This model is a fine-tuned version of unsloth/llama-3-8b-bnb-4bit using the vicgalle/alpaca-gpt4 dataset.
Model Description
- Developed by: sujeetgund
- Model type: Causal Language Model
- Language(s) (NLP): English
- License: Apache 2.0
Training Details
This model was trained using Unsloth and Hugging Face's TRL library on a Google Colab Tesla T4 instance.
- Notebook Reference: View Training Notebook
- Epochs: 1 (60 steps used for demonstration)
- Learning Rate: 2e-4
- Optimizer: AdamW 8-bit
- LoRA Rank (r): 16
- LoRA Alpha: 16
How to use
You can load this model using unsloth for 2x faster inference:
from unsloth import FastLanguageModel
model, tokenizer = FastLanguageModel.from_pretrained(
model_name = "sujeetgund/llama_3_lora_model",
max_seq_length = 1024,
load_in_4bit = True,
)
FastLanguageModel.for_inference(model)
About Unsloth
Fine-tuned with Unsloth. Unsloth makes fine-tuning Llama-3, Mistral, and Gemma models 2x faster and use 70% less memory.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for sujeetgund/llama_3_lora_model
Base model
meta-llama/Meta-Llama-3-8B Quantized
unsloth/llama-3-8b-bnb-4bit