Skip to content

Latest commit

 

History

2 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 

Repository files navigation

HTGM v2 Hindi LLM – Session 16 | 145 Hours Training Progress

Overview

Session 16 of HTGM v2 Hindi LLM training is now completed.

This session focused on stable checkpoint continuation, stronger validation improvement, and pushing the model beyond the important 37K+ optimizer steps milestone.

After 145+ total training hours, the model is showing better consistency, lower validation loss, improved perplexity, and stronger long-session reliability.

This repository documents the real progress of building a Hindi LLM from scratch using a GPT-style Transformer architecture.


Session Information

  • Session Number: 16
  • Total Training Time: 145+ Hours
  • Total Sessions Completed: 16
  • Major Milestone: 37K+ Optimizer Steps Crossed

Technical Details

  • Model Name: HTGM v2
  • Type: Hindi Large Language Model (Hindi LLM)
  • Architecture: GPT-style Transformer
  • Model Size: 163.4M Parameters
  • Sequence Length: 2048
  • Vocabulary Size: 100,000
  • Dataset Size: ~41GB
  • Dataset Source: AI4Bharat Sangrah + Manual Cleaning
  • Training Platform: Kaggle
  • GPU: 2x T4
  • Tokenizer: Hugging Face BPE-based Tokenizer

Training Progress

  • Total Planned Optimizer Steps: 923,382
  • Resumed From Checkpoint Step: 1,118,962
  • Current Training Step: 37,512+
  • Mid Checkpoints Saved: ckpt_s35000.pt, ckpt_s36000.pt, ckpt_s37000.pt
  • Final Checkpoint Saved: ckpt_final_s1200400.pt

Dataset Used

Loaded Chunks

  • chunk_1.txt
  • chunk_2.txt
  • chunk_3.txt

Training Metrics

Core Metrics

  • Early Session Loss: 3.7838
  • Mid Training Loss: 4.8249
  • Later Training Loss: 4.8023
  • Latest Observed Loss: 5.0656
  • Final Validation Loss: 4.9147
  • Final Validation Perplexity (PPL): 136.27

Optimization Metrics

  • Learning Rate: 7.63e-05
  • Initial Learning Rate: 9.90e-07
  • Tokens Per Second: Stable throughout training

Checkpoint Strategy

Training resumed from a previous checkpoint instead of restarting from scratch.

Benefits

  • Safe long training workflow
  • Stable continuation across sessions
  • Hyperparameter testing without losing progress
  • Reliable optimizer and scheduler recovery

This makes long Hindi LLM experiments practical and efficient.


Observations

What Improved

  • Validation loss improved to 4.91
  • Perplexity improved to 136
  • Stronger checkpoint reliability
  • GPU performance stayed strong
  • No major crashes
  • W&B monitoring active
  • Smooth checkpoint saving

Why It Matters

Crossing 37K+ steps is important, but improving validation quality matters even more.

This session proves that better optimization is more valuable than simply training longer.


Current Status

  • Training is stable
  • Validation is improving
  • Model learning quality is getting better
  • Long-session reliability is strong

Still not production-ready, but the foundation is becoming much stronger.


Next Steps

  • Apply Supervised Fine-Tuning (SFT)
  • Improve instruction-following capability
  • Better dataset structuring
  • Continue longer training sessions
  • Improve response quality and coherence

Author

Mahesh Editor
India AI Official


Build in Public

This project is shared publicly to document real AI development, not just final results.

Every checkpoint, problem, fix, and improvement is part of the journey.

About

Building HTGM v2 — a Hindi LLM from scratch using GPT architecture. Real training logs, checkpoints, experiments, and progress shared publicly.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors