---
license: apache-2.0
datasets:
- BrainGPT/train_valid_split_pmc_neuroscience_2002-2022_filtered_subset
tags:
- text-generation-inference
- peft
---

## Training details:
We fine-tuned Llama2-7b-chat using LoRA. We used a batch size of 1 and a chunk size of 2048. Training involved the use of the AdamW optimizer with a learning rate of 2e-5 and gradient accumulation steps set at 8. A single training epoch was performed, along with a warm-up step of 0.03 and a weight decay rate of 0.001. The learning rate was controlled using a cosine learning rate scheduler. LoRA adapters, characterized by a rank of 8, an alpha value of 32, and a dropout rate of 0.1, were applied after all self-attention blocks and fully-connected layers. This results in total 17,891,328 trainable parameters, roughly 0.26% of the entire parameters of the base model. To optimize training performance, bf16 mixed precision training and data parallelism were employed. We used 4 Nvidia A100 (80GB) GPUs hosted on the Microsoft Azure platform. An epoch of training takes roughly 42 GPU hours.

## Training data:
Please refer to Dataset card: https://huggingface.co/datasets/BrainGPT/train_valid_split_pmc_neuroscience_2002-2022_filtered_subset

## Model weights:
The current version of BrainGPT was fine-tuned on Llama-2-7b-chat-hf with LoRA, `adapter_model.bin` contains the LoRA adapter weights. To load and use the full model, you need to be granted access to Llama-2-7b-chat-hf via https://huggingface.co/meta-llama/Llama-2-7b-chat-hf. 

## Load and use model:
```python
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM
from transformers import AutoTokenizer

config = PeftConfig.from_pretrained("BrainGPT/BrainGPT-7B-v0.1")

# Load model
model = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path)
model = PeftModel.from_pretrained(model, "BrainGPT/BrainGPT-7B-v0.1")

# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)

```