Add Adafactor and Adahessian optimizers, cleanup optimizer arg passing, add gradient clipping support.

sotabench_setup.sh

@@ -7,7 +7,8 @@ pip install -r requirements-sotabench.txt
 apt-get update
 apt-get install -y libjpeg-dev zlib1g-dev libpng-dev libwebp-dev
 pip uninstall -y pillow
-CC="cc -mavx2" pip install -U --force-reinstall pillow-simd
+CFLAGS="${CFLAGS} -mavx2" pip install -U --no-cache-dir --force-reinstall --no-binary :all:--compile https://github.com/mrT23/pillow-simd/zipball/simd/7.0.x
+#CC="cc -mavx2" pip install -U --force-reinstall pillow-simd
 
 # FIXME this shouldn't be needed but sb dataset upload functionality doesn't seem to work
 apt-get install wget

timm/optim/__init__.py

@@ -1,10 +1,13 @@
-from .nadam import Nadam
-from .rmsprop_tf import RMSpropTF
+from .adamp import AdamP
 from .adamw import AdamW
-from .radam import RAdam
+from .adafactor import Adafactor
+from .adahessian import Adahessian
+from .lookahead import Lookahead
+from .nadam import Nadam
 from .novograd import NovoGrad
 from .nvnovograd import NvNovoGrad
-from .lookahead import Lookahead
-from .adamp import AdamP
+from .radam import RAdam
+from .rmsprop_tf import RMSpropTF
 from .sgdp import SGDP
-from .optim_factory import create_optimizer
+
+from .optim_factory import create_optimizer

timm/optim/adafactor.py

@@ -0,0 +1,174 @@
+""" Adafactor Optimizer
+
+Lifted from https://github.com/pytorch/fairseq/blob/master/fairseq/optim/adafactor.py
+
+Original header/copyright below.
+
+"""
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+import torch
+import math
+
+
+class Adafactor(torch.optim.Optimizer):
+    """Implements Adafactor algorithm.
+    This implementation is based on: `Adafactor: Adaptive Learning Rates with Sublinear Memory Cost`
+    (see https://arxiv.org/abs/1804.04235)
+
+    Note that this optimizer internally adjusts the learning rate depending on the
+    *scale_parameter*, *relative_step* and *warmup_init* options.
+
+    To use a manual (external) learning rate schedule you should set `scale_parameter=False` and
+    `relative_step=False`.
+
+    Arguments:
+        params (iterable): iterable of parameters to optimize or dicts defining parameter groups
+        lr (float, optional): external learning rate (default: None)
+        eps (tuple[float, float]): regularization constants for square gradient
+            and parameter scale respectively (default: (1e-30, 1e-3))
+        clip_threshold (float): threshold of root mean square of final gradient update (default: 1.0)
+        decay_rate (float): coefficient used to compute running averages of square gradient (default: -0.8)
+        beta1 (float): coefficient used for computing running averages of gradient (default: None)
+        weight_decay (float, optional): weight decay (L2 penalty) (default: 0)
+        scale_parameter (bool): if True, learning rate is scaled by root mean square of parameter (default: True)
+        relative_step (bool): if True, time-dependent learning rate is computed
+            instead of external learning rate (default: True)
+        warmup_init (bool): time-dependent learning rate computation depends on
+            whether warm-up initialization is being used (default: False)
+    """
+
+    def __init__(self, params, lr=None, eps=1e-30, eps_scale=1e-3, clip_threshold=1.0,
+                 decay_rate=-0.8, betas=None, weight_decay=0.0, scale_parameter=True, warmup_init=False):
+        relative_step = lr is None
+        if warmup_init and not relative_step:
+            raise ValueError('warmup_init requires relative_step=True')
+
+        beta1 = None if betas is None else betas[0]   # make it compat with standard betas arg
+        defaults = dict(lr=lr, eps=eps, eps_scale=eps_scale, clip_threshold=clip_threshold, decay_rate=decay_rate,
+                        beta1=beta1, weight_decay=weight_decay, scale_parameter=scale_parameter,
+                        relative_step=relative_step, warmup_init=warmup_init)
+        super(Adafactor, self).__init__(params, defaults)
+
+    @staticmethod
+    def _get_lr(param_group, param_state):
+        if param_group['relative_step']:
+            min_step = 1e-6 * param_state['step'] if param_group['warmup_init'] else 1e-2
+            lr_t = min(min_step, 1.0 / math.sqrt(param_state['step']))
+            param_scale = 1.0
+            if param_group['scale_parameter']:
+                param_scale = max(param_group['eps_scale'], param_state['RMS'])
+            param_group['lr'] = lr_t * param_scale
+        return param_group['lr']
+
+    @staticmethod
+    def _get_options(param_group, param_shape):
+        factored = len(param_shape) >= 2
+        use_first_moment = param_group['beta1'] is not None
+        return factored, use_first_moment
+
+    @staticmethod
+    def _rms(tensor):
+        return tensor.norm(2) / (tensor.numel() ** 0.5)
+
+    def _approx_sq_grad(self, exp_avg_sq_row, exp_avg_sq_col):
+        r_factor = (exp_avg_sq_row / exp_avg_sq_row.mean(dim=-1, keepdim=True)).rsqrt_().unsqueeze(-1)
+        c_factor = exp_avg_sq_col.unsqueeze(-2).rsqrt()
+        return torch.mul(r_factor, c_factor)
+
+    def step(self, closure=None):
+        """Performs a single optimization step.
+        Arguments:
+            closure (callable, optional): A closure that reevaluates the model and returns the loss.
+        """
+        loss = None
+        if closure is not None:
+            loss = closure()
+
+        for group in self.param_groups:
+            for p in group['params']:
+                if p.grad is None:
+                    continue
+                grad = p.grad.data
+                if grad.dtype in {torch.float16, torch.bfloat16}:
+                    grad = grad.float()
+                if grad.is_sparse:
+                    raise RuntimeError('Adafactor does not support sparse gradients.')
+
+                state = self.state[p]
+                grad_shape = grad.shape
+
+                factored, use_first_moment = self._get_options(group, grad_shape)
+                # State Initialization
+                if len(state) == 0:
+                    state['step'] = 0
+
+                    if use_first_moment:
+                        # Exponential moving average of gradient values
+                        state['exp_avg'] = torch.zeros_like(grad)
+                    if factored:
+                        state['exp_avg_sq_row'] = torch.zeros(grad_shape[:-1]).to(grad)
+                        state['exp_avg_sq_col'] = torch.zeros(grad_shape[:-2] + grad_shape[-1:]).to(grad)
+                    else:
+                        state['exp_avg_sq'] = torch.zeros_like(grad)
+
+                    state['RMS'] = 0
+                else:
+                    if use_first_moment:
+                        state['exp_avg'] = state['exp_avg'].to(grad)
+                    if factored:
+                        state['exp_avg_sq_row'] = state['exp_avg_sq_row'].to(grad)
+                        state['exp_avg_sq_col'] = state['exp_avg_sq_col'].to(grad)
+                    else:
+                        state['exp_avg_sq'] = state['exp_avg_sq'].to(grad)
+
+                p_data_fp32 = p.data
+                if p.data.dtype in {torch.float16, torch.bfloat16}:
+                    p_data_fp32 = p_data_fp32.float()
+
+                state['step'] += 1
+                state['RMS'] = self._rms(p_data_fp32)
+                lr_t = self._get_lr(group, state)
+
+                beta2t = 1.0 - math.pow(state['step'], group['decay_rate'])
+                update = grad ** 2 + group['eps']
+                if factored:
+                    exp_avg_sq_row = state['exp_avg_sq_row']
+                    exp_avg_sq_col = state['exp_avg_sq_col']
+
+                    exp_avg_sq_row.mul_(beta2t).add_(1.0 - beta2t, update.mean(dim=-1))
+                    exp_avg_sq_col.mul_(beta2t).add_(1.0 - beta2t, update.mean(dim=-2))
+                    #exp_avg_sq_row.mul_(beta2t).add_(update.mean(dim=-1), alpha=1.0 - beta2t)  # pytorch 1.6+
+                    #exp_avg_sq_col.mul_(beta2t).add_(update.mean(dim=-2), alpha=1.0 - beta2t)
+
+                    # Approximation of exponential moving average of square of gradient
+                    update = self._approx_sq_grad(exp_avg_sq_row, exp_avg_sq_col)
+                    update.mul_(grad)
+                else:
+                    exp_avg_sq = state['exp_avg_sq']
+
+                    exp_avg_sq.mul_(beta2t).add_(1.0 - beta2t, update)
+                    #exp_avg_sq.mul_(beta2t).add_(update, alpha=1.0 - beta2t)  # pytorch 1.6+
+                    update = exp_avg_sq.rsqrt().mul_(grad)
+
+                update.div_((self._rms(update) / group['clip_threshold']).clamp_(min=1.0))
+                update.mul_(lr_t)
+
+                if use_first_moment:
+                    exp_avg = state['exp_avg']
+                    exp_avg.mul_(group["beta1"]).add_(1 - group["beta1"], update)
+                    #exp_avg.mul_(group['beta1']).add_(update, alpha=1 - group['beta1'])  # pytorch 1.6+
+                    update = exp_avg
+
+                if group['weight_decay'] != 0:
+                    p_data_fp32.add_(-group["weight_decay"] * lr_t, p_data_fp32)
+                    #p_data_fp32.add_(p_data_fp32, alpha=-group['weight_decay'] * lr_t)  # pytorch 1.6+
+
+                p_data_fp32.add_(-update)
+
+                if p.data.dtype in {torch.float16, torch.bfloat16}:
+                    p.data.copy_(p_data_fp32)
+
+        return loss

timm/optim/adahessian.py

@@ -0,0 +1,156 @@
+""" AdaHessian Optimizer
+
+Lifted from https://github.com/davda54/ada-hessian/blob/master/ada_hessian.py
+Originally licensed MIT, Copyright 2020, David Samuel
+"""
+import torch
+
+
+class Adahessian(torch.optim.Optimizer):
+    """
+    Implements the AdaHessian algorithm from "ADAHESSIAN: An Adaptive Second OrderOptimizer for Machine Learning"
+
+    Arguments:
+        params (iterable): iterable of parameters to optimize or dicts defining parameter groups
+        lr (float, optional): learning rate (default: 0.1)
+        betas ((float, float), optional): coefficients used for computing running averages of gradient and the
+            squared hessian trace (default: (0.9, 0.999))
+        eps (float, optional): term added to the denominator to improve numerical stability (default: 1e-8)
+        weight_decay (float, optional): weight decay (L2 penalty) (default: 0.0)
+        hessian_power (float, optional): exponent of the hessian trace (default: 1.0)
+        update_each (int, optional): compute the hessian trace approximation only after *this* number of steps
+            (to save time) (default: 1)
+        n_samples (int, optional): how many times to sample `z` for the approximation of the hessian trace (default: 1)
+    """
+
+    def __init__(self, params, lr=0.1, betas=(0.9, 0.999), eps=1e-8, weight_decay=0.0,
+                 hessian_power=1.0, update_each=1, n_samples=1, avg_conv_kernel=False):
+        if not 0.0 <= lr:
+            raise ValueError(f"Invalid learning rate: {lr}")
+        if not 0.0 <= eps:
+            raise ValueError(f"Invalid epsilon value: {eps}")
+        if not 0.0 <= betas[0] < 1.0:
+            raise ValueError(f"Invalid beta parameter at index 0: {betas[0]}")
+        if not 0.0 <= betas[1] < 1.0:
+            raise ValueError(f"Invalid beta parameter at index 1: {betas[1]}")
+        if not 0.0 <= hessian_power <= 1.0:
+            raise ValueError(f"Invalid Hessian power value: {hessian_power}")
+
+        self.n_samples = n_samples
+        self.update_each = update_each
+        self.avg_conv_kernel = avg_conv_kernel
+
+        # use a separate generator that deterministically generates the same `z`s across all GPUs in case of distributed training
+        self.seed = 2147483647
+        self.generator = torch.Generator().manual_seed(self.seed)
+
+        defaults = dict(lr=lr, betas=betas, eps=eps, weight_decay=weight_decay, hessian_power=hessian_power)
+        super(Adahessian, self).__init__(params, defaults)
+
+        for p in self.get_params():
+            p.hess = 0.0
+            self.state[p]["hessian step"] = 0
+
+    @property
+    def is_second_order(self):
+        return True
+
+    def get_params(self):
+        """
+        Gets all parameters in all param_groups with gradients
+        """
+
+        return (p for group in self.param_groups for p in group['params'] if p.requires_grad)
+
+    def zero_hessian(self):
+        """
+        Zeros out the accumalated hessian traces.
+        """
+
+        for p in self.get_params():
+            if not isinstance(p.hess, float) and self.state[p]["hessian step"] % self.update_each == 0:
+                p.hess.zero_()
+
+    @torch.no_grad()
+    def set_hessian(self):
+        """
+        Computes the Hutchinson approximation of the hessian trace and accumulates it for each trainable parameter.
+        """
+
+        params = []
+        for p in filter(lambda p: p.grad is not None, self.get_params()):
+            if self.state[p]["hessian step"] % self.update_each == 0:  # compute the trace only each `update_each` step
+                params.append(p)
+            self.state[p]["hessian step"] += 1
+
+        if len(params) == 0:
+            return
+
+        if self.generator.device != params[0].device:  # hackish way of casting the generator to the right device
+            self.generator = torch.Generator(params[0].device).manual_seed(self.seed)
+
+        grads = [p.grad for p in params]
+
+        for i in range(self.n_samples):
+            # Rademacher distribution {-1.0, 1.0}
+            zs = [torch.randint(0, 2, p.size(), generator=self.generator, device=p.device) * 2.0 - 1.0 for p in params]
+            h_zs = torch.autograd.grad(
+                grads, params, grad_outputs=zs, only_inputs=True, retain_graph=i < self.n_samples - 1)
+            for h_z, z, p in zip(h_zs, zs, params):
+                p.hess += h_z * z / self.n_samples  # approximate the expected values of z*(H@z)
+
+    @torch.no_grad()
+    def step(self, closure=None):
+        """
+        Performs a single optimization step.
+        Arguments:
+            closure (callable, optional) -- a closure that reevaluates the model and returns the loss (default: None)
+        """
+
+        loss = None
+        if closure is not None:
+            loss = closure()
+
+        self.zero_hessian()
+        self.set_hessian()
+
+        for group in self.param_groups:
+            for p in group['params']:
+                if p.grad is None or p.hess is None:
+                    continue
+
+                if self.avg_conv_kernel and p.dim() == 4:
+                    p.hess = torch.abs(p.hess).mean(dim=[2, 3], keepdim=True).expand_as(p.hess).clone()
+
+                # Perform correct stepweight decay as in AdamW
+                p.mul_(1 - group['lr'] * group['weight_decay'])
+
+                state = self.state[p]
+
+                # State initialization
+                if len(state) == 1:
+                    state['step'] = 0
+                    # Exponential moving average of gradient values
+                    state['exp_avg'] = torch.zeros_like(p)
+                    # Exponential moving average of Hessian diagonal square values
+                    state['exp_hessian_diag_sq'] = torch.zeros_like(p)
+
+                exp_avg, exp_hessian_diag_sq = state['exp_avg'], state['exp_hessian_diag_sq']
+                beta1, beta2 = group['betas']
+                state['step'] += 1
+
+                # Decay the first and second moment running average coefficient
+                exp_avg.mul_(beta1).add_(p.grad, alpha=1 - beta1)
+                exp_hessian_diag_sq.mul_(beta2).addcmul_(p.hess, p.hess, value=1 - beta2)
+
+                bias_correction1 = 1 - beta1 ** state['step']
+                bias_correction2 = 1 - beta2 ** state['step']
+
+                k = group['hessian_power']
+                denom = (exp_hessian_diag_sq / bias_correction2).pow_(k / 2).add_(group['eps'])
+
+                # make update
+                step_size = group['lr'] / bias_correction1
+                p.addcdiv_(exp_avg, denom, value=-step_size)
+
+        return loss

timm/optim/optim_factory.py

@@ -3,7 +3,18 @@ Hacked together by / Copyright 2020 Ross Wightman
 """
 import torch
 from torch import optim as optim
-from timm.optim import Nadam, RMSpropTF, AdamW, RAdam, NovoGrad, NvNovoGrad, Lookahead, AdamP, SGDP
+
+from .adafactor import Adafactor
+from .adahessian import Adahessian
+from .adamp import AdamP
+from .lookahead import Lookahead
+from .nadam import Nadam
+from .novograd import NovoGrad
+from .nvnovograd import NvNovoGrad
+from .radam import RAdam
+from .rmsprop_tf import RMSpropTF
+from .sgdp import SGDP
+
 try:
     from apex.optimizers import FusedNovoGrad, FusedAdam, FusedLAMB, FusedSGD
     has_apex = True
@@ -29,11 +40,6 @@ def add_weight_decay(model, weight_decay=1e-5, skip_list=()):
 def create_optimizer(args, model, filter_bias_and_bn=True):
     opt_lower = args.opt.lower()
     weight_decay = args.weight_decay
-    if 'adamw' in opt_lower or 'radam' in opt_lower:
-        # Compensate for the way current AdamW and RAdam optimizers apply LR to the weight-decay
-        # I don't believe they follow the paper or original Torch7 impl which schedules weight
-        # decay based on the ratio of current_lr/initial_lr
-        weight_decay /= args.lr
     if weight_decay and filter_bias_and_bn:
         parameters = add_weight_decay(model, weight_decay)
         weight_decay = 0.
@@ -43,66 +49,59 @@ def create_optimizer(args, model, filter_bias_and_bn=True):
     if 'fused' in opt_lower:
         assert has_apex and torch.cuda.is_available(), 'APEX and CUDA required for fused optimizers'
 
+    opt_args = dict(lr=args.lr, weight_decay=weight_decay)
+    if args.opt_eps is not None:
+        opt_args['eps'] = args.opt_eps
+    if args.opt_betas is not None:
+        opt_args['betas'] = args.opt_betas
+
     opt_split = opt_lower.split('_')
     opt_lower = opt_split[-1]
     if opt_lower == 'sgd' or opt_lower == 'nesterov':
-        optimizer = optim.SGD(
-            parameters, lr=args.lr, momentum=args.momentum, weight_decay=weight_decay, nesterov=True)
+        optimizer = optim.SGD(parameters, momentum=args.momentum, nesterov=True, **opt_args)
     elif opt_lower == 'momentum':
-        optimizer = optim.SGD(
-            parameters, lr=args.lr, momentum=args.momentum, weight_decay=weight_decay, nesterov=False)
+        optimizer = optim.SGD(parameters, momentum=args.momentum, nesterov=False, **opt_args)
     elif opt_lower == 'adam':
-        optimizer = optim.Adam(
-            parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = optim.Adam(parameters, **opt_args)
     elif opt_lower == 'adamw':
-        optimizer = AdamW(
-            parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = optim.AdamW(parameters, **opt_args)
     elif opt_lower == 'nadam':
-        optimizer = Nadam(
-            parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = Nadam(parameters, **opt_args)
     elif opt_lower == 'radam':
-        optimizer = RAdam(
-            parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = RAdam(parameters, **opt_args)
     elif opt_lower == 'adamp':        
-        optimizer = AdamP(
-            parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps,
-            delta=0.1, wd_ratio=0.01, nesterov=True)
+        optimizer = AdamP(parameters, wd_ratio=0.01, nesterov=True, **opt_args)
     elif opt_lower == 'sgdp':        
-        optimizer = SGDP(
-            parameters, lr=args.lr, momentum=args.momentum, weight_decay=weight_decay, 
-            eps=args.opt_eps, nesterov=True)        
+        optimizer = SGDP(parameters, momentum=args.momentum, nesterov=True, **opt_args)
     elif opt_lower == 'adadelta':
-        optimizer = optim.Adadelta(
-            parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = optim.Adadelta(parameters, **opt_args)
+    elif opt_lower == 'adafactor':
+        if not args.lr:
+            opt_args['lr'] = None
+        optimizer = Adafactor(parameters, **opt_args)
+    elif opt_lower == 'adahessian':
+        optimizer = Adahessian(parameters, **opt_args)
     elif opt_lower == 'rmsprop':
-        optimizer = optim.RMSprop(
-            parameters, lr=args.lr, alpha=0.9, eps=args.opt_eps,
-            momentum=args.momentum, weight_decay=weight_decay)
+        optimizer = optim.RMSprop(parameters, alpha=0.9, momentum=args.momentum, **opt_args)
     elif opt_lower == 'rmsproptf':
-        optimizer = RMSpropTF(
-            parameters, lr=args.lr, alpha=0.9, eps=args.opt_eps,
-            momentum=args.momentum, weight_decay=weight_decay)
+        optimizer = RMSpropTF(parameters, alpha=0.9, momentum=args.momentum, **opt_args)
     elif opt_lower == 'novograd':
-        optimizer = NovoGrad(parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = NovoGrad(parameters, **opt_args)
     elif opt_lower == 'nvnovograd':
-        optimizer = NvNovoGrad(parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = NvNovoGrad(parameters, **opt_args)
     elif opt_lower == 'fusedsgd':
-        optimizer = FusedSGD(
-            parameters, lr=args.lr, momentum=args.momentum, weight_decay=weight_decay, nesterov=True)
+        optimizer = FusedSGD(parameters, momentum=args.momentum, nesterov=True, **opt_args)
     elif opt_lower == 'fusedmomentum':
-        optimizer = FusedSGD(
-            parameters, lr=args.lr, momentum=args.momentum, weight_decay=weight_decay, nesterov=False)
+        optimizer = FusedSGD(parameters, momentum=args.momentum, nesterov=False, **opt_args)
     elif opt_lower == 'fusedadam':
-        optimizer = FusedAdam(
-            parameters, lr=args.lr, adam_w_mode=False, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = FusedAdam(parameters, adam_w_mode=False, **opt_args)
     elif opt_lower == 'fusedadamw':
-        optimizer = FusedAdam(
-            parameters, lr=args.lr, adam_w_mode=True, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = FusedAdam(parameters, adam_w_mode=True, **opt_args)
     elif opt_lower == 'fusedlamb':
-        optimizer = FusedLAMB(parameters, lr=args.lr, weight_decay=weight_decay, eps=args.opt_eps)
+        optimizer = FusedLAMB(parameters, **opt_args)
     elif opt_lower == 'fusednovograd':
-        optimizer = FusedNovoGrad(
-            parameters, lr=args.lr, betas=(0.95, 0.98), weight_decay=weight_decay, eps=args.opt_eps)
+        opt_args.setdefault('betas', (0.95, 0.98))
+        optimizer = FusedNovoGrad(parameters, **opt_args)
     else:
         assert False and "Invalid optimizer"
         raise ValueError

timm/utils/cuda.py

@@ -15,10 +15,10 @@ except ImportError:
 class ApexScaler:
     state_dict_key = "amp"
 
-    def __call__(self, loss, optimizer, clip_grad=None, parameters=None):
+    def __call__(self, loss, optimizer, clip_grad=None, parameters=None, create_graph=False):
         with amp.scale_loss(loss, optimizer) as scaled_loss:
-            scaled_loss.backward()
-        if clip_grad:
+            scaled_loss.backward(create_graph=create_graph)
+        if clip_grad is not None:
             torch.nn.utils.clip_grad_norm_(amp.master_params(optimizer), clip_grad)
         optimizer.step()
 
@@ -37,9 +37,9 @@ class NativeScaler:
     def __init__(self):
         self._scaler = torch.cuda.amp.GradScaler()
 
-    def __call__(self, loss, optimizer, clip_grad=None, parameters=None):
-        self._scaler.scale(loss).backward()
-        if clip_grad:
+    def __call__(self, loss, optimizer, clip_grad=None, parameters=None, create_graph=False):
+        self._scaler.scale(loss).backward(create_graph=create_graph)
+        if clip_grad is not None:
             assert parameters is not None
             self._scaler.unscale_(optimizer)  # unscale the gradients of optimizer's assigned params in-place
             torch.nn.utils.clip_grad_norm_(parameters, clip_grad)

train.py

@@ -98,12 +98,18 @@ parser.add_argument('-vb', '--validation-batch-size-multiplier', type=int, defau
 # Optimizer parameters
 parser.add_argument('--opt', default='sgd', type=str, metavar='OPTIMIZER',
                     help='Optimizer (default: "sgd"')
-parser.add_argument('--opt-eps', default=1e-8, type=float, metavar='EPSILON',
-                    help='Optimizer Epsilon (default: 1e-8)')
+parser.add_argument('--opt-eps', default=None, type=float, metavar='EPSILON',
+                    help='Optimizer Epsilon (default: None, use opt default)')
+parser.add_argument('--opt-betas', default=None, type=float, nargs='+', metavar='BETA',
+                    help='Optimizer Betas (default: None, use opt default)')
 parser.add_argument('--momentum', type=float, default=0.9, metavar='M',
-                    help='SGD momentum (default: 0.9)')
+                    help='Optimizer momentum (default: 0.9)')
 parser.add_argument('--weight-decay', type=float, default=0.0001,
                     help='weight decay (default: 0.0001)')
+parser.add_argument('--clip-grad', type=float, default=None, metavar='NORM',
+                    help='Clip gradient norm (default: None, no clipping)')
+
+
 
 # Learning rate schedule parameters
 parser.add_argument('--sched', default='step', type=str, metavar='SCHEDULER',
@@ -595,6 +601,7 @@ def train_epoch(
         elif mixup_fn is not None:
             mixup_fn.mixup_enabled = False
 
+    second_order = hasattr(optimizer, 'is_second_order') and optimizer.is_second_order
     batch_time_m = AverageMeter()
     data_time_m = AverageMeter()
     losses_m = AverageMeter()
@@ -623,9 +630,12 @@ def train_epoch(
 
         optimizer.zero_grad()
         if loss_scaler is not None:
-            loss_scaler(loss, optimizer)
+            loss_scaler(
+                loss, optimizer, clip_grad=args.clip_grad, parameters=model.parameters(), create_graph=second_order)
         else:
-            loss.backward()
+            loss.backward(create_graph=second_order)
+            if args.clip_grad is not None:
+                torch.nn.utils.clip_grad_norm_(model.parameters(), args.clip_grad)
             optimizer.step()
 
         torch.cuda.synchronize()