model/pytorch_model.bin

import torch
import torch.nn as nn
import torch.optim as optim
from torch.utils.data import DataLoader, Dataset
import json
import os

# Step 1: Define Your Dataset Class
class CustomDataset(Dataset):
    def __init__(self, texts, labels):
        self.texts = texts
        self.labels = labels

    def __len__(self):
        return len(self.texts)

    def __getitem__(self, idx):
        return self.texts[idx], self.labels[idx]

# Step 2: Define Your Model Class
class LSTMModel(nn.Module):
    def __init__(self, input_size, hidden_size, output_size):
        super(LSTMModel, self).__init__()
        self.lstm = nn.LSTM(input_size, hidden_size, batch_first=True)
        self.fc = nn.Linear(hidden_size, output_size)

    def forward(self, x):
        lstm_out, _ = self.lstm(x)
        out = self.fc(lstm_out[:, -1, :])  # Get the last time step output
        return out

# Step 3: Initialize Hyperparameters and Model
input_size = 100  # Example input size (e.g., embedding size)
hidden_size = 64  # Number of LSTM units
output_size = 10  # Number of output classes
num_epochs = 5
learning_rate = 0.001

# Initialize the model
model = LSTMModel(input_size, hidden_size, output_size)

# Step 4: Set Up Loss and Optimizer
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=learning_rate)

# Step 5: Sample Data (You would replace this with your actual data)
texts = torch.randn(100, 10, input_size)  # 100 samples, sequence length of 10
labels = torch.randint(0, output_size, (100,))  # 100 random labels

# Create a DataLoader
dataset = CustomDataset(texts, labels)
data_loader = DataLoader(dataset, batch_size=16, shuffle=True)

# Step 6: Training Loop
for epoch in range(num_epochs):
    for inputs, targets in data_loader:
        # Forward pass
        outputs = model(inputs)
        loss = criterion(outputs, targets)

        # Backward pass and optimization
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
    
    print(f'Epoch [{epoch+1}/{num_epochs}], Loss: {loss.item():.4f}')

# Step 7: Save the Model
model_save_path = "model"  # Change this to your desired path
os.makedirs(model_save_path, exist_ok=True)  # Create the directory if it doesn't exist

# Save the model weights as pytorch_model.bin
torch.save(model.state_dict(), os.path.join(model_save_path, "pytorch_model.bin"))

# Step 8: Create and Save the Configuration File
config = {
    "input_size": input_size,
    "hidden_size": hidden_size,
    "output_size": output_size,
    "num_layers": 1,  # Add more parameters as needed
    "dropout": 0.2
}

# Save the configuration to a JSON file
with open(os.path.join(model_save_path, "config.json"), "w") as f:
    json.dump(config, f)

print("Model and configuration saved successfully!")