoos

  • Tensorflow model zoos

  • Kipoi - a model zoo for genomics. Examples of transfer learning, predicting pathogenic variants, TFBSs. Avsec et al., “The Kipoi Repository Accelerates Community Exchange and Reuse of Predictive Models for Genomics.”, GitHub repo

  • Caffe Model Zoo

  • BERT, Bidirectional Encoder Representations from Transformers, for natural language processing tasks. Model architecture, implemented using TensorFlow. Applications - Masked Language Model, next sentence prediction. Excels in several benchmarks. Pretrained models and code. See also BioBert

  • DNABERT - pre-trained Bidirectional Encoder Representation from Transformers (BERT) model for DNA sequence, captures contextual information, 512bp window. Input - overlapping k-mer tokens (6-mer perform best). Architecture details (12 transformer encoder layers, learning rate increased then decreased, 120 epochs sufficient), comparison with RNN, CNN, Transformers. Outperforms DeePromoter, PromID, FPROM, FirstEF promoter prediction methods, DeepBind, DeepSEA, Basset, DeepSite, DanQ, DESSO on on TFBS prediction. Pretrained on 10-510-long sequences, to predict missing nucleotides. Fine-tuned for promoter region prediction, transcription factor binding sites, splicing sites, genetic variants. visualization of learned attention patterns. [CLS] classification token at the beginning of sequence, [SEP] at the end. Visualization of attention scores. Supplementary material contains detailed description of the model's architecture, input data, training, evaluation, various case scenarios, visualization. Bringing BERT to the field blog post contains more details and links to transformer resources.

    Paper Ji, Yanrong, Zhihan Zhou, Han Liu, and Ramana V Davuluri. “[DNABERT: Pre-Trained Bidirectional Encoder Representations from Transformers Model for DNA-Language in Genome](https://doi.org/10.1093/bioinformatics/btab083).” Bioinformatics, (August 9, 2021)
  • TabPFN - the Tabular Prior-data Fitted Network, a tabular foundational model (generative transformer-based), outperforms other models, excels on small datasets but scalable, fast. Classification or regression problems Leverages in-context learning (ICL), generates a large corpus of synthetic tabular datasets (prior in Bayesian terms, approach based on structural causal models, SCMs) and then train a self-attention-based transformer neural network. Pre-training - predict masked targets of all synthetic datasets. Z-normalization of inputs. Sample-specific learning, feature-specific attention attending to features across samples, invariant to the order of samples and features. Tested on two benchmark datasets, AutoML benchmark and OpenML-CTR23, outperforms XGBoost, CatBoost, random forest. Approaches for interpretation. Python implementation.
    Paper Hollmann, N., Müller, S., Purucker, L. et al. Accurate predictions on small data with a tabular foundation model. Nature 637, 319–326 (2025). https://doi.org/10.1038/s41586-024-08328-6

  • folding_tools - a collection of protein folding tools

  • progressive_growing_of_gans - Progressive Growing of GANs for Improved Quality, Stability, and Variation — Official TensorFlow implementation of the ICLR 2018 paper

DL projects

Language models

ChatGPT, Gemini, NotebookLM, Claude, DeepSeek, Together.ai, Kimi, Qwen, Grok, OpenRouter, Groq, Storm, [Perplexity.ai, Suna, MiniMax, Stackoverflow.ai

  • awesome-chatgpt - Curated list of awesome tools, demos, docs for ChatGPT and GPT-3

  • awesome-llm-courses - A curated list of awesome online courses about Large Langage Models (LLMs)

  • agenticSeek - A 100% local alternative to Manus AI, this voice-enabled AI assistant autonomously browses the web, writes code, and plans tasks while keeping all data on your device. Tailored for local reasoning models, it runs entirely on your hardware, ensuring complete privacy and zero cloud dependency.

  • chatbox - User-friendly Desktop Client App for AI Models/LLMs (GPT, Claude, Gemini, Ollama...)

  • chatgpt-clone - Build Yo'own ChatGPT with OpenAI API & Gradio. A Python app for web browser intercage to ChatGPT.

  • DeepSeek-V3 - a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token. Excellent performance at a fraction of the cost. Web chatbot.

  • Emergent - an LLM model for website/app creation. Uses Claude, ChatGPT, connects to GitHub

  • h2ogpt - open-source GPT with document and image Q&A, 100% private chat, no data leaks, Apache 2.0 https://arxiv.org/pdf/2306.08161.pdf Live Demo: https://gpt.h2o.ai/

  • Hands-On-Large-Language-Models - Official code repo for the O'Reilly Book - "Hands-On Large Language Models" by Jay Alammar and Maarten Grootendorst. PyTorch, Jupyter notebooks, can be opened in Google colab.

  • litellm - Python SDK, Proxy Server (LLM Gateway) to call 100+ LLM APIs in OpenAI format - [Bedrock, Azure, OpenAI, VertexAI, Cohere, Anthropic, Sagemaker, HuggingFace, Replicate, Groq]. Docs

  • llm-course - Course to get into Large Language Models (LLMs) with roadmaps and Colab notebooks.

  • llm_engineering - repo for the Mastering AI and LLM Engineering course. Python, includes videos.

  • LLMs-from-scratch - Implementing a ChatGPT-like LLM from scratch, step by step, the Python code for coding, pretraining, and finetuning a GPT-like LLM. By Sebastian Raschka. For the book Build a Large Language Model (From Scratch).

  • LLMsPracticalGuide - A curated list of practical guide resources of LLMs (LLMs Tree, Examples, Papers)

  • lobe-chat - an open-source, modern-design AI chat framework. Supports Multi AI Providers( OpenAI / Claude 3 / Gemini / Ollama / Azure / DeepSeek), Knowledge Base (file upload / knowledge management / RAG ), Multi-Modals (Vision/TTS/Plugins/Artifacts). One-click FREE deployment of your private ChatGPT/ Claude application.

  • mlc-llm - Enable everyone to develop, optimize and deploy AI models natively on everyone's devices. Documentation

  • nanoGPT - The simplest, fastest repository for training/finetuning medium-sized GPTs.

  • ollama - Get up and running with Llama 2 and other large language models locally.

  • openai-cookbook - Examples and guides for using the OpenAI API. Rendered version

  • privateGPT - Interact privately with your documents using the power of GPT, 100% privately, no data leaks.

  • prompt-eng-interactive-tutorial - Anthropic's Interactive Prompt Engineering Tutorial

  • Qwen - LLM created by Alibaba Cloud. Transformer-based, lightweight but among top performers. Website

  • Storm - An LLM-powered knowledge curation system that researches a topic and generates a full-length report with citations. GitHub

Audio, music, voice

  • abogen - Generate audiobooks from EPUBs, PDFs and text with synchronized captions.

  • buzz - Buzz transcribes and translates audio offline on your personal computer. Powered by OpenAI's Whisper.

  • ebook2audiobook - Convert ebooks to audiobooks with chapters and metadata using dynamic AI models and voice cloning. Supports 1,107+ languages!

  • ElevenReader - text to speech converter, for ebooks (EPUB), PDFs, text files. iOS and Android mobile apps.

  • Flow Music - Google’s AI-powered music creation platform for generation of full songs from prompts, creating AI music videos, remix/split audio, build music tools, and publish/share.

  • insanely-fast-whisper - An opinionated CLI to transcribe Audio files w/ Whisper on-device! Powered by Transformers, Optimum & flash-attn. Transcribe 150 minutes (2.5 hours) of audio in less than 98 seconds - with OpenAI's Whisper Large v3. Blazingly fast transcription is now a reality!

  • Jukebox - music generation neural network. Hierarchical Vector Quantised-Variational AutoEncoder (VQ-VAE) architecture, three separate temporal resolutions. Able to generate singing from lyrics, extend music examples. Dhariwal et al., “Jukebox: A Generative Model for Music.”, Blog post with examples of generated music

  • june - Local voice chatbot for engaging conversations, powered by Ollama, Hugging Face Transformers, and Coqui TTS Toolkit

  • Magenta - Music and Art Generation with Machine Intelligence

  • MINIMAX - text-to-speech, music creation. Voice clone, voice design, noise removal.

  • OpenVoice - voice cloning tool, transfer voice tones to pronounce different words, even in different language.

  • Real-Time-Voice-Cloning - Clone a voice in 5 seconds to generate arbitrary speech in real-time. Learn voice characteristics from a short audio clip and perform text-to-speech conversion using this voice.

  • Project DeepSpeech - A TensorFlow implementation of Baidu's DeepSpeech architecture. Transcribe audio data, English model available. Documentation

  • SpeechBrain - A PyTorch-based Speech Toolkit for speech/speaker recognition, speech enhancement, processing, and more. GitHub repo

  • Unmute (Kyutai) - open source text-to-speech and speech-to-text model, GitHub.

  • vampnet - music generation with masked transformers. arXiv paper, supplementary page

  • VibeVoice - VibeVoice is a novel framework designed for generating expressive, long-form, multi-speaker conversational audio, such as podcasts, from text.

  • voicebox - The open-source voice synthesis studio powered by Qwen3-TTS.

Image, video, vision

  • PaperBanana - Automating Academic Illustration for AI Scientists. Visual language model trained on NeurIPS diagrams. GitHub

  • Krea - image and video generation from text prompt, enhancing image quality, image editing from text prompt.

  • Sora - OpenAI's text-to-video generation tool that allows users to create, animate, and remix videos from text prompts, images, and existing videos.

  • 18 All-Time Classic Open Source Computer Vision Projects for Beginners by Analytics Vidhya

  • 500-AI-Machine-learning-Deep-learning-Computer-vision-NLP-Projects-with-code - links to many ML/DL projects and resources

  • Top 10 Computer Vision Papers 2020 - A list of the top 10 computer vision papers in 2020 with video demos, articles, code and paper reference.

  • awesome-ai-art-image-synthesis - A list of awesome tools, ideas, prompt engineering tools, colabs, models, and helpers for the prompt designer playing with aiArt and image synthesis. Covers Dalle2, MidJourney, StableDiffusion, and open source tools.

  • Bringing-Old-Photos-Back-to-Life - Old Photo Restoration (Official PyTorch Implementation)

  • CLIP - CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image

  • CogView2 - official code repo for paper "CogView2: Faster and Better Text-to-Image Generation via Hierarchical Transformers" (ArXiv), Example

  • Cosmos - Cosmos is a world model development platform that consists of world foundation models, tokenizers and video processing pipeline to accelerate the development of Physical AI at Robotics & AV labs. Cosmos is purpose built for physical AI. The Cosmos repository will enable end users to run the Cosmos models, run inference scripts and generate videos. Docker options.

  • HunyuanVideo - A Systematic Framework For Large Video Generation Model

  • DALL-E-2 - a new AI system that can create realistic images and art from a description in natural language. Tweet. How This A.I. Draws Anything You Describe 16m video by ColdFusion. HOW DALL-E COULD POWER A CREATIVE REVOLUTION, The Verge. DALLE2-pytorch - Implementation of DALL-E 2, OpenAI's updated text-to-image synthesis neural network, in Pytorch. min-dalle - min(DALL·E) is a fast, minimal port of DALL·E Mini to PyTorch

  • Deep-Live-Cam - real time face swap and one-click video deepfake with only a single image.

  • JoJoGAN - Official PyTorch repo for JoJoGAN: One Shot Face Stylization. Arxiv

  • Photo restoration with GFP-GAN - GFPGAN aims at developing Practical Algorithms for Real-world Face Restoration. Online version

  • Reve - image editing model

  • seedream - advanced image generation/editing model

  • Selfie2Anime online tool and a GitHub repo

  • stablediffusion - High-Resolution Image Synthesis with Latent Diffusion Models

  • VGG - Visual Geometry Group. Large source of image datasets. Also, practicals, demos, software, publications, more.

DL Misc

  • 100_days_of_CUDA - notes about learning CUDA

  • geospy.ai - location identification from photos

  • app.wombo.art - deep generative model dreaming awesome images from text, Android and iOS apps available. Tweet describing the VQGAN+CLIP technology behind it

  • CSrankings - A web app for ranking computer science departments according to their research output in selective venues, and for finding active faculty across a wide range of areas. Website

  • ColossalAI - A Unified Deep Learning System for Big Model Era. Scaling deep learning models using data, pipeline, tensor, and sequence parallelism. 1D, 2D, 2.5D, 3D distributed operators. Examples of each. Written in PyTorch, needs a configuration file defining parallelism. Benchmarked against DeepSpeed, Megatron-LM.

    Paper Li, Shenggui, Jiarui Fang, Zhengda Bian, Hongxin Liu, Yuliang Liu, Haichen Huang, Boxiang Wang, and Yang You. “Colossal-AI: A Unified Deep Learning System For Large-Scale Parallel Training,” n.d.

Awesome Machine learning

ML Books

ML Courses & Tutorials

ML Videos

ML Papers

  • Whalen, Sean, Jacob Schreiber, William S. Noble, and Katherine S. Pollard. “Navigating the Pitfalls of Applying Machine Learning in Genomics.” Nature Reviews Genetics 23, no. 3 (March 2022): 169–81. https://doi.org/10.1038/s41576-021-00434-9. - Five machine learning problems in genomics, distributional differences, dependency structure, confounding variables, information leakage, unbalanced data. Description, examples, solutions.

  • Domingos, Pedro. “A Few Useful Things to Know about Machine Learning.” Communications of the ACM 55, no. 10 (October 1, 2012): 78. https://doi.org/10.1145/2347736.2347755. Twelve lessons for machine learning. Overview of machine learning problems and algorithms, problem of overfitting, causes and solutions, curse of dimensionality, issues with high-dimensional data, feature engineering, bagging, boosting, stacking, model sparsity. Video lectures

ML Tools

  • mlr3 - Machine learning in R R package, the unified interface to classification, regression, survival analysis, and other machine learning tasks. GitHub repo, mlr3gallery - Examples of problems and code solutions, mlr3 Manual - mlr3 bookdown. More on the mlr3 package site, including videos

ML Misc

Material in Chinese

  • Autopilot-Notes - Autonomous driving notes summarizing the basics, hardware, perception, position, planning, control, product, tools, and manufacturing plan topics.

Material in Russian

  • GigaChat - Russian language-focused language model. Based on the mixture of experts (MoE) architecture, GPT-3, 13B parameters. 2 min Video demo. Leaderboard. Alternatives: YandexGPT5, Cotype. GigaAM for speech processing. Models on Huggingface. @gigachat_bot Telegram bot.
    Paper team, GigaChat, Mamedov Valentin, Evgenii Kosarev, Gregory Leleytner, Ilya Shchuckin, Valeriy Berezovskiy, Daniil Smirnov, et al. “GigaChat Family: Efficient Russian Language Modeling Through Mixture of Experts Architecture.” arXiv, June 11, 2025. https://doi.org/10.48550/arXiv.2506.09440.

  • Scientific_graphics_in_python - matplotlib for scientific graphics. 3 parts, 13 chapters. By Pavel Shabanov

  • ml-course-hse - machine learning course at the Computer Sciences Department, High Schoool of Economy. Multiple years, videos

  • mlcourse_open - OpenDataScience Machine Learning course (Both in English and Russian). Python-based ML course, with video lectures. Video

  • DL_CSHSE_spring2018 - Deep learning, Anton Osokin, Higher School of Economics, Computer Sciences Department (Russian), course material, and video lectures

  • Ordinary Differential Equations - Обыкновенные дифференциальные уравнения, Интерактивный учебник, Илья Щуров (НИУ ВШЭ)

  • Calculus - Математический анализ, Записки лекций, Илья Щуров (НИУ ВШЭ). Tweet

  • mathprofi.ru - Высшая математика – просто и доступно. Mirror

deep-learning
machine-learning

Similar repos

peggy1502/Amazing-Resources

List of references and online resources related to data science, machine learning and deep learning.

435

mikeroyal/Machine-Learning-Guide

Machine learning Guide. Learn all about Machine Learning Tools, Libraries, Frameworks, Large…

710

kandeldeepak46/Keep-Learning

Resources and blogs, curated for learning, computer science, mathematics, data science, computer…

26

ashara12/awesome-deeplearning

A curated list of awesome Deep Learning tutorials, projects and communities.

29,040

ChristosChristofidis/awesome-deep-learning

A curated list of awesome Deep Learning tutorials, projects and communities.

29,040

imteekay/machine-learning-research

✨ AI/ML Research

458

aurimas13/Machine-Learning-Goodness

The Machine Learning project including ML/DL projects, notebooks, cheat codes of ML/DL, useful…

286

guillaume-chevalier/Awesome-Deep-Learning-Resources

Rough list of my favorite deep learning resources, useful for revisiting topics or for reference. I…

1,821

mdozmorov/MachineLearning_notes

Machine learning and deep learning resources

574

367 commits

updated Oct 13, 2026

See the code

README

https://github.com/ed-donner/llm_engineering?tab=readme-ov-file# Machine- and Deep Learning resources

License: MIT PR's Welcome

Machine and AI/deep learning and data analysis resources. Please, contribute and get in touch! See MDmisc notes for other programming and genomics-related notes.

Table of content

Cheatsheets

Awesome Deep Learning

Keras, Tensorflow

PyTorch

JAX

JAX is a combination of Automatic Differentiation and XLA (Accelerated Linear ALgebra). XLA is a compiler developed by Google to work on TPU units. Jax has Numpy as its higher layer of abstraction, and works the same way on CPU, GPU, and TPU (much faster).

  • awesome-jax - JAX - A curated list of resources

  • JAX - Jupyter (Colab) notebooks introducing JAX basic (jit, vmap, pmap, grad, and other) and advanced concepts, by @yvrjsharma

Graph Neural Networks

Transformers

DL Books

DL Courses & Tutorials

DL Videos

DL Papers

  • Model Context Protocol - analysis of MCP ecosystem, architecture (host, client, server), adoption (Table 1), servers (Table 2), security issues (Table 3) and recommendations.
    Paper Hou, Xinyi, Yanjie Zhao, Shenao Wang, and Haoyu Wang. “Model Context Protocol (MCP): Landscape, Security Threats, and Future Research Directions.” ACM Transactions on Software Engineering and Methodology 35, no. 10 (2026): 1–37. https://doi.org/10.1145/3796519.
  • Paper2Agent - converting papers into agents by constructing a model context protocol (MCP) server (Paper2Skill, Paper2MCP). Implemented with Claude Code. Outperforms Biomni. AlphaGenome agent for variant analysis, 22 MCP tools. Scanpy agent for single-cell analysis, 7 tools. TISSUE (Transcript Imputation with Spatial Single-cell Uncertainty Estimation) agent. Tested on 100 papers. Agents can be hosted at https://huggingface.co/spaces. GitHub explaining how parts play together. Online paper2agent.ai.
    Paper Miao, Jiacheng, Joe R. Davis, Yaohui Zhang, Jonathan K. Pritchard, and James Zou. “Reimagining Research Papers as Interactive and Reliable AI Agents.” Nature, ahead of print, September 16, 2026. https://doi.org/10.1038/s41586-026-11044-y.
  • HNet - Hierarchical network (U-net-inspired, tokenization-free architecture for end-to-end sequencing modeling. Raw data processed by encoder (SSM, Mamba-2 layers, downsampled, passed through the main network (Transformer layers, can itself be an H-net), and upsampled by a decoder (a smoothing module and an upsampler). Novel dynamic chunking (a routing module for boundary prediction and a downsampler module retaining those boundaries). Dynamically predicts chunking boundaries as spots where the embeddings of nearby tokens diverge, and those boundaries are selected for processing. Careful sequence of projection and normalization, adjustment of optimization parameters for each layer, regularization, other techniques for hierarchical sequence modeling and training. Outperforms MambaByte, LlamaByte, SpaceByte and its modifications (HellaSwag, high computational efficiency. HuggingFace.
    Paper Hwang, Sukjun, Brandon Wang, and Albert Gu. “Dynamic Chunking for End-to-End Hierarchical Sequence Modeling.” arXiv:2507.07955. Preprint, arXiv, July 15, 2025. https://doi.org/10.48550/arXiv.2507.07955.
  • Vision Transformer - pure transformer applied to sequences of 16x16 image patches. Architecture experiments. GitHub references other papers about visual transformers.
    Paper Dosovitskiy, Alexey, Lucas Beyer, Alexander Kolesnikov, et al. “An Image Is Worth 16x16 Words: Transformers for Image Recognition at Scale.” arXiv:2010.11929. Preprint, arXiv, June 3, 2021. https://doi.org/10.48550/arXiv.2010.11929.

  • annotated_deep_learning_paper_implementations - a collection of simple PyTorch implementations of neural networks and related algorithms, with explanations. Jupyter notebooks. Website rendering

  • best_AI_papers_2022 - A curated list of the latest breakthroughs in AI (in 2022) by release date with a clear video explanation, link to a more in-depth article, and code.

  • DeepMind Research - implementations and illustrative code to accompany DeepMind publications. Jupyter notebooks and data, list of projects

  • Lee, Benjamin D, Anthony Gitter, Casey S Greene, Sebastian Raschka, Finlay Maguire, Alexander J Titus, Michael D Kessler, et al. “Ten Quick Tips for Deep Learning in Biology.” ArXiv 29 May 2021 - 1. Use appropriate method; 2. Establish baseline; 3. Train reproducibly; 4. Know your data; 5. Select sensible architecture; 6. Optimize hyperparameters; 7. Mitigate overfitting; 8. Maximize interpretability; 9. Avoid over-interpretation; 10. Prioritize research ethics. Summary in Figure 1. References. Latest version

  • Sebastian Ruder, “An Overview of Gradient Descent Optimization Algorithms.” June 15, 2017 - Gradient descent optimization algorithm review, by . Definitions, intuitive progression of algorithm improvements. Gradient descent variants: Batch, Stochastic, Mini-batch. Gradient descent algorithme: Momentum, Nesterov accelerated gradient, Adagrad, Adadelta, RMSprop, Adam, AdaMax, Nadam. Visualizattion. Parallel implementations.

  • eugeneyan/applied-ml - Papers & tech blogs by companies sharing their work on data science & machine learning in production.

  • 2020: A Year Full of Amazing AI papers- A Review - A curated list of the latest breakthroughs in AI by release date with a clear video explanation, link to a more in-depth article, and code

  • awesome-deepbio - A curated list of awesome deep learning publications in the field of computational biology

  • Deep Learning Papers Reading Roadmap - Deep Learning papers reading roadmap for anyone who are eager to learn this amazing tech!

  • Papers with code - Systematic collection of machine- and deep learning papers with code, State-of-the-art, Methods, Datasets

  • Deep_learning_examples - Examples of using deep learning in Bioinformatics. Deep Learning in Bioinformatics

  • deep_learning_papers - A place to collect papers that are related to deep learning and computational biology, by Harold Pimentel

  • Deep-Learning-Papers-Reading-Roadmap - Deep Learning papers reading roadmap for anyone who are eager to learn this amazing tech!

  • deeplearning-biology - A list of papers on deep learning implementations in biology

  • Machine-learning-for-proteins - List of papers about machine learning for proteins

  • Key Papers in Deep Reinforcement Learning, Twitter

  • LeCun, Bengio, and Hinton, “Deep Learning.” - Classical deep learning review. Areas of application, historical development, principles of supervised learning, stochastic gradient descent (Figure 1 - illustration of forward and backpropagation, with equations), convolutional neural networks for image recognition and in other areas, language processing, recurrent neural networks, LSTMs

  • Vincent et al., “Extracting and Composing Robust Features with Denoising Autoencoders.” - Denoising autoencoder paper, statistical formulations

  • Schmidhuber, “Deep Learning in Neural Networks.” - Deep overview of deep learning history. Year-by-year description of types of DL, approaches, algorithmic (backpropagation) improvements, problems, and solutions

  • Scientific-LLM-Survey - Scientific Large Language Models: A Survey on Biological & Chemical Domains

  • TextGrad - text (prompt) optimization by backpropagating LLM-generated feedback. Each AI system is transformed into a computational graph where components exchange text, code, and images through complex (not necessarily differentiable) functions. TextGrad generates feedback in the form of text to guide the modifications of unstructured variables in a system - textual gradient descent analogy. Tested on LeetCode Hard, Google-proof Question Answering (GPQA), Massive Multitask Language Understanding (MMLU), MathVista, ScienceQA, GSM8k datasets. Outperforms Reflexion, OPRO, DSPy.

    Paper Yuksekgonul, Mert, Federico Bianchi, Joseph Boen, Sheng Liu, Pan Lu, Zhi Huang, Carlos Guestrin, and James Zou. “Optimizing Generative AI by Backpropagating Language Model Feedback.” Nature 639, no. 8055 (March 20, 2025): 609–16. https://doi.org/10.1038/s41586-025-08661-4.

DL Genomics

  • dnaHNet - the HNet architecture specifically tuned for DNA end-to-end tokenizer-free modeling. Differentiable dynamic chunking mechanism, balancing compression with predictive accuracy. Outperforms StripedHyena2, Evo2. Evaluated on protein variant efects (MaveDB), Gene essentiality (DEG), Genomic structure interpretation (NCBI). Computationally efficient, scales well.
    Paper Shah, Arnav, Junzhe Li, Parsa Idehpour, et al. “dnaHNet: A Scalable and Hierarchical Foundation Model for Genomic Sequence Learning.” Version 3. Preprint, arXiv, 2026. https://doi.org/10.48550/ARXIV.2602.10603.
  • VoxTell - text-guided 3D image segmentation (CT, MRI, PET). nnU-Net framework, ResEncL encoder for vision, Qwen3-Embedding-4B encoder for text, cross-modal integration via a two-layer MLP. Repeated image-text integration through the decoding hierarchy. Loss function promoting this integration. Tested on 158 public datasets with over 62K volumetric scans (Appendix B). Outperforms TotalSegmentator, BioMedParse, Text3dSAM, SegVol, SAT on zero-shot segmentation performance and cross-modality and unseen concept generalization, Mask2Former and MaskFormed on ablation study.
    Paper Rokuss, Maximilian, Moritz Langenberg, Yannick Kirchhoff, et al. “VoxTell: Free-Text Promptable Universal 3D Medical Image Segmentation.” Version 1. Preprint, arXiv, 2025. https://doi.org/10.48550/ARXIV.2511.11450.
  • EVO 2 - generative model trained on 9.3 trillion DNA bases across all domains of life (OpenGenome2 dataset with non-redundant nucleotide sequence data). StripedHyena 2 architecture, convolutional multi-hybrid architecture, operators arranged in a striped pattern. IM token context window with single-nucleotide resolution. 7B and 40B parameters versions. Predicts the functional impact of genetic variation on protein, RNA, organismal fitness, regulatory motifs. Zero-shot performance better than other models (AlphaMissense, ESM-1b, GPN-MSA). Sparse autoencoders to interpret what the model learns. Inference mode to generate epigenomic structure. EvoMech - mechanical interpretability visualizer, predicted features for various genomes. Evo designer - DNA generation from nucleotides and specific prompts or score a sequence's entropy. Models on Huggingface.
    Paper Brixi, Garyk, Matthew G. Durrant, Jerome Ku, et al. “Genome Modeling and Design across All Domains of Life with Evo 2.” Preprint, Genomics, February 21, 2025. https://doi.org/10.1101/2025.02.18.638918.
  • AlphaGenome - DNA deep learning model with 1Mb input to predict base-level genomic tracks (gene expression, transcription initiation, chromatin accessibility, histone modifications, TF binding, Hi-C contact maps, splice sites). 1Mb input partitioned into 131kb overlapping sequences, U-net-style design (encoder-transformer-decoder, rotary position embedding). Trained on ENCODE, GTEx, 4D Nucleome, FANTOM5. Outperforms nearly all models (Enformer, Borzoi, Pangolin, SpliceAI, DeltaSplice, others). Implemented using JAX and Haiku, pseudocode for main methods steps. API, Google Deepmind blog.
    Paper Avsec, Žiga, Natasha Latysheva, Jun Cheng, et al. “Advancing Regulatory Variant Effect Prediction with AlphaGenome.” Nature 649, no. 8099 (2026): 1206–18. https://doi.org/10.1038/s41586-025-10014-0.
  • Chai-2 model for antibody development. Compared with AlphaProteo, RFDiffusion. Tested on IL7R, InsulinR, PD-L1, PDGFb, TNFa, experimental validation. Target-specific and cross-reactive proteins.
Paper Chai Discovery, Jacques Boitreaud, Jack Dent, Matthew McPartlon, Joshua Meier, Vinicius Reis, Alex Rogozhonikov, and Kevin Wu. “Chai-1: Decoding the Molecular Interactions of Life.” BioRxiv, 2024, 2024–10. https://doi.org/10.1101/2024.10.10.615955
  • Evo - foundational model for sequence prediction (mutational effect, activity of regulatory DNA, gene essentiality) and generation (CRISPR-Cas complexes, transposons). Striped Hyena-based architecture using 29 Hyena layers, 3 rotary position embedding, over 131K single-nucleotide token window. Pretrained on the OpenGenome dataset with more than 80K bacterial and archaeal genomes, millions of phage and plasmid sequences, viruses excluded. Compared Transformer, Mamba, Hyena, StripedHyena architectures.
    Paper Nguyen, Eric, Michael Poli, Matthew G Durrant, Brian Kang, Dhruva Katrekar, David B Li, Liam J Bartie, et al. “Sequence Modeling and Design from Molecular to Genome Scale with Evo,” 2024. https://www.science.org/doi/10.1126/science.ado9336
  • Review or genome Language Models (gLMs) using the transformer architecture. Intro about deep learning in genomics, references. DNA tokenization approaches (one-hot, k-mer (overlapping and not), byte-pair encoding). Masked Language Modeling (MLM), Autoregressive Language Modeling (ALM), State Space Models (SSM). The Hyena layer and the Mamba layer. Description of individual models: SATORI (self-attention-based model to detect regulatory element interactions), Enformer, Borzoi, C.Origami, DNABERT (-XL, -2), the Nucleotide transformer, GENA-LM, GPN, HyenaDNA, Evo (Striped Hyena), summarized in Table 1. Benchmarking datasets, the BEND benchmarking paper shows gLMs do not consistently outperform supervised baselines. Limitations - not accounting for long-range interactions, cell type specificity, high computing cost. Interpretability approaches - attention scores, attention flow, rollout, SHAP. Excellent supplementary texts describing the attention mechanisms, each model (SSM, Hyena), add-and-norm layers, skip connections, fully connected layers, encoder-decoder interplay. Supplementary Table S1 - an in-depth summary of all models. Supplementary figures - schematic representation.
    Paper Consens, Micaela E., Cameron Dufault, Michael Wainberg, Duncan Forster, Mehran Karimzadeh, Hani Goodarzi, Fabian J. Theis, Alan Moses, and Bo Wang. “Transformers and Genome Language Models.” Nature Machine Intelligence 7, no. 3 (March 13, 2025): 346–62. https://doi.org/10.1038/s42256-025-01007-9.

  • genomicsnotebook - Genomics Data Analysis with Jupyter Notebooks on Azure.

  • LLM-for-genomics-training - Tutorial on large language models for genomics

  • Machine Learning for Genomics - ML4GLand is a community for that develops and maintains tools (primarily in Python) for genomics sequence based machine learning.

  • SEQUOIA - a linearized transformer model for gene expression prediction from pathology slides. Uses UNI (foundational model for slides), compared with ResNet50. Compared with tRNAsformer, HE2RNA. Trained on 7584 tumor samples across 16 cancer types (TCGA), validated on independent cohorts (CPTAC, Tempus). BRCA shows best performance, the model predicts the risk of breast cancer recurrence. About 15K out of 20K genes can be predicted, well-known signature genes are predicted best. Detected 272 genes significantly associated with recurrence. Predicts spatially-specific gene expression. Python, GitHub.

    Paper Pizurica, Marija. “Digital Profiling of Gene Expression from Histology Images with Linearized Attention.” Nature Communications, 2024.
  • MAMBA original paper. Selective state space models (SSMs), integrated into a simple architecture without attention or MLP blocks. Performance improves with longer sequence length, up to million tokens. Detailed theory. Compared with H3, Gated MLP architectures. Outperforms other models on language tasks, DNA modeling, audio modeling and generation, fast.
    Paper Gu, Albert, and Tri Dao. “Mamba: Linear-Time Sequence Modeling with Selective State Spaces.” arXiv, May 31, 2024. https://doi.org/10.48550/arXiv.2312.00752.
  • HyenaDNA - a genomic foundation model with context lengths of up to 1M tokens at the single nucleotide-level. A Hyena operator (long convolutions and element-wise gate layers). Pretrained on the human reference genome to predict next token, masked tokens. Scales sub-quadratically. Tested on GenomicBenchmarks dataset (classification, prediction), compared with DNABERT, Nucleotide Transformer.
    Paper Nguyen, Eric, Michael Poli, Marjan Faizi, Armin W Thomas, Callum Birch Sykes, Michael Wornow, Aman Patel, et al. “HyenaDNA: Long-Range Genomic Sequence Modeling at Single Nucleotide Resolution,” https://arxiv.org/abs/2306.15794
  • Hyena original paper. Replaces attention with long convolutions and data-controlled gating. The Hyena operator definition as a recurrence of long convolution and multiplicative element-wise gating of the (projected) input, Figure 1.1. Excels on long sequences (8192 tokens), compared with FlashAttention. Theory, prior work on state-space models.
    Paper Poli, Michael, Stefano Massaroli, Eric Nguyen, Daniel Y. Fu, Tri Dao, Stephen Baccus, Yoshua Bengio, Stefano Ermon, and Christopher Ré. “Hyena Hierarchy: Towards Larger Convolutional Language Models.” arXiv, April 19, 2023. https://doi.org/10.48550/arXiv.2302.10866.

  • Avsec, Žiga, Vikram Agarwal, Daniel Visentin, Joseph R. Ledsam, Agnieszka Grabska-Barwinska, Kyle R. Taylor, Yannis Assael, John Jumper, Pushmeet Kohli, and David R. Kelley. “Effective Gene Expression Prediction from Sequence by Integrating Long-Range Interactions.” Nature Methods, (October 2021) - Enformer - a deep learning model (transformers) to predict epigenetic and gene expression profiles (128bp resolution) from human and mouse cell types using only the DNA sequence as input, incorporating information from up to 100 kb on either side of the target locus (200kb of input DNA sequence). Transformers, allow to increase the receptive field up to 100kb, in contrast to 20kb for Basenji2 or ExPecto. Significant increase in performance. Predictions using CAGE data. Improves variant effect prediction on eQTL data. Excellent transformer description.

  • Foundation models in bioinformatics, language, vision, graph, multimodel. Specific models, tools, briefly described, organized in categories (Figures, tables). Genes, proteomics, drug discovery, cell clustering, multi-omics integration.

    Paper Guo, Fei, Renchu Guan, Yaohang Li, et al. “Foundation Models in Bioinformatics.” National Science Review 12, no. 4 (2025): nwaf028. https://doi.org/10.1093/nsr/nwaf028.
  • Deep learning for genomics. Network architectures - CNN, RNN, autoencoders, hybrid architectures, transformer-based LLMs. Model interpretation - saliency maps, attention scores. Transfer, multitask, multi-view learning. Applications - prediction of gene expression, promoters, enhancers, splicing, DNA and RNA binding sites, mutation effect, subcellular locations. Proteins, structural classification, prediction, secondary, tertiary structure.
    Paper Yue, Tianwei, Yuanxin Wang, Longxiang Zhang, et al. “Deep Learning for Genomics: A Concise Overview.” arXiv:1802.00810. Preprint, arXiv, October 4, 2023. https://doi.org/10.48550/arXiv.1802.00810.

DL Tools

  • Interactive_Tools - Interactive Tools for Machine Learning, Deep Learning and Math. Play with deep neural network in browser

  • ivy - The Unified Machine Learning Framework supporting JAX, TensorFlow, PyTorch, MXNet, and Numpy. Python module. Documentation

  • keras - Deep Learning for humans http://keras.io/

  • MXNet-Gluon-Style-Transfer - neural artistic style transfer using MXNet. PyTorch and Torch implementations available

  • openai.com - GPT-3 Access Without the Wait (API access to GPT-3)

  • OpenCV - Open Source Computer Vision library. GitHub, opencv-python - CPU-only OpenCV packages for Python. Documentation. Video - 3h OpenCV crash course

  • pathology_learning - Using traditional machine learning and deep learning methods to predict stuff from TCGA pathology slides

  • ruta - Unsupervised Deep Architechtures in R, autoencoders. Requires Keras and TensorFlow. Book

  • tensor2tensor - Library of deep learning models and datasets designed to make deep learning more accessible and accelerate ML research

  • Janggu - deep learning interface to genomic data (FASTA, BAM, BigWig, BED, GFF). Numpy-like Bioseq and Cover objects accessable by Keras. Includes model evaluation and interpretation features. Pypi, Docs, Janggu - Deep learning for genomics

  • maui - Multi-omics Autoencoder Integration. Latent factors from different data types (stacked variational autoencoders), and their clustering, testing for association with survival. Tested vs. latent factors extracted using Multifactor Analysis (MFA) and iCluster+, on TCGA colorectal cancer RNA-seq, SNPs, CNVs. Evaluation of Colorectal Cancer Subtypes and Cell Lines Using Deep Learning

  • Ludwig is a toolbox built on top of TensorFlow that allows to train and test deep learning models without the need to write code. GitHub

  • Mask_RCNN - Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow

  • PennAI - AI-Driven Data Science, entry-level machine learning interface for non-experts. A System for Accessible Artificial Intelligence

Auto ML

DL models

  • PyTorch model zoos

  • Keras model zoos

  • Tensorflow model zoos

  • Kipoi - a model zoo for genomics. Examples of transfer learning, predicting pathogenic variants, TFBSs. Avsec et al., “The Kipoi Repository Accelerates Community Exchange and Reuse of Predictive Models for Genomics.”, GitHub repo

  • Caffe Model Zoo

  • BERT, Bidirectional Encoder Representations from Transformers, for natural language processing tasks. Model architecture, implemented using TensorFlow. Applications - Masked Language Model, next sentence prediction. Excels in several benchmarks. Pretrained models and code. See also BioBert

  • DNABERT - pre-trained Bidirectional Encoder Representation from Transformers (BERT) model for DNA sequence, captures contextual information, 512bp window. Input - overlapping k-mer tokens (6-mer perform best). Architecture details (12 transformer encoder layers, learning rate increased then decreased, 120 epochs sufficient), comparison with RNN, CNN, Transformers. Outperforms DeePromoter, PromID, FPROM, FirstEF promoter prediction methods, DeepBind, DeepSEA, Basset, DeepSite, DanQ, DESSO on on TFBS prediction. Pretrained on 10-510-long sequences, to predict missing nucleotides. Fine-tuned for promoter region prediction, transcription factor binding sites, splicing sites, genetic variants. visualization of learned attention patterns. [CLS] classification token at the beginning of sequence, [SEP] at the end. Visualization of attention scores. Supplementary material contains detailed description of the model's architecture, input data, training, evaluation, various case scenarios, visualization. Bringing BERT to the field blog post contains more details and links to transformer resources.

    Paper Ji, Yanrong, Zhihan Zhou, Han Liu, and Ramana V Davuluri. “[DNABERT: Pre-Trained Bidirectional Encoder Representations from Transformers Model for DNA-Language in Genome](https://doi.org/10.1093/bioinformatics/btab083).” Bioinformatics, (August 9, 2021)
  • TabPFN - the Tabular Prior-data Fitted Network, a tabular foundational model (generative transformer-based), outperforms other models, excels on small datasets but scalable, fast. Classification or regression problems Leverages in-context learning (ICL), generates a large corpus of synthetic tabular datasets (prior in Bayesian terms, approach based on structural causal models, SCMs) and then train a self-attention-based transformer neural network. Pre-training - predict masked targets of all synthetic datasets. Z-normalization of inputs. Sample-specific learning, feature-specific attention attending to features across samples, invariant to the order of samples and features. Tested on two benchmark datasets, AutoML benchmark and OpenML-CTR23, outperforms XGBoost, CatBoost, random forest. Approaches for interpretation. Python implementation.
    Paper Hollmann, N., Müller, S., Purucker, L. et al. Accurate predictions on small data with a tabular foundation model. Nature 637, 319–326 (2025). https://doi.org/10.1038/s41586-024-08328-6

  • folding_tools - a collection of protein folding tools

  • progressive_growing_of_gans - Progressive Growing of GANs for Improved Quality, Stability, and Variation — Official TensorFlow implementation of the ICLR 2018 paper

DL projects

Language models

ChatGPT, Gemini, NotebookLM, Claude, DeepSeek, Together.ai, Kimi, Qwen, Grok, OpenRouter, Groq, Storm, [Perplexity.ai, Suna, MiniMax, Stackoverflow.ai

  • awesome-chatgpt - Curated list of awesome tools, demos, docs for ChatGPT and GPT-3

  • awesome-llm-courses - A curated list of awesome online courses about Large Langage Models (LLMs)

  • agenticSeek - A 100% local alternative to Manus AI, this voice-enabled AI assistant autonomously browses the web, writes code, and plans tasks while keeping all data on your device. Tailored for local reasoning models, it runs entirely on your hardware, ensuring complete privacy and zero cloud dependency.

  • chatbox - User-friendly Desktop Client App for AI Models/LLMs (GPT, Claude, Gemini, Ollama...)

  • chatgpt-clone - Build Yo'own ChatGPT with OpenAI API & Gradio. A Python app for web browser intercage to ChatGPT.

  • DeepSeek-V3 - a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token. Excellent performance at a fraction of the cost. Web chatbot.

  • Emergent - an LLM model for website/app creation. Uses Claude, ChatGPT, connects to GitHub

  • h2ogpt - open-source GPT with document and image Q&A, 100% private chat, no data leaks, Apache 2.0 https://arxiv.org/pdf/2306.08161.pdf Live Demo: https://gpt.h2o.ai/

  • Hands-On-Large-Language-Models - Official code repo for the O'Reilly Book - "Hands-On Large Language Models" by Jay Alammar and Maarten Grootendorst. PyTorch, Jupyter notebooks, can be opened in Google colab.

  • litellm - Python SDK, Proxy Server (LLM Gateway) to call 100+ LLM APIs in OpenAI format - [Bedrock, Azure, OpenAI, VertexAI, Cohere, Anthropic, Sagemaker, HuggingFace, Replicate, Groq]. Docs

  • llm-course - Course to get into Large Language Models (LLMs) with roadmaps and Colab notebooks.

  • llm_engineering - repo for the Mastering AI and LLM Engineering course. Python, includes videos.

  • LLMs-from-scratch - Implementing a ChatGPT-like LLM from scratch, step by step, the Python code for coding, pretraining, and finetuning a GPT-like LLM. By Sebastian Raschka. For the book Build a Large Language Model (From Scratch).

  • LLMsPracticalGuide - A curated list of practical guide resources of LLMs (LLMs Tree, Examples, Papers)

  • lobe-chat - an open-source, modern-design AI chat framework. Supports Multi AI Providers( OpenAI / Claude 3 / Gemini / Ollama / Azure / DeepSeek), Knowledge Base (file upload / knowledge management / RAG ), Multi-Modals (Vision/TTS/Plugins/Artifacts). One-click FREE deployment of your private ChatGPT/ Claude application.

  • mlc-llm - Enable everyone to develop, optimize and deploy AI models natively on everyone's devices. Documentation

  • nanoGPT - The simplest, fastest repository for training/finetuning medium-sized GPTs.

  • ollama - Get up and running with Llama 2 and other large language models locally.

  • openai-cookbook - Examples and guides for using the OpenAI API. Rendered version

  • privateGPT - Interact privately with your documents using the power of GPT, 100% privately, no data leaks.

  • prompt-eng-interactive-tutorial - Anthropic's Interactive Prompt Engineering Tutorial

  • Qwen - LLM created by Alibaba Cloud. Transformer-based, lightweight but among top performers. Website

  • Storm - An LLM-powered knowledge curation system that researches a topic and generates a full-length report with citations. GitHub

Audio, music, voice

  • abogen - Generate audiobooks from EPUBs, PDFs and text with synchronized captions.

  • buzz - Buzz transcribes and translates audio offline on your personal computer. Powered by OpenAI's Whisper.

  • ebook2audiobook - Convert ebooks to audiobooks with chapters and metadata using dynamic AI models and voice cloning. Supports 1,107+ languages!

  • ElevenReader - text to speech converter, for ebooks (EPUB), PDFs, text files. iOS and Android mobile apps.

  • Flow Music - Google’s AI-powered music creation platform for generation of full songs from prompts, creating AI music videos, remix/split audio, build music tools, and publish/share.

  • insanely-fast-whisper - An opinionated CLI to transcribe Audio files w/ Whisper on-device! Powered by Transformers, Optimum & flash-attn. Transcribe 150 minutes (2.5 hours) of audio in less than 98 seconds - with OpenAI's Whisper Large v3. Blazingly fast transcription is now a reality!

  • Jukebox - music generation neural network. Hierarchical Vector Quantised-Variational AutoEncoder (VQ-VAE) architecture, three separate temporal resolutions. Able to generate singing from lyrics, extend music examples. Dhariwal et al., “Jukebox: A Generative Model for Music.”, Blog post with examples of generated music

  • june - Local voice chatbot for engaging conversations, powered by Ollama, Hugging Face Transformers, and Coqui TTS Toolkit

  • Magenta - Music and Art Generation with Machine Intelligence

  • MINIMAX - text-to-speech, music creation. Voice clone, voice design, noise removal.

  • OpenVoice - voice cloning tool, transfer voice tones to pronounce different words, even in different language.

  • Real-Time-Voice-Cloning - Clone a voice in 5 seconds to generate arbitrary speech in real-time. Learn voice characteristics from a short audio clip and perform text-to-speech conversion using this voice.

  • Project DeepSpeech - A TensorFlow implementation of Baidu's DeepSpeech architecture. Transcribe audio data, English model available. Documentation

  • SpeechBrain - A PyTorch-based Speech Toolkit for speech/speaker recognition, speech enhancement, processing, and more. GitHub repo

  • Unmute (Kyutai) - open source text-to-speech and speech-to-text model, GitHub.

  • vampnet - music generation with masked transformers. arXiv paper, supplementary page

  • VibeVoice - VibeVoice is a novel framework designed for generating expressive, long-form, multi-speaker conversational audio, such as podcasts, from text.

  • voicebox - The open-source voice synthesis studio powered by Qwen3-TTS.

Image, video, vision

  • PaperBanana - Automating Academic Illustration for AI Scientists. Visual language model trained on NeurIPS diagrams. GitHub

  • Krea - image and video generation from text prompt, enhancing image quality, image editing from text prompt.

  • Sora - OpenAI's text-to-video generation tool that allows users to create, animate, and remix videos from text prompts, images, and existing videos.

  • 18 All-Time Classic Open Source Computer Vision Projects for Beginners by Analytics Vidhya

  • 500-AI-Machine-learning-Deep-learning-Computer-vision-NLP-Projects-with-code - links to many ML/DL projects and resources

  • Top 10 Computer Vision Papers 2020 - A list of the top 10 computer vision papers in 2020 with video demos, articles, code and paper reference.

  • awesome-ai-art-image-synthesis - A list of awesome tools, ideas, prompt engineering tools, colabs, models, and helpers for the prompt designer playing with aiArt and image synthesis. Covers Dalle2, MidJourney, StableDiffusion, and open source tools.

  • Bringing-Old-Photos-Back-to-Life - Old Photo Restoration (Official PyTorch Implementation)

  • CLIP - CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image

  • CogView2 - official code repo for paper "CogView2: Faster and Better Text-to-Image Generation via Hierarchical Transformers" (ArXiv), Example

  • Cosmos - Cosmos is a world model development platform that consists of world foundation models, tokenizers and video processing pipeline to accelerate the development of Physical AI at Robotics & AV labs. Cosmos is purpose built for physical AI. The Cosmos repository will enable end users to run the Cosmos models, run inference scripts and generate videos. Docker options.

  • HunyuanVideo - A Systematic Framework For Large Video Generation Model

  • DALL-E-2 - a new AI system that can create realistic images and art from a description in natural language. Tweet. How This A.I. Draws Anything You Describe 16m video by ColdFusion. HOW DALL-E COULD POWER A CREATIVE REVOLUTION, The Verge. DALLE2-pytorch - Implementation of DALL-E 2, OpenAI's updated text-to-image synthesis neural network, in Pytorch. min-dalle - min(DALL·E) is a fast, minimal port of DALL·E Mini to PyTorch

  • Deep-Live-Cam - real time face swap and one-click video deepfake with only a single image.

  • JoJoGAN - Official PyTorch repo for JoJoGAN: One Shot Face Stylization. Arxiv

  • Photo restoration with GFP-GAN - GFPGAN aims at developing Practical Algorithms for Real-world Face Restoration. Online version

  • Reve - image editing model

  • seedream - advanced image generation/editing model

  • Selfie2Anime online tool and a GitHub repo

  • stablediffusion - High-Resolution Image Synthesis with Latent Diffusion Models

  • VGG - Visual Geometry Group. Large source of image datasets. Also, practicals, demos, software, publications, more.

DL Misc

  • 100_days_of_CUDA - notes about learning CUDA

  • geospy.ai - location identification from photos

  • app.wombo.art - deep generative model dreaming awesome images from text, Android and iOS apps available. Tweet describing the VQGAN+CLIP technology behind it

  • CSrankings - A web app for ranking computer science departments according to their research output in selective venues, and for finding active faculty across a wide range of areas. Website

  • ColossalAI - A Unified Deep Learning System for Big Model Era. Scaling deep learning models using data, pipeline, tensor, and sequence parallelism. 1D, 2D, 2.5D, 3D distributed operators. Examples of each. Written in PyTorch, needs a configuration file defining parallelism. Benchmarked against DeepSpeed, Megatron-LM.

    Paper Li, Shenggui, Jiarui Fang, Zhengda Bian, Hongxin Liu, Yuliang Liu, Haichen Huang, Boxiang Wang, and Yang You. “Colossal-AI: A Unified Deep Learning System For Large-Scale Parallel Training,” n.d.

Awesome Machine learning

ML Books

ML Courses & Tutorials

ML Videos

ML Papers

  • Whalen, Sean, Jacob Schreiber, William S. Noble, and Katherine S. Pollard. “Navigating the Pitfalls of Applying Machine Learning in Genomics.” Nature Reviews Genetics 23, no. 3 (March 2022): 169–81. https://doi.org/10.1038/s41576-021-00434-9. - Five machine learning problems in genomics, distributional differences, dependency structure, confounding variables, information leakage, unbalanced data. Description, examples, solutions.

  • Domingos, Pedro. “A Few Useful Things to Know about Machine Learning.” Communications of the ACM 55, no. 10 (October 1, 2012): 78. https://doi.org/10.1145/2347736.2347755. Twelve lessons for machine learning. Overview of machine learning problems and algorithms, problem of overfitting, causes and solutions, curse of dimensionality, issues with high-dimensional data, feature engineering, bagging, boosting, stacking, model sparsity. Video lectures

ML Tools

  • mlr3 - Machine learning in R R package, the unified interface to classification, regression, survival analysis, and other machine learning tasks. GitHub repo, mlr3gallery - Examples of problems and code solutions, mlr3 Manual - mlr3 bookdown. More on the mlr3 package site, including videos

ML Misc

Material in Chinese

  • Autopilot-Notes - Autonomous driving notes summarizing the basics, hardware, perception, position, planning, control, product, tools, and manufacturing plan topics.

Material in Russian

  • GigaChat - Russian language-focused language model. Based on the mixture of experts (MoE) architecture, GPT-3, 13B parameters. 2 min Video demo. Leaderboard. Alternatives: YandexGPT5, Cotype. GigaAM for speech processing. Models on Huggingface. @gigachat_bot Telegram bot.
    Paper team, GigaChat, Mamedov Valentin, Evgenii Kosarev, Gregory Leleytner, Ilya Shchuckin, Valeriy Berezovskiy, Daniil Smirnov, et al. “GigaChat Family: Efficient Russian Language Modeling Through Mixture of Experts Architecture.” arXiv, June 11, 2025. https://doi.org/10.48550/arXiv.2506.09440.

  • Scientific_graphics_in_python - matplotlib for scientific graphics. 3 parts, 13 chapters. By Pavel Shabanov

  • ml-course-hse - machine learning course at the Computer Sciences Department, High Schoool of Economy. Multiple years, videos

  • mlcourse_open - OpenDataScience Machine Learning course (Both in English and Russian). Python-based ML course, with video lectures. Video

  • DL_CSHSE_spring2018 - Deep learning, Anton Osokin, Higher School of Economics, Computer Sciences Department (Russian), course material, and video lectures

  • Ordinary Differential Equations - Обыкновенные дифференциальные уравнения, Интерактивный учебник, Илья Щуров (НИУ ВШЭ)

  • Calculus - Математический анализ, Записки лекций, Илья Щуров (НИУ ВШЭ). Tweet

  • mathprofi.ru - Высшая математика – просто и доступно. Mirror

deep-learning
machine-learning

Similar repos

peggy1502/Amazing-Resources

List of references and online resources related to data science, machine learning and deep learning.

435

mikeroyal/Machine-Learning-Guide

Machine learning Guide. Learn all about Machine Learning Tools, Libraries, Frameworks, Large…

710

kandeldeepak46/Keep-Learning

Resources and blogs, curated for learning, computer science, mathematics, data science, computer…

26

ashara12/awesome-deeplearning

A curated list of awesome Deep Learning tutorials, projects and communities.

29,040

ChristosChristofidis/awesome-deep-learning

A curated list of awesome Deep Learning tutorials, projects and communities.

29,040

imteekay/machine-learning-research

✨ AI/ML Research

458

aurimas13/Machine-Learning-Goodness

The Machine Learning project including ML/DL projects, notebooks, cheat codes of ML/DL, useful…

286

guillaume-chevalier/Awesome-Deep-Learning-Resources

Rough list of my favorite deep learning resources, useful for revisiting topics or for reference. I…

1,821