🦙 llama38bbatch1model-merged

This is a fine-tuned LLaMA-3 8B model using QLoRA + PEFT + SFTTrainer, trained on a custom dataset (final_batch1, 2, and 3.csv) by @hemanthvamsikrishna.

---

📚 Training Details

• Base Model: meta-llama/Meta-Llama-3-8B
• Fine-tuning Method: QLoRA (4-bit), PEFT (LoRA)
• Trainer: trl.SFTTrainer
• Epochs: 3
• Batch Size: 4 (with gradient accumulation = 4)
• Max Sequence Length: 512
• LoRA Targets: ['q_proj', 'k_proj', 'v_proj', 'o_proj', 'gate_proj', 'up_proj', 'down_proj']

---

📦 Files Included

• ✅ Sharded model weights (pytorch_model-*.bin)
• ✅ Tokenizer & Config files
• ✅ LoRA merged with base weights

---

🧪 Example Usage

from transformers import pipeline

pipe = pipeline("text-generation", model="hemanthvamsikrishna/llama38bbatch1model-merged")
output = pipe("What is the future of AI?", max_new_tokens=100)
print(output[0]['generated_text'])