Gemma2_Virtual_doctor

Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->

Model Details

Model Description

<!-- Provide a longer summary of what this model is. -->

This model is fined tune based on Google's Gemma model for creating virtual doctor or medical Asistant. It can be used in medical and healthcare AI assitant apps and chatbots.

• Developed by: [Ali Bidaran]

Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig, GemmaTokenizer

model_id = "alibidaran/Gemma2_Virtual_doctor"
bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_compute_dtype=torch.bfloat16
)


tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=bnb_config, device_map={"":0})

prompt = " Hi doctor, I feel a pain on my ankle, I walk hardly and with pain what do you recommend me?"
text=f"<s> ###Human: {prompt} ###Asistant: "
inputs=tokenizer(text,return_tensors='pt').to('cuda')
with torch.no_grad():
    outputs=model.generate(**inputs,max_new_tokens=200,do_sample=True,top_p=0.92,top_k=10,temperature=0.7)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training Parameters

    per_device_train_batch_size=1,
    gradient_accumulation_steps=8,
    warmup_steps=2,
    #max_steps=200,
   
    num_train_epochs=1,
    learning_rate=2e-4,
    fp16=True,
    logging_steps=100,
    output_dir="outputs",
    optim="paged_adamw_8bit",
    save_steps=500,
    ddp_find_unused_parameters=False // for training on multiple GPU