Implementation of the RemBG library

A tiny model for testing out Cog

BLIP3 is a series of foundational Large Multimodal Models (LMMs) developed by Salesforce AI Research

Blip 3 (blip3-phi3-mini-instruct-r-v1), Answers questions about images

viⓍTTS vixTTS là mô hình tạo sinh giọng nói cho phép bạn sao chép giọng nói sang các ngôn ngữ khác nhau chỉ bằng cách sử dụng một đoạn âm thanh nhanh dài 6 giây

Transcribe audios using OpenAI's Whisper with stabilizing timestamps by stable-ts python package.

Auto-magically relights your images

Use a face to make images. Uses SDXL fine-tuned checkpoints.

Use a face to instantly make images. Uses SDXL Lightning checkpoints.

Run any ComfyUI workflow. Guide: https://github.com/fofr/cog-comfyui

Cog to turn minimally-formatted plaintext into pdfs (using tex on the backend)

Dark Sushi Mix 2.25D Model with vae-ft-mse-840000-ema (Text2Img, Img2Img and Inpainting)

DeepSeek LLM, an advanced language model comprising 67 billion parameters. Trained from scratch on a vast dataset of 2 trillion tokens in both English and Chinese

turns text into pdf files with TeX

Meta Llama Guard 2 is an 8B parameter Llama 3-based LLM safeguard model

Hermes 2 Pro is an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house

Virtual try-on using Stable Diffusion and IP-Adapter

a fine-tuned model to detect dragon in images.

InstantID. ControlNets. More base SDXL models. And the latest ByteDance's ⚡️SDXL-Lightning !⚡️

The img2img pipeline that makes an anime-style image of a person. It uses one of sd1.5 models as a base, depth-estimation as a ControleNet and IPadapter model for face consistency.

GPU accelerated replay renderer / video data clipper for comma.ai connect's openpilot route data. SEE README.

Consistent Self-Attention for Long-Range Image and Video Generation

Robust face restoration algorithm for old photos / AI-generated faces (adapted to work with video inputs)

📖 PuLID: Pure and Lightning ID Customization via Contrastive Alignment

Semantic Segmentation

A SDXL Model trained from another SDXL-hiroshinagai model images

Train SDXL 1.0 with LoRA | mixed precision bf16 and save precision fp16

Just some good ole beautifulsoup scrapping URL magic. (some sites don't work as they block scrapping, but still useful)

High resolution image Upscaler and Enhancer. Use at ClarityAI.cc. A free Magnific alternative. Twitter/X: @philz1337x

Realistic Inpainting with ControlNET (M-LSD + SEG)

Tango2: LLM-guided Diffusion-based Text-to-Audio Generation and DPO-based Alignment

🗣️ TalkNet-ASD: Detect who is speaking in a video

Transfer a material from an image to a subject

Demucs is an audio source separator created by Facebook Research.

Creates voxels like game asset

Projection module trained to add vision capabilties to Llama 3 using SigLIP