main.py

from transformers import AutoModelForCausalLM, AutoTokenizer
from flask import Flask, request, jsonify

device = "cuda" # the device to load the model onto
from transformers import set_cache_dir
set_cache_dir("/code/.cache/huggingface")

model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.1")
tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.1")



@app.route('/recommend', methods=['POST'])
def recommendation():
    content = request.json
    user_degree = content.get('degree')
    user_stream = content.get('stream')
    user_semester = content.get('semester')
    messages = [
    {"role": "user", "content": f"""
    You need to act like as recommendataion engine for course recommendation based on below details.

    Degree: {user_degree}
    Stream: {user_stream}
    Current Semester: {user_semester}


    Based on above details recommend the courses that realtes to above details 
    Note: Output should bevalid json format in below format:
    {{"course1:ABC,course2:DEF,course3:XYZ,...}}
    
    """},
    
    ]
    
    encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")
    
    model_inputs = encodeds.to(device)
    model.to(device)
    
    generated_ids = model.generate(model_inputs, max_new_tokens=1000, do_sample=True)
    decoded = tokenizer.batch_decode(generated_ids)
    return jsonify({"res":decoded[0]})

if __name__ == '__main__':
    app.run(debug=True)