# thirumurugan240/fine_tuning_concepts architecture

> Architecture diagram of the GitHub repository thirumurugan240/fine_tuning_concepts, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-02). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/thirumurugan240/fine_tuning_concepts
- Repository: https://github.com/thirumurugan240/fine_tuning_concepts

## Overview

This repository is a hands-on fine-tuning course, with notebooks teaching the workflow from model selection and data preparation through training, evaluation, inference, and serving. The intended artifact path is an adapter produced by notebook 11, used in notebooks 13–14, then merged for notebooks 15–16 and the FastAPI app. Notebook capabilities are inferred from filenames; the notebooks themselves were not excerpted, so other internal calls and transitions are not shown.

Main entry points: [`merged_model`](https://github.com/thirumurugan240/fine_tuning_concepts/tree/main/merged_model) (Merged Model), [`app.py`](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/app.py) (Generation API), [`finetuned_model`](https://github.com/thirumurugan240/fine_tuning_concepts/tree/main/finetuned_model) (LoRA Adapter), [`14_Merging_LoRA_and_Saving.ipynb`](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/14_Merging_LoRA_and_Saving.ipynb) (Adapter Merging)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_foundations["Fine-Tuning Foundations"]
  node_intro["Fine-Tuning Intro"]
  node_approaches["Method Comparison"]
  node_pipeline["Training Pipeline"]
  node_types["Fine-Tuning Types"]
  node_hardware["Hardware Setup"]
  node_base_model["Base Model Choice"]
end

subgraph group_preparation["Data Preparation"]
  node_dataset["Dataset Concepts"]
  node_dataset_creation["Dataset Creation"]
  node_tokenization["Tokenization Setup"]
end

subgraph group_training["Training and Evaluation"]
  node_lora["LoRA Configuration"]
  node_finetune["Model Fine-Tuning"]
  node_adapter["LoRA Adapter"]
  node_monitoring["Training Monitoring"]
  node_evaluation["Model Evaluation"]
  node_merge["Adapter Merging"]
end

subgraph group_serving["Inference and Serving"]
  node_merged_model["Merged Model"]
  node_inference["Model Inference"]
  node_api_tutorial["API Tutorial"]
  node_api["Generation API<br/>[app.py]"]
end

node_caller(("API Caller"))

node_finetune -->|"creates"| node_adapter
node_evaluation -->|"uses"| node_adapter
node_merge -->|"uses"| node_adapter
node_merge -->|"creates"| node_merged_model
node_inference -->|"uses"| node_merged_model
node_api_tutorial -->|"uses"| node_merged_model
node_api -->|"loads"| node_merged_model
node_caller -->|"requests generation"| node_api
node_api -->|"returns answer"| node_caller

click node_intro "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/01_Introduction_to_Fine_Tuning.ipynb"
click node_approaches "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/02_FineTuning_vs_Prompting_vs_RAG.ipynb"
click node_pipeline "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/03_LLM_Training_Pipeline.ipynb"
click node_types "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/04_Types_of_Fine_Tuning.ipynb"
click node_hardware "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/05_Hardware_and_Environment_Setup.ipynb"
click node_base_model "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/06_Choosing_the_Base_Model.ipynb"
click node_dataset "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/07_Understanding_Datasets.ipynb"
click node_dataset_creation "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/08_Create_Your_Own_Dataset.ipynb"
click node_tokenization "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/09_Tokenization_and_Chat_Templates.ipynb"
click node_lora "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/10_LoRA_QLoRA_Configuration.ipynb"
click node_finetune "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/11_Fine_Tuning_Your_First_LLM.ipynb"
click node_adapter "https://github.com/thirumurugan240/fine_tuning_concepts/tree/main/finetuned_model"
click node_monitoring "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/12_Monitoring_and_Overfitting.ipynb"
click node_evaluation "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/13_Evaluating_the_Model.ipynb"
click node_merge "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/14_Merging_LoRA_and_Saving.ipynb"
click node_merged_model "https://github.com/thirumurugan240/fine_tuning_concepts/tree/main/merged_model"
click node_inference "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/15_Running_Inference.ipynb"
click node_api_tutorial "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/16_Building_an_API_with_FastAPI.ipynb"
click node_api "https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/app.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_intro,node_approaches,node_pipeline,node_types,node_hardware,node_base_model toneBlue
class node_dataset,node_dataset_creation,node_tokenization toneAmber
class node_lora,node_finetune,node_adapter,node_monitoring,node_evaluation,node_merge,node_caller toneMint
class node_merged_model,node_inference,node_api_tutorial,node_api toneRose
```

## Components

### Fine-Tuning Foundations

- [Fine-Tuning Intro](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/01_Introduction_to_Fine_Tuning.ipynb): `01_Introduction_to_Fine_Tuning.ipynb`
- [Method Comparison](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/02_FineTuning_vs_Prompting_vs_RAG.ipynb): `02_FineTuning_vs_Prompting_vs_RAG.ipynb`
- [Training Pipeline](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/03_LLM_Training_Pipeline.ipynb): `03_LLM_Training_Pipeline.ipynb`
- [Fine-Tuning Types](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/04_Types_of_Fine_Tuning.ipynb): `04_Types_of_Fine_Tuning.ipynb`
- [Hardware Setup](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/05_Hardware_and_Environment_Setup.ipynb): `05_Hardware_and_Environment_Setup.ipynb`
- [Base Model Choice](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/06_Choosing_the_Base_Model.ipynb): `06_Choosing_the_Base_Model.ipynb`

### Data Preparation

- [Dataset Concepts](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/07_Understanding_Datasets.ipynb): `07_Understanding_Datasets.ipynb`
- [Dataset Creation](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/08_Create_Your_Own_Dataset.ipynb): `08_Create_Your_Own_Dataset.ipynb`
- [Tokenization Setup](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/09_Tokenization_and_Chat_Templates.ipynb): `09_Tokenization_and_Chat_Templates.ipynb`

### Training and Evaluation

- [LoRA Configuration](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/10_LoRA_QLoRA_Configuration.ipynb): `10_LoRA_QLoRA_Configuration.ipynb`
- [Model Fine-Tuning](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/11_Fine_Tuning_Your_First_LLM.ipynb): `11_Fine_Tuning_Your_First_LLM.ipynb`
- [LoRA Adapter](https://github.com/thirumurugan240/fine_tuning_concepts/tree/main/finetuned_model): `finetuned_model`
- [Training Monitoring](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/12_Monitoring_and_Overfitting.ipynb): `12_Monitoring_and_Overfitting.ipynb`
- [Model Evaluation](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/13_Evaluating_the_Model.ipynb): `13_Evaluating_the_Model.ipynb`
- [Adapter Merging](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/14_Merging_LoRA_and_Saving.ipynb): `14_Merging_LoRA_and_Saving.ipynb`

### Inference and Serving

- [Merged Model](https://github.com/thirumurugan240/fine_tuning_concepts/tree/main/merged_model): `merged_model`
- [Model Inference](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/15_Running_Inference.ipynb): `15_Running_Inference.ipynb`
- [API Tutorial](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/16_Building_an_API_with_FastAPI.ipynb): `16_Building_an_API_with_FastAPI.ipynb`
- [Generation API](https://github.com/thirumurugan240/fine_tuning_concepts/blob/main/app.py): `app.py`

### Other components

- **API Caller**

## Connections

- Model Fine-Tuning → LoRA Adapter: creates
- Model Evaluation → LoRA Adapter: uses
- Adapter Merging → LoRA Adapter: uses
- Adapter Merging → Merged Model: creates
- Model Inference → Merged Model: uses
- API Tutorial → Merged Model: uses
- Generation API → Merged Model: loads (evidence: `app.py`)
- API Caller → Generation API: requests generation (evidence: `app.py`)
- Generation API → API Caller: returns answer (evidence: `app.py`)
