Projects in Awesome Lists by genji970

https://github.com/genji970/llm_api_service_deploying_in_aws

collecting pdf data by using ray. fine tuning pretrained model gpt2. building rest api, using fine tuned model. deploying in aws. aws 배포용 llm 서비스 파이프라인 구축

api-rest aws gpt llm service

Last synced: 04 Mar 2025

https://github.com/genji970/robot_simulation_with_llm_gpt2_reinforcement_learning

env, programming language : c++ , ubuntu22 on virtualbox, aim is to build robot by ros2 and simulate this robot in a gazebo env,further plan is to merge llm code pipeline and rl to this system

ai llm rl robotics ros2-humble

Last synced: 04 Mar 2025

https://github.com/genji970/text_to_image_3d-reconstruction

using u net, depth estimation -> 2d to 3d image(3d 프로젝트)

Last synced: 04 Mar 2025

https://github.com/genji970/text_to_3d_point_cloud_image_and_classification

writing prompt -> loaded diffusion model outputs 2d image -> using midas depth estimation, making depth image -> making point cloud data with (image,depth image) -> using loaded pointnet, doing object detection to ply data

Last synced: 04 Mar 2025

https://github.com/genji970/llm-rl-dl-paper-review-coding

논문 리뷰 , 구현

Last synced: 04 Mar 2025

https://github.com/genji970/llm_project_from_data_process_to_inference

project covers from data_process ~ inference . 프로젝트(데이터 처리부터 추론까지)

Last synced: 04 Mar 2025

https://github.com/genji970/reinforcement_learning_mujoco_half_cheetah_with_two_stage_methods_to_llm_lora_finetuning_sac_tuning

To solve half cheeath v4 by using two stage methods, first lora fine tuning to LLM and second do soft actor critic to LLM(in progress)

Last synced: 04 Mar 2025

https://github.com/genji970/online-transformer_encoder_rl-mujoco-half-cheetah-reward-change-experiment-bad-teacher-model

This experiment follows three steps. collecting data from normal rl env. training custome transformer model(for bad teacher model) with collected data trajectory. training new rl agent with trained transformer model in normal rl env(in progress)

Last synced: 04 Mar 2025

https://github.com/genji970/llm_rl_fine_tuning_for_solving_rl_problem

This repo follows three steps. First, do actor critic method and get trajectory. Second, Put these trajectory to pretrained llm, meta llama for lora fine tuning. Third, with fine tuned model, do rl training

Last synced: 04 Mar 2025

https://github.com/genji970/2024-june-3-llm-paper-contrastive-preference-optimization--review-implementation

contrastive preference optimization: pushing the boundaries of LLM performance in machine translation

Last synced: 04 Mar 2025

https://github.com/genji970/llm_with_quantization_deepspeed

Last synced: 04 Mar 2025

https://github.com/genji970/llm-rl-actor-critic

for code record

Last synced: 04 Mar 2025

https://github.com/genji970/aws_distributed_learning_llm_code

in progress

Last synced: 04 Mar 2025

https://github.com/genji970/llm-rl-portfolio

자연어 강화학습 포트폴리오

Last synced: 04 Mar 2025

https://github.com/genji970/ai_stock_prediction_with_convex_transformation_dnn

This repo contains one paper and result of experiment. By simply transform the form of data in the candle chart, can get about 12% profit for each stock in usa_market(for 9 stocks) and about 13% profit for each 39 stocks in kospi market.

Last synced: 04 Mar 2025

https://github.com/genji970/vision-paper-study

study

Last synced: 04 Mar 2025

https://github.com/genji970/llm_thought_record

llm 생각, 느낀점 정리

Last synced: 04 Mar 2025

https://github.com/genji970/vision-kaggle-competition

Competition code I used

Last synced: 04 Mar 2025

https://github.com/genji970/llm_pipeline_from_training_to_deploying

pipeline consist of two stage. first, pdf data collecting using ray, fine tuning(llm training) pipeline. second, deploying in aws using rest api. for opensource contributions

api contributions-welcome finetuning good-first-issue llm

Last synced: 04 Mar 2025

https://github.com/genji970/covid_detection_by_reconstructing_2d_to_3d_image_using_depth_estimation

mostly annotations are written in korean. Detecting covid using 3d , depth estimation, U net, diffusion. 3d 프로젝트용

Last synced: 04 Mar 2025

https://github.com/genji970/chatbot_reduction-in-execution-time_with-reference-to-paper-enhancing-robustness-in-llm-

Chatbot_Reduction in execution time_with reference to paper "Enhancing Robustness in Large Language Models : Prompting for Mitigating the Impact of Irrelevant Information"

Last synced: 04 Mar 2025

https://github.com/genji970/distributed_learning_for_rl_mujoco_cheetah_by_ray

using ray, multi process learning for half cheetah v4. ray를 사용한 half cheetah v4 분산학습

Last synced: 04 Mar 2025

https://github.com/genji970/recent-llm-paper-review-coding-contrastive-preference-optimization-

llm paper contrastive preference opt code implementation in colab env for data prerprocessing and model fine tuning parts. colab 환경에서 llm 논문 데이터 전처리 ~ model 파인튜닝까지의 과정

Last synced: 04 Mar 2025