For example i expect temperature values between 20-30°C, but I’m getting negative numbers or very large values.

So collect data of humidity and temperature with dht11 and train it with Tansorflow and convert it to tflite and am using tflite micro

Did you quantize your model?

What’s that, sorry am new to this!

Try this collab code for the quantization

# Google Colab script to quantize a model to int8
import tensorflow as tf
import numpy as np

# Load model
model = tf.keras.models.load_model("your_model.h5")

# Representative dataset
def representative_dataset():
    for _ in range(100):
        yield [np.random.rand(1, 28, 28, 1).astype(np.float32)]

# Convert model
converter = tf.lite.TFLiteConverter.from_keras_model(model)
converter.optimizations = [tf.lite.Optimize.DEFAULT]
converter.representative_dataset = representative_dataset
converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8]
converter.inference_input_type = tf.int8
converter.inference_output_type = tf.int8

tflite_model = converter.convert()

# Save model
with open("model_int8.tflite", "wb") as f:
    f.write(tflite_model)

print("Quantized model saved as model_int8.tflite")

I appreciate it @wafa_ath

Am still getting wrong output

Not sure how to do that

Mean and standard deviation…
I have two input data one for temperature and one for humidity should i apply on both of them

Yes for both

Mean= total data/member of the total data

Ow okay okay i get it , i will try that

What do you think, did try to use it and compare it to other AI?

Yes i tried it but the server is always busy, i know that faced open AI from the beginning too and it will be fixed with time, but what i don’t understand that why openAI didn’t used automated reinforcement learning to reduce the cost too or there is a catch !