Added experiments and fixes some timings

Added gradient monitoring logs

Added experiments, attention, network, index-layer, gradient-monitoring, calculation and implementation documentation

Added missing changelog

Reordering attentions and networks

Removed conditional architectural flow

Fixed naming conventions

Fixed Comments

Added expirements, built skeleton for dynamic tokenizer

Added experiments, fixed attention code

Added experiments

Added experiments and matrixes

Added expiriments, float type and temperature options

Aligned Expirement

Added expirements, addded library information, enhanced expirement visual management and evaluation

Added expirements, Moved configs, fixed naming conventions, added c_shuffle and r_split

Added expirements and minor fix

Centralized functions and added image and video

Fixed configs, attentions and networks information

Fixed blank charts

Added optimizer warmup and fixed naming conventions

Added expirements and attention, network comment step matching

Added cuda handling, added flops estimation, added expirements, fixed model nullabilities

Added expirements, tokenizers, descriptions, charts, materials, attentions, networks and more

Fixed naming conventions, added networks, abstracted model import and export

Refactored code, added devices, attentions, networks

Added expirement

Fixed Version

Moved config files into Hugging Face Repo

Added Multiple Architectures

Fixed epoch and batch report prints

Switched model file extension from weights to json

Refactored the code, added tokenizer class, added finetune option, added inference cache, removed python loops

Enhanched readme

Enhanched outputs information

Enhanced runtime naming conventions

Enhanced runtime variables and functions

Recreated main branch