src/train.py [196:210]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    loss = loss / opts.accumulation_steps

    # Perform backward pass
    loss.backward()
    
    # Clip gradient norms and get (clipped) gradient norms for logging
    grad_norms = clip_grad_norms(optimizer.param_groups, opts.max_grad_norm)
    
    # Perform optimization step after accumulating gradients
    if step % opts.accumulation_steps == 0:
        optimizer.step()
        optimizer.zero_grad()

    # Logging
    if step % int(opts.log_step) == 0:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



src/train.py [291:305]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    loss = loss / opts.accumulation_steps
    
    # Perform backward pass
    loss.backward()
    
    # Clip gradient norms and get (clipped) gradient norms for logging
    grad_norms = clip_grad_norms(optimizer.param_groups, opts.max_grad_norm)
    
    # Perform optimization step after accumulating gradients
    if step % opts.accumulation_steps == 0:
        optimizer.step()
        optimizer.zero_grad()

    # Logging
    if step % int(opts.log_step) == 0:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -