Maidari Tsydenov's Projects
A Unified Library for Parameter-Efficient and Modular Transfer Learning
This repository holds all the code for the site http://www.adventuresinmachinelearning.com
Π¨Π°Π±Π»ΠΎΠ½Ρ Π·Π°Π΄Π°Ρ Ρ Π²Π²ΠΎΠ΄ΠΎΠΌ-Π²ΡΠ²ΠΎΠ΄ΠΎΠΌ Π΄Π»Ρ Π²Π²ΠΎΠ΄Π½ΠΎΠΉ ΡΠ°ΡΡΠΈ ΠΈ ΠΏΠ΅ΡΠ²ΠΎΠ³ΠΎ ΡΠΏΡΠΈΠ½ΡΠ° ΠΊΡΡΡΠ° ΠΠ»Π³ΠΎΡΠΈΡΠΌΡ Π΄Π»Ρ ΡΠ°Π·ΡΠ°Π±ΠΎΡΡΠΈΠΊΠΎΠ².
This repo includes ChatGPT prompt curation to use ChatGPT better.
Distilled variant of Whisper for speech recognition. 6x faster, 50% smaller, within 1% word error rate.
Apache Hadoop docker image
Foodgram project template
Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.
infra_sp2
Implementation of the LLaMA language model based on nanoGPT. Supports flash attention, Int8 and GPTQ 4bit quantization, LoRA and LLaMA-Adapter fine-tuning, pre-training. Apache 2.0-licensed.
Fine-tuning LLaMA to follow Instructions within 1 Hour and 1.2M Parameters
Config files for my GitHub profile.
Telegram bot for user interaction with chatgpt, dalle, salute-speech services, etc.
Inferencing on Mistral 7B LLM with 4-bit quantization
Π Π΅ΡΠ΅Π½ΠΈΠ΅ ΠΊΠ΅ΠΉΡΠ° ΠΎΡ Π ΠΠ "ΠΠ½ΡΠ΅Π»Π»Π΅ΠΊΡΡΠ°Π»ΡΠ½ΡΠΉ Π³ΠΎΠ»ΠΎΡΠΎΠ²ΠΎΠΉ ΠΏΠΎΠΌΠΎΡΠ½ΠΈΠΊ ΠΌΠ°ΡΠΈΠ½ΠΈΡΡΠ°". Π₯Π°ΠΊΠ°ΡΠΎΠ½ "Π¦ΠΈΡΡΠΎΠ²ΠΎΠΉ ΠΏΡΠΎΡΡΠ² 2023 ΠΠΈΠΆΠ½ΠΈΠΉ Π½ΠΎΠ²Π³ΠΎΡΠΎΠ΄"
rag for a fictional company
Russian GPT3 models.
This bot filters comments from a YouTube broadcast and sends them to Telegram.
Silero Models: pre-trained speech-to-text, text-to-speech and text-enhancement models made embarrassingly simple
Π’Π΅ΡΡΠΎΠ²ΠΎΠ΅ Π΄Π»Ρ ΡΠΌΠ΅Π½Ρ Π² ΡΠΈΡΠΈΡΡ 2023
High-Resolution Image Synthesis with Latent Diffusion Models
StableDiffusionTelegram is a telegram bot that allows to generate images using the stable diffusion AI from a telegram bot, in a much more comfortable and simple way.
A Gradio web UI for Large Language Models. Supports transformers, GPTQ, llama.cpp (GGUF), Llama models.
Thin wrapper around OpenAI Whisper API with streaming support
ΠΠΎΠΌΠ°Π½Π΄Π½ΡΠΉ ΠΏΡΠΎΠ΅ΠΊΡ "YaMDb" - ΡΠ΅ΡΠ²ΠΈΡ, ΠΏΠΎΠ·Π²ΠΎΠ»ΡΡΡΠΈΠΉ ΡΠΈΡΠ°ΡΡ ΠΈ Π΄ΠΎΠ±Π°Π²Π»ΡΡΡ ΠΎΡΠ·ΡΠ²Ρ ΠΎ ΡΠ°Π·Π»ΠΈΡΠ½ΡΡ
ΠΏΡΠΎΠΈΠ·Π²Π΅Π΄Π΅Π½ΠΈΡΡ
ΠΈΡΠΊΡΡΡΡΠ²Π°: ΠΊΠ½ΠΈΠ³Π°Ρ
, ΡΠΈΠ»ΡΠΌΠ°Ρ
, ΠΌΡΠ·ΡΠΊΠ΅ ΠΈ Π½Π΅ ΡΠΎΠ»ΡΠΊΠΎ