DAY 31

Notes

DLRM is a open source Deep Learning Recommendation Model introduced by Facebook for tensorflow and deeplearning.
NLP has
1. NLTK
2. SPACY
3. TEXT BLOB
4. Stemming
5. Lemmatization
6. Sentimental Analysis
fbcli -> a facebook command line interface for linux

pip3 install fbcli

the rate of transfer from a pendrive to HDD is depended upon the IO rate of pendrive which is based on the material used for making the HDD
GPU is required when higher graphical work is necessary
Python is an interpreted language
Namespace is the private space of different compilers over a same system RAM which is not accessible by other compilers
Provision Space - > the minimum amount of space that is required by some service

Make a combination of text,image/video,voice and then search for sentiment analysis

In deep learning the most important component is GPU and gym , a library of openai uses a lot of GPU for processing
it is a concept developed in deeplearning which interprets or calculates the requirements
Tensor is a dimension - with some value ( Tensor can 1D / 2D or 3D )
Tensorflow is a mixture of tensor and graph
this library is made for heavier and high level mathematical calculations
It can be though as SuperAdvanced Version of numpy
To install tensorflow

pip3 install tensorflow

to interact with gpu in system -- heavy software 377M and only for system with gpu

pip3 install tensorflow-gpu

Tensorflow has 3 categories for taking data as input
1. Constant -> defined by developer
2. Variables -> defined by user
3. Placeholder -> takes input from user at run time just like input function
Compilation - > checking the syntax of some file and creating a file stored in RAM to be initialsed anytime
Tensorflow is not python , its just supported by it
Its background is made in complete C++

* CNN
* poet -> for image classification 
* posenet  - > object detection(simplest in tensorflow )
* yolo