AI/ML Hub

Welcome to the UCR Research Computing AI and ML Hub, your gateway to exploring artificial intelligence and machine learning resources. Dive into the latest technologies, tools, and platforms designed to empower your research and innovation.

How to run your own copy of a open source LLM on UCR Ursa Major or your workstation right now! How To run offline LLMs see it in action Here

Leading AI chatbots for your research needs.

Warning: Ensure you have the necessary permissions or take the appropriate steps before using AI chatbots with sensitive or proprietary data.

Chatbot Description
Groq Offers lightning-fast computations, ideal for complex ML models.
Perplexity Excels in understanding and generating human-like text, perfect for NLP tasks.
Gemini Best for creative content generation, including writing, art, and music composition.
ChatGPT Known for its versatility in engaging in detailed conversations and providing informative responses.
Claude Stands out for its deep contextual understanding, making it great for nuanced dialogues.
Pi.ai Pi.ai is a personal AI assistant offering empathetic conversations, factual information, and emotional support to users, with a user-friendly interface accessible on various platforms.

AI & ML Software Tools for Researchers

Leverage powerful software tools designed to harness the full potential of HPC clusters for AI and ML tasks, across various domains.

Note: Installation and usage of some HPC tools may require advanced configuration. Contact us! research-computing@ucr.edu

General Purpose and Cross-Domain Tools

Software Use Case Example Command
TensorFlow Deep Learning tensorflow_model_server --port=8500 --model_name=my_model --model_base_path=/models/my_model
PyTorch Machine Learning Models python -m torch.distributed.launch --nproc_per_node=NUM_GPUS_YOU_HAVE my_train_script.py
Horovod Distributed Training horovodrun -np 4 -H localhost:4 python train.py
CUDA Parallel Computing nvcc my_cuda_program.cu -o my_cuda_program
Hugging Face Transformers Natural Language Processing transformers-cli convert --model_type bert --tf_checkpoint /path/to/checkpoint --config /path/to/bert_config.json --pytorch_dump_output /path/to/pytorch_model.bin

Life Sciences & Biology

Software Use Case Example Command
AlphaFold Protein Structure Prediction alphafold --fasta_path=my_sequence.fasta --output_dir=output
FreeSurfer Neuroimaging Analysis recon-all -s subject_name -i input_volume.nii -all

Earth Science

Software Use Case Example Command
GDAL Geospatial Data Processing gdal_translate -of GTiff input.tif output.tif
Pangeo Climate and Ocean Science Example usage involves deploying Pangeo on cloud platforms for scalable data processing.

Chemistry & Physics

Software Use Case Example Command
GROMACS Molecular Dynamics gmx mdrun -s topol.tpr -deffnm output
Quantum ESPRESSO Materials Modeling pw.x -in Si.scf.in > Si.scf.out

Humanities & Social Sciences

Software Use Use Case Example Command
SpaCy NLP & Text Processing SpaCy is optimized for performance, with capabilities for syntactic analysis, tokenization, and entity recognition.
Gensim Topic Modeling & Similarity Gensim is specialized in unsupervised topic modeling and natural language document similarity.
TextBlob Simplified Text Processing TextBlob simplifies text processing, providing an intuitive interface for common NLP tasks like sentiment analysis.
NLTK Natural Language Toolkit NLTK is a leading platform for building Python programs to work with human language data, offering interfaces to over 50 corpora and lexical resources.

Top Open Source AI Models

Explore a handpicked selection of top open-source AI models that are driving innovation across various domains. From sophisticated text generation to dynamic text-to-image models, these contributions represent the cutting edge of AI research and application.

How to run your own copy of a open source LLM on UCR Ursa Major or your workstation right now! How To run offline LLMs

Name Description Name Description
google/gemma-7b Text Generation ByteDance/SDXL-Lightning Text-to-Image
bigcode/starcoder2-15b Text Generation playgroundai/playground-v2.5-1024px-aesthetic Text-to-Image
google/gemma-7b-it Text Generation mistralai/Mixtral-8x7B-Instruct-v0.1 Text Generation
google/gemma-2b Text Generation stabilityai/stable-video-diffusion-img2vid-xt Image-to-Video
stabilityai/stable-cascade Text-to-Image meta-llama/Llama-2-7b-chat-hf Text Generation

AI Models by Domains and Applications

This section outlines critical areas within AI, including natural language processing, audio processing, computer vision, and multimodal systems. Discover the transformative power of AI through these domains, each linking to foundational tasks and state-of-the-art models.

Natural Language Processing Audio Processing Computer Vision Multimodal
Text classification Audio classification Image classification Image captioning
Token classification Automatic speech recognition Image segmentation Document Question Answering
Question answering   Video classification Visual Question Answering
Causal language modeling   Object detection Text to speech
Masked language modeling   Zero-shot object detection  
TranslationSummarization   Zero-shot image classification  
Multiple choice   Depth estimation  
    Image-to-Image  
    Mask Generation  
    Knowledge Distillation for Computer Vision  

Large Language Model Development Tools

Tool Name Description Usage in AI Development
Flowise AI A software platform for building LLM applications on custom data, demonstrated to answer questions from PDFs. Enables the creation and deployment of LLM applications on various data sources.
LangChain A framework for building applications around language models, simplifying embedding, tokenization, and more. Facilitates the creation and deployment of LLM-based applications.
Vector Stores Solutions for storing and retrieving vectors, essential for models like RAG. Enhances the efficiency of retrieval-augmented generation models.
Embedding Tools Tools for creating vector embeddings of textual data. Used in processing text for NLP models and improving model understanding.
Tokenization Libraries Libraries that support the conversion of text into tokens, suitable for language models. Essential for preparing data for training and fine-tuning AI models.
Fine-Tuning & Hosting Tools for customizing pre-trained models and hosting them for inference. Allows for the customization and deployment of AI models for specific tasks.

Tip: Experiment with different development tools to find the best fit for your project’s needs. Innovation often comes from trying out new approaches.


Other Online AI Tools

RAG Enabled Tools

Tool Name Description Application in Research
Notebook LM An AI-first notebook designed to help users summarize information, understand complex ideas, and brainstorm new connections. Ideal for synthesizing facts and generating document guides.
Other Tool Description of the tool How it assists in specific research tasks

Image Generation

Explore the latest in AI-powered image generation, where these tools leverage deep learning models to transform textual descriptions into vivid, high-resolution images. From generating unique artworks to designing specific visuals, these platforms enable an unprecedented level of creative expression.

AI Tool Application Description
DALL·E 3 Image Generation Generates high-quality images from textual descriptions, offering unparalleled creative freedom.
Stable Diffusion Image Synthesis State-of-the-art image synthesis from user inputs, perfect for creating detailed and imaginative visuals.
MidJourney Artistic Image Creation Creates artistic images and visualizations based on textual descriptions, blending art and AI.

Video Generation

The realm of video generation is being transformed by AI, making it possible to produce engaging and dynamic videos from simple descriptions. These tools are revolutionizing content creation, offering new ways to create educational, entertaining, and promotional videos with minimal effort.

AI Tool Application Description
Kaiber AI Video Generation Transforms textual descriptions into engaging videos, simplifying content creation.
DeepBrain AI Video Synthesis Produces lifelike AI-generated videos, including virtual human content, for a range of applications.
OpenAI’s Sora Video Generation (Upcoming) Promises to revolutionize video creation with AI, generating videos from descriptions.
Haiper AI Video Creation Develops AI products for video creation, enabling users to express themselves creatively on social media.

Music Generation

AI is also making waves in the music industry, enabling creators to generate unique compositions, experiment with sound, and even innovate in the realm of music production. These tools provide a new avenue for musical exploration and creativity.

AI Tool Application Description
Suno.ai Music Generation Empowers users to create unique music compositions using AI, making music creation accessible to all.
Google’s Magenta Music & Art Innovation A research project exploring the role of AI in the process of creating art and music.
Google Labs’ Music FX Music Effects & Tools Offers a suite of experimental music tools and effects designed to expand creative possibilities through AI.

Voice Cloning and Synthesis

Voice cloning and synthesis technologies are breaking new ground, offering tools to create realistic and customizable voiceovers and audio content. From enhancing accessibility to enabling more natural and engaging user experiences, these tools are at the forefront of audio innovation.

AI Tool Application Description
ElevenLabs Voice Cloning & Synthesis Offers revolutionary voice cloning and synthesis capabilities, ideal for creating lifelike audio from text.

These LLM resources are designed to push the boundaries of digital creativity and audio production, providing researchers, artists, and developers with powerful tools to explore new realms of audio-visual expression. Whether you’re looking to generate unique musical pieces or clone voices for various applications, these platforms offer the technology and flexibility to bring your visions to life.