GitHub - Joshua-Immanuel/Building-LLM-from-scratch: Explored tokenisation, gpt2's byte pair encoding technique, data loader and dataset from pytorch, token embeddings and positional embeddings

Name		Name	Last commit message	Last commit date
Latest commit History 17 Commits
124M_parameters_GPT_2.ipynb		124M_parameters_GPT_2.ipynb
Attention_mechanism.ipynb		Attention_mechanism.ipynb
Classification finetuning data loading.ipynb		Classification finetuning data loading.ipynb
Data_Preparation_stage_1_Tokenisation_.ipynb		Data_Preparation_stage_1_Tokenisation_.ipynb
Final_finetuning_project.ipynb		Final_finetuning_project.ipynb
FineTuning_LLM_for_spam_classifier.ipynb		FineTuning_LLM_for_spam_classifier.ipynb
LayerNormalization+GELU+FeedForward.ipynb		LayerNormalization+GELU+FeedForward.ipynb
Loading_Data_+_Training_+_Evaluation.ipynb		Loading_Data_+_Training_+_Evaluation.ipynb
Loading_GPT_2_weights.ipynb		Loading_GPT_2_weights.ipynb
Temperature_scaling_+_top_k_Sampling_.ipynb		Temperature_scaling_+_top_k_Sampling_.ipynb
Transformer_Block.ipynb		Transformer_Block.ipynb
finetune_deepseek_coder_v2_lora.py		finetune_deepseek_coder_v2_lora.py
finetune_qwen25_coder_lora.py		finetune_qwen25_coder_lora.py
finetuning_app.py		finetuning_app.py
gpt_download3.py		gpt_download3.py
inference.ipynb		inference.ipynb
theverdict.txt		theverdict.txt