Paper Replications

Paper Replication Collection

Collection of my re-implementation of seminal AI/ML papers from scratch in PyTorch.

Filter by category:
ViT

ViT

From scratch implementation of ViT
GPT

GPT

From scratch implementation of GPT
BERT

BERT

From scratch implementation of BERT
CycleGANs

CycleGANs

From scratch implementation of CycleGANs
Differential Transformer

Differential Transformer

From scratch implementation of Differential Transformer
Pix2Pix

Pix2Pix

From scratch implementation of Pix2Pix
SigLip

SigLip

From scratch implementation of SigLip
WGANs

WGANs

From scratch implementation of WGANs
Fine Tuning using PEFT

Fine Tuning using PEFT

From scratch implementation of Fine Tuning using PEFT
ORPO

ORPO

From scratch implementation of ORPO
GRU

GRU

From scratch implementation of GRU
Attention Mechanisms

Attention Mechanisms

From scratch implementation of Attention Mechanisms
Encoder-Decoder

Encoder-Decoder

From scratch implementation of Encoder-Decoder
RNNs

RNNs

From scratch implementation of RNNs
Transformer

Transformer

From scratch implementation of Transformer
Mixtral

Mixtral

From scratch implementation of Mixtral
TTS

TTS

From scratch implementation of TTS
Moonshine

Moonshine

From scratch implementation of Moonshine
DPO

DPO

From scratch implementation of DPO
SimplePO

SimplePO

From scratch implementation of SimplePO
LoRA

LoRA

From scratch implementation of LoRA
Gemma

Gemma

From scratch implementation of Gemma
Gemma3

Gemma3

From scratch implementation of Gemma3
Llama

Llama

From scratch implementation of Llama
Llama4

Llama4

From scratch implementation of Llama4
PaliGemma

PaliGemma

From scratch implementation of PaliGemma
CLiP

CLiP

From scratch implementation of CLiP
DDP

DDP

From scratch implementation of DDP
Llava

Llava

From scratch implementation of Llava
Seq2Seq

Seq2Seq

From scratch implementation of Seq2Seq
Whisper

Whisper

From scratch implementation of Whisper
LSTM

LSTM

From scratch implementation of lstm
VAE

VAE

From scratch implementation of VAE
CGANs

CGANs

From scratch implementation of CGANs
CLAP

CLAP

From scratch implementation of CLAP
DCGANs

DCGANs

From scratch implementation of DCGANs
DeepSeekV3

DeepSeekV3

From scratch implementation of DeepSeekV3
Kimi-K2

Kimi-K2

From scratch implementation of Kimi-K2