일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | |||
5 | 6 | 7 | 8 | 9 | 10 | 11 |
12 | 13 | 14 | 15 | 16 | 17 | 18 |
19 | 20 | 21 | 22 | 23 | 24 | 25 |
26 | 27 | 28 | 29 | 30 | 31 |
- mPLUG-2
- pretrained model layer
- 특정 layer 동결
- 논문 작성
- MLLM
- def train
- Video Understanding
- 가중치 없이 모델 로드
- mqtt
- 논문 작성 요령
- 주식 용어정리
- 주식
- DeepNet
- Foundation Transformers
- 모델 구조 변경
- 가상환경
- 특정 layer 추출
- def validation
- Instruction dataset
- 모델 동결
- 모델 freeze
- KOSMOS-2
- 3C4P
- cnn
- 파라미터 수
- 파라미터 수 확인
- 강화학습
- 논문리뷰
- Multimodal Large Language Model
- layer 추출
- Today
- Total
목록인공지능/딥러닝 사이드 Project (10)
시작은 미약하였으나 , 그 끝은 창대하리라
1. 필요 모듈 설치 ➢ 중요한 모듈만 언급- torch (https://pytorch.org/get-started/previous-versions/) local 컴퓨터 버전에 맞게 알맞게 설치하기..# CUDA 11.7conda install pytorch==1.13.0 torchvision==0.14.0 torchaudio==0.13.0 pytorch-cuda=11.7 -c pytorch -c nvidia- transformerspip install transformers- pytorch-ligtningpip install pytorch-lightning 2. 데이터 셋 준비- 주의사항: DETR 의 데이터셋 format은 COCO format 으로 준비되어야 함. (다운 받은 데이터 셋의 ..
❏ 목적: 제스쳐를 통한 제어 (맨 아래 github 들어가면 모든 코드 확인가능함) ❏ Mediapipe Gesture Recognition 간단설명 이미지로 손을 학습하나 생각할 수 있으나 그게아니였다.. 손에 대한 관절 포인트(= landmark) 21개를 이미 다 학습을 시켜놓아서 손을 보면 실시간으로 손의 좌표(x,y,z) 따주는것임. where, x는 가로, y 세로, z 는 카메라로 부터의 거리. 학습에 사용되는 feature = landmark data[21 *3] + landmark visibility[21]+ 손가락 각도 데이터[15개] =총 99개 ❏ 모델 : CNN- LSTM 데이터를 수집을 많이 한것도 아니고 모델이 깊은것도 아닌데. 생각보다 잘되더라..( 카메라 각도와 손의 각도..
논문: https://arxiv.org/abs/2010.11929 참고 깃허브: https://github.com/FrancescoSaverioZuppichini/ViT/tree/main 필자 깃허브( 코드 다운) : https://github.com/YongTaeIn/ViT ❏ 아래의 ppt와 같은 구조로 모델이 수행된다. ➢사실상 Transformer모델을 간파하고 있다면 초기의 input에서 patch로 변환하는것 이외에는 색다른 부분이 없다. ➢ Patching +CLS Token+ embedding # patch embedding # Add CLS Token # position embedding import torch import torch.nn.functional as F import matp..
참고글: https://huggingface.co/blog/autoformer Autoformer 공식 github: https://github.com/thuml/Autoformer 논문: https://arxiv.org/pdf/2106.13008.pdf 필자의 Autoformer_encdoer를 이용한 시계열 예측 github: https://github.com/YongTaeIn/Autoformer_encoder_time_series/tree/master ✓ Auto Former 는 크게 두가지 기술이 들어간다. -> 그리고 아래 두가지 기술에 대해 적용하는 방법을 설명할 것이다. 1. Series Decomposition (시계열 분해) 2 Attention layer - > Auto Correlat..
목표: 시계열 데이터를 트랜스포머 인코더를 이용하여 예측해보자. 전체코드: https://github.com/YongTaeIn/Transformer-Encoder-model-for-time-series.git 0. Input shape 의 변환 순서 1 번과정, embedding = [batch size, time_seq, feature_dim] 2번 과정, positional Encoding =[batch size, time_seq, d_model] 3번과정~4번과정Encoder blcok= [batch size, time_seq, d_model] 5번과정, FFW =[batch size, time_seq, 본인설정] 6번과정 ,차원축소=[batch size, time_seq] 7번과정, Dense =..
이미지 데이터 출처: https://www.kaggle.com/datasets/pranavraikokte/covid19-image-dataset Covid-19 Image Dataset 3 Way Classification - COVID-19, Viral Pneumonia, Normal www.kaggle.com 1.모듈 불러옴 # impot module import tensorflow as tf import matplotlib.pyplot as plt #이미지 확인용 모듈 import matplotlib.image as mpimg #이미지 확인용 모듈 import os import numpy as np from tensorflow import keras from tensorflow.keras.model..
1) Yolov5를 설치함. (참고 https://github.com/ultralytics/yolov5) git clone https://github.com/ultralytics/yolov5 # clone cd yolov5 pip install -r requirements.txt # install module 설치하고 나면 사진과 같은 폴더및 파일들이 생겨날것이다. 2) (핵심) Image detection 하고싶은 데이터 수집 (나의 경우 포트홀,맨홀, 크랙 등등을 수집) (라벨링 툴 :https://github.com/heartexlabs/labelImg) -1) train 폴더 생성후 images와 labels폴더를 생성한다. -2) 라벨링 툴을 이용하여 ' 클래스 넘버와, x1, y1, x2, y..
(들아가기전에-> 예시코드를 이용해 시계열 데이터를 비교하며, 커널(필터)이 어떻게 작동되는지 써놓은 글 입니다.) (많은 예시와 코드들이 있지만 시계열 데이터 와 아닌 것에는 엄연히 코드의 차이가 있다(데이터 가공하는 부분에 있어서 다름 )) 1. 모듈 불러옴. import numpy as np import pandas as pd import tensorflow as tf from keras.models import Sequential from keras.layers import Dense,Conv1D,Flatten from tensorflow import keras from sklearn.datasets import load_iris 2.seed 설정(하이퍼 파라미터 바꿀시 다른 값들 안바뀌도록) #..