Update README.md

README.md

@@ -20,126 +20,102 @@ tags:
 Fine-tuning a Small Language Model (SLM) for Step-by-Step Math Reasoning  
 
 ## Overview  
-OpenMath is an open-source project focused on fine-tuning a small language model for math reasoning using QLoRA (4-bit LoRA).  
+OpenMath is an open-source project focused on fine-tuning a small language model for mathematical reasoning using parameter-efficient LoRA training.  
 
-This repository contains only a LoRA adapter trained on GSM8K. Users must load the base model separately and attach the adapter.  
+This repository contains **only a LoRA adapter** trained on the full GSM8K dataset. Users must load the base model separately and attach the adapter using PEFT.  
 
-The latest version of this model was trained on an AMD MI300X GPU using ROCm, showing that modern non-NVIDIA accelerators can successfully support large-scale fine-tuning with Hugging Face and PyTorch.  
+The latest version of this model was trained on an **AMD MI300X GPU using ROCm**, demonstrating that high-performance non-NVIDIA accelerators can successfully support modern large language model fine-tuning with PyTorch and Hugging Face.  
 
 ---
 
 ## Base Model  
-Qwen/Qwen2.5-Math-1.5B  
+**Qwen/Qwen2.5-Math-1.5B**  
 
-This repository does not contain the base model weights — they must be loaded from Hugging Face.  
+This repository **does not contain the base model weights** — they must be loaded directly from Hugging Face before applying this LoRA adapter.  
 
 ---
 
 ## Hardware Used (Latest Training Run)  
 
-GPU: AMD MI300X (ROCm 7.0)  
-VRAM: 192 GB  
-Operating System: Ubuntu 24.04  
-Framework: PyTorch + Hugging Face  
-Backend: ROCm  
+- **GPU:** AMD MI300X (ROCm 7.0)  
+- **VRAM:** 192 GB  
+- **OS:** Ubuntu 24.04  
+- **Framework:** PyTorch + Hugging Face  
+- **Backend:** ROCm  
 
 ---
 
 ## Dataset  
 
-GSM8K (Grade School Math 8K)  
-Training samples: 1,000  
-Evaluation: Full GSM8K test split (1,319 problems)  
+**GSM8K (Grade School Math 8K)**  
+- **Training samples:** 7,473 (full training split)  
+- **Evaluation:** Full GSM8K test split (1,319 problems)  
 
-Only the solution portion of each example was used for loss computation through loss masking.  
+Only the solution portion of each example was used for loss computation via loss masking to encourage stronger reasoning behavior.  
 
 ---
 
 ## Training Configuration  
 
-## Training Configuration (MI300X Run)
-
 **Method:** LoRA (full precision, bfloat16)  
-**Precision:** bfloat16 (no 4-bit quantization)  
+**Precision:** bfloat16 (no 4-bit quantization in this run)  
 
-**LoRA settings**
+### LoRA settings  
 - Rank: 16  
 - Alpha: 32  
 - Dropout: 0.05  
 - Target modules: `q_proj`, `k_proj`, `v_proj`, `o_proj`  
 
-**Data & sequence**
+### Data & sequence  
 - Max sequence length: 1024  
 
-**Optimization**
-- Batch size: 2  
+### Optimization  
+- Per-device batch size: 2  
 - Gradient accumulation: 8  
-- **Effective batch size:** 16  
+- Effective batch size: 16  
 - Learning rate: 1e-4  
 - Optimizer: `adamw_torch`  
 - Scheduler: cosine  
 - Warmup: 5%  
 
-**Training**
-- Epochs: 3  
-
+### Training  
+- **Epochs:** 3  
 
 ---
 
 ## Results  
 
-GSM8K Accuracy (Full Test Set):  
-750 out of 1319 correct, which equals 56.86 percent accuracy.  
+**GSM8K Accuracy (Full Test Set):**  
+750 / 1319 = **56.86% accuracy**  
 
-This is significantly stronger than the earlier Colab T4 run and is a strong result for a 1.5B model trained with LoRA.  
-
----
-
-## What This Repository Contains  
-
-adapter_model.safetensors — LoRA weights  
-adapter_config.json — LoRA configuration  
-chat_template.jinja — chat formatting template  
-tokenizer.json — tokenizer file  
-tokenizer_config.json — tokenizer settings  
-README.md — documentation  
-
-This repository does not include checkpoints, optimizer states, or full base model weights.  
+This represents a substantial improvement over earlier small-scale Colab experiments and is a strong result for a 1.5B model trained with LoRA on the full dataset.  
 
 ---
 
 ## How to Use This Model  
 
-Load the base model Qwen/Qwen2.5-Math-1.5B from Hugging Face, then attach this LoRA adapter using PEFT. Generate answers using a prompt that includes an instruction, problem, and solution section.  
+1. Load the base model **Qwen/Qwen2.5-Math-1.5B** from Hugging Face.  
+2. Attach this LoRA adapter using PEFT.  
+3. Use a structured prompt that includes an instruction, problem, and solution section for best results.  
 
 ---
 
 ## Why This Matters  
 
-This project demonstrates that AMD MI300X can train modern language models with Hugging Face and QLoRA.  
-It shows that high-quality math reasoning is possible at 1.5B parameters using efficient fine-tuning.  
-It provides a lightweight adapter instead of requiring users to download a massive full model.  
+- Demonstrates that **AMD MI300X** can effectively train modern LLMs with Hugging Face + LoRA.  
+- Shows strong math reasoning at **1.5B parameters** with lightweight fine-tuning.  
+- Provides a compact adapter instead of requiring users to download a massive full model.  
 
 ---
 
 ## Limitations  
 
-The model can make reasoning mistakes.  
-It should not be used for exams, assignments, or professional decisions.  
-Performance depends heavily on prompt formatting.  
-
----
-
-## Future Work  
-
-Train on 3,000 to 5,000 GSM8K samples.  
-Add SVAMP and ASDiv datasets.  
-Improve decoding to reduce repetition.  
-Experiment with multi-GPU scaling on MI300X.  
-Add a Streamlit demo for interactive use.  
+- The model can make reasoning mistakes.  
+- It should not be used for exams, assignments, or professional decisions.  
+- Performance depends heavily on prompt formatting.  
 
 ---
 
 ## License  
 
-cc-by-nc-4.0
+**cc-by-nc-4.0**