reporesolutefoldermain

This is a PEFT Multilingual fine-tuned version of meta-llama/Llama-2-7b-hf using Bhagwad Gita verses.

Model Details

• Base Model: meta-llama/Llama-2-7b-hf
• Fine-tuning Method: PEFT (Parameter Efficient Fine-Tuning)
• Task: Translation
• Language: Hindi/Sanskrit to English

Directory Structure

├── adapter_model.safetensors    # Main model weights
├── adapter_config.json          # Model configuration
├── all_results.json            # Training results
├── train_results.json          # Detailed training metrics
├── trainer_state.json          # Trainer state
└── checkpoint/                 # Checkpoint directory
    ├── training_args.bin       # Training arguments
    ├── optimizer.pt            # Optimizer state
    ├── scheduler.pt            # Scheduler state
    ├── adapter_model.safetensors
    ├── adapter_config.json
    ├── rng_state.pth          # Random state
    └── trainer_state.json

Usage

from peft import AutoPeftModelForCausalLM
from transformers import AutoTokenizer
model = AutoPeftModelForCausalLM.from_pretrained("aneeshm44/reporesolutefoldermain")
tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-2-7b-hf")