George Grigorev

facemesh.pytorch. This is the PyTorch implementation of paper Real-time Facial Surface Geometry from Monocular Video on Mobile GPUs (https://arxiv.org/pdf/1907.06724.pdf)

sample_efficient_gpt. Training framework with a goal to explore the frontier of sample efficiency of small language models

null-text-inversion. Unofficial implementation of paper NULL-text Inversion for Editing Real Images using Guided Diffusion Models ( https://arxiv.org/abs/2211.09794 )

cs336_solutions. Here's my solutions to all assignments of Stanford CS336 course: LLM from Scratch

OLMo. My fork os allen AI's OLMo for educational purposes.

rvc-trainer. Jupyter Notebook

qwen2_1_5b_summarize.

age_gender_classifier. EDA and training procedure for multitask age and gender classification

Qwen3-GRPO-text-reversal. Text reversal task using Qwen3-Base-0.6B

tacotron2_improved. This is my reimplementation of Tacotron2 based on nvidia implementation

Tasks. Differents tasks I did in Python

llm-cli. Python

tacotron2. My implementation of tacotron2 model

TA_analyzer. Target Audience VK Public analyzer

llava_utils. One-file utils file that helps to access official LLaVA repository

sagan_pytorch. Self-Attention Generative Adversarial Networks implementation using this paper: https://arxiv.org/abs/1805.08318v1