feat($solver): change scheduler call methods

using name of lr scheduler in config to call

fastreid/solver/build.py

@@ -34,21 +34,22 @@ def build_optimizer(cfg, model):
 
 
 def build_lr_scheduler(cfg, optimizer):
-    if cfg.SOLVER.SCHED == "warmup":
+    scheduler_args = {
-        return lr_scheduler.WarmupMultiStepLR(
+        "optimizer": optimizer,
-            optimizer,
+
-            cfg.SOLVER.STEPS,
+        # warmup options
-            cfg.SOLVER.GAMMA,
+        "warmup_factor": cfg.SOLVER.WARMUP_FACTOR,
-            warmup_factor=cfg.SOLVER.WARMUP_FACTOR,
+        "warmup_iters": cfg.SOLVER.WARMUP_ITERS,
-            warmup_iters=cfg.SOLVER.WARMUP_ITERS,
+        "warmup_method": cfg.SOLVER.WARMUP_METHOD,
-            warmup_method=cfg.SOLVER.WARMUP_METHOD
+
-        )
+        # multi-step lr scheduler options
-    elif cfg.SOLVER.SCHED == "delay":
+        "milestones": cfg.SOLVER.STEPS,
-        return lr_scheduler.DelayedCosineAnnealingLR(
+        "gamma": cfg.SOLVER.GAMMA,
-            optimizer,
+
-            cfg.SOLVER.DELAY_ITERS,
+        # cosine annealing lr scheduler options
-            cfg.SOLVER.COS_ANNEAL_ITERS,
+        "max_iters": cfg.SOLVER.MAX_ITER,
-            warmup_factor=cfg.SOLVER.WARMUP_FACTOR,
+        "delay_iters": cfg.SOLVER.DELAY_ITERS,
-            warmup_iters=cfg.SOLVER.WARMUP_ITERS,
+        "eta_min_lr": cfg.SOLVER.ETA_MIN_LR,
-            warmup_method=cfg.SOLVER.WARMUP_METHOD
+
-        )
+    }
+    return getattr(lr_scheduler, cfg.SOLVER.SCHED)(**scheduler_args)

fastreid/solver/lr_scheduler.py

@@ -10,7 +10,7 @@ from typing import List
 import torch
 from torch.optim.lr_scheduler import _LRScheduler, CosineAnnealingLR
 
-__all__ = ["WarmupMultiStepLR", "DelayerScheduler"]
+__all__ = ["WarmupMultiStepLR", "DelayedScheduler"]
 
 
 class WarmupMultiStepLR(_LRScheduler):
@@ -23,6 +23,7 @@ class WarmupMultiStepLR(_LRScheduler):
             warmup_iters: int = 1000,
             warmup_method: str = "linear",
             last_epoch: int = -1,
+            **kwargs,
     ):
         if not list(milestones) == sorted(milestones):
             raise ValueError(
@@ -76,16 +77,16 @@ def _get_warmup_factor_at_iter(
         raise ValueError("Unknown warmup method: {}".format(method))
 
 
-class DelayerScheduler(_LRScheduler):
+class DelayedScheduler(_LRScheduler):
     """ Starts with a flat lr schedule until it reaches N epochs the applies a scheduler
     Args:
         optimizer (Optimizer): Wrapped optimizer.
-        delay_epochs: number of epochs to keep the initial lr until starting aplying the scheduler
+        delay_iters: number of epochs to keep the initial lr until starting applying the scheduler
         after_scheduler: after target_epoch, use this scheduler(eg. ReduceLROnPlateau)
     """
 
-    def __init__(self, optimizer, delay_epochs, after_scheduler, warmup_factor, warmup_iters, warmup_method):
+    def __init__(self, optimizer, delay_iters, after_scheduler, warmup_factor, warmup_iters, warmup_method):
-        self.delay_epochs = delay_epochs
+        self.delay_epochs = delay_iters
         self.after_scheduler = after_scheduler
         self.finished = False
         self.warmup_factor = warmup_factor
@@ -94,7 +95,6 @@ class DelayerScheduler(_LRScheduler):
         super().__init__(optimizer)
 
     def get_lr(self):
-
         if self.last_epoch >= self.delay_epochs:
             if not self.finished:
                 self.after_scheduler.base_lrs = self.base_lrs
@@ -113,10 +113,11 @@ class DelayerScheduler(_LRScheduler):
             else:
                 self.after_scheduler.step(epoch - self.delay_epochs)
         else:
-            return super(DelayerScheduler, self).step(epoch)
+            return super(DelayedScheduler, self).step(epoch)
 
 
-def DelayedCosineAnnealingLR(optimizer, delay_epochs, cosine_annealing_epochs, warmup_factor,
+def DelayedCosineAnnealingLR(optimizer, delay_iters, max_iters, eta_min_lr, warmup_factor,
-                             warmup_iters, warmup_method):
+                             warmup_iters, warmup_method, **kwargs, ):
-    base_scheduler = CosineAnnealingLR(optimizer, cosine_annealing_epochs, eta_min=0)
+    cosine_annealing_iters = max_iters - delay_iters
-    return DelayerScheduler(optimizer, delay_epochs, base_scheduler, warmup_factor, warmup_iters, warmup_method)
+    base_scheduler = CosineAnnealingLR(optimizer, cosine_annealing_iters, eta_min_lr)
+    return DelayedScheduler(optimizer, delay_iters, base_scheduler, warmup_factor, warmup_iters, warmup_method)

fastreid/solver/optim/ranger.py

@@ -1,14 +1,177 @@
-####
+# Ranger deep learning optimizer - RAdam + Lookahead + Gradient Centralization, combined into one optimizer.
-# CODE TAKEN FROM https://github.com/lessw2020/Ranger-Deep-Learning-Optimizer
+
-# Blog post: https://medium.com/@lessw/new-deep-learning-optimizer-ranger-synergistic-combination-of-radam-lookahead-for-the-best-of-2dc83f79a48d
+# https://github.com/lessw2020/Ranger-Deep-Learning-Optimizer
-####
+# and/or
+# https://github.com/lessw2020/Best-Deep-Learning-Optimizers
+
+# Ranger has now been used to capture 12 records on the FastAI leaderboard.
+
+# This version = 20.4.11
+
+# Credits:
+# Gradient Centralization --> https://arxiv.org/abs/2004.01461v2 (a new optimization technique for DNNs), github:  https://github.com/Yonghongwei/Gradient-Centralization
+# RAdam -->  https://github.com/LiyuanLucasLiu/RAdam
+# Lookahead --> rewritten by lessw2020, but big thanks to Github @LonePatient and @RWightman for ideas from their code.
+# Lookahead paper --> MZhang,G Hinton  https://arxiv.org/abs/1907.08610
+
+# summary of changes:
+# 4/11/20 - add gradient centralization option.  Set new testing benchmark for accuracy with it, toggle with use_gc flag at init.
+# full code integration with all updates at param level instead of group, moves slow weights into state dict (from generic weights),
+# supports group learning rates (thanks @SHolderbach), fixes sporadic load from saved model issues.
+# changes 8/31/19 - fix references to *self*.N_sma_threshold;
+# changed eps to 1e-5 as better default than 1e-8.
 
 import math
+
 import torch
-from .lookahead import Lookahead
+from torch.optim.optimizer import Optimizer
-from .radam import RAdam
 
 
-def Ranger(params, alpha=0.5, k=6, betas=(.95, 0.999), *args, **kwargs):
+class Ranger(Optimizer):
-    radam = RAdam(params, betas=betas, *args, **kwargs)
+
-    return Lookahead(radam, alpha, k)
+    def __init__(self, params, lr=1e-3,  # lr
+                 alpha=0.5, k=6, N_sma_threshhold=5,  # Ranger options
+                 betas=(.95, 0.999), eps=1e-5, weight_decay=0,  # Adam options
+                 use_gc=True, gc_conv_only=False
+                 # Gradient centralization on or off, applied to conv layers only or conv + fc layers
+                 ):
+
+        # parameter checks
+        if not 0.0 <= alpha <= 1.0:
+            raise ValueError(f'Invalid slow update rate: {alpha}')
+        if not 1 <= k:
+            raise ValueError(f'Invalid lookahead steps: {k}')
+        if not lr > 0:
+            raise ValueError(f'Invalid Learning Rate: {lr}')
+        if not eps > 0:
+            raise ValueError(f'Invalid eps: {eps}')
+
+        # parameter comments:
+        # beta1 (momentum) of .95 seems to work better than .90...
+        # N_sma_threshold of 5 seems better in testing than 4.
+        # In both cases, worth testing on your dataset (.90 vs .95, 4 vs 5) to make sure which works best for you.
+
+        # prep defaults and init torch.optim base
+        defaults = dict(lr=lr, alpha=alpha, k=k, step_counter=0, betas=betas, N_sma_threshhold=N_sma_threshhold,
+                        eps=eps, weight_decay=weight_decay)
+        super().__init__(params, defaults)
+
+        # adjustable threshold
+        self.N_sma_threshhold = N_sma_threshhold
+
+        # look ahead params
+
+        self.alpha = alpha
+        self.k = k
+
+        # radam buffer for state
+        self.radam_buffer = [[None, None, None] for ind in range(10)]
+
+        # gc on or off
+        self.use_gc = use_gc
+
+        # level of gradient centralization
+        self.gc_gradient_threshold = 3 if gc_conv_only else 1
+
+        print(f"Ranger optimizer loaded. \nGradient Centralization usage = {self.use_gc}")
+        if (self.use_gc and self.gc_gradient_threshold == 1):
+            print(f"GC applied to both conv and fc layers")
+        elif (self.use_gc and self.gc_gradient_threshold == 3):
+            print(f"GC applied to conv layers only")
+
+    def __setstate__(self, state):
+        print("set state called")
+        super(Ranger, self).__setstate__(state)
+
+    def step(self, closure=None):
+        loss = None
+        # note - below is commented out b/c I have other work that passes back the loss as a float, and thus not a callable closure.
+        # Uncomment if you need to use the actual closure...
+
+        # if closure is not None:
+        # loss = closure()
+
+        # Evaluate averages and grad, update param tensors
+        for group in self.param_groups:
+
+            for p in group['params']:
+                if p.grad is None:
+                    continue
+                grad = p.grad.data.float()
+
+                if grad.is_sparse:
+                    raise RuntimeError('Ranger optimizer does not support sparse gradients')
+
+                p_data_fp32 = p.data.float()
+
+                state = self.state[p]  # get state dict for this param
+
+                if len(state) == 0:  # if first time to run...init dictionary with our desired entries
+                    # if self.first_run_check==0:
+                    # self.first_run_check=1
+                    # print("Initializing slow buffer...should not see this at load from saved model!")
+                    state['step'] = 0
+                    state['exp_avg'] = torch.zeros_like(p_data_fp32)
+                    state['exp_avg_sq'] = torch.zeros_like(p_data_fp32)
+
+                    # look ahead weight storage now in state dict
+                    state['slow_buffer'] = torch.empty_like(p.data)
+                    state['slow_buffer'].copy_(p.data)
+
+                else:
+                    state['exp_avg'] = state['exp_avg'].type_as(p_data_fp32)
+                    state['exp_avg_sq'] = state['exp_avg_sq'].type_as(p_data_fp32)
+
+                # begin computations
+                exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq']
+                beta1, beta2 = group['betas']
+
+                # GC operation for Conv layers and FC layers
+                if grad.dim() > self.gc_gradient_threshold:
+                    grad.add_(-grad.mean(dim=tuple(range(1, grad.dim())), keepdim=True))
+
+                state['step'] += 1
+
+                # compute variance mov avg
+                exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad)
+                # compute mean moving avg
+                exp_avg.mul_(beta1).add_(1 - beta1, grad)
+
+                buffered = self.radam_buffer[int(state['step'] % 10)]
+
+                if state['step'] == buffered[0]:
+                    N_sma, step_size = buffered[1], buffered[2]
+                else:
+                    buffered[0] = state['step']
+                    beta2_t = beta2 ** state['step']
+                    N_sma_max = 2 / (1 - beta2) - 1
+                    N_sma = N_sma_max - 2 * state['step'] * beta2_t / (1 - beta2_t)
+                    buffered[1] = N_sma
+                    if N_sma > self.N_sma_threshhold:
+                        step_size = math.sqrt(
+                            (1 - beta2_t) * (N_sma - 4) / (N_sma_max - 4) * (N_sma - 2) / N_sma * N_sma_max / (
+                                    N_sma_max - 2)) / (1 - beta1 ** state['step'])
+                    else:
+                        step_size = 1.0 / (1 - beta1 ** state['step'])
+                    buffered[2] = step_size
+
+                if group['weight_decay'] != 0:
+                    p_data_fp32.add_(-group['weight_decay'] * group['lr'], p_data_fp32)
+
+                # apply lr
+                if N_sma > self.N_sma_threshhold:
+                    denom = exp_avg_sq.sqrt().add_(group['eps'])
+                    p_data_fp32.addcdiv_(-step_size * group['lr'], exp_avg, denom)
+                else:
+                    p_data_fp32.add_(-step_size * group['lr'], exp_avg)
+
+                p.data.copy_(p_data_fp32)
+
+                # integrated look ahead...
+                # we do it at the param level instead of group level
+                if state['step'] % group['k'] == 0:
+                    slow_p = state['slow_buffer']  # get access to slow param tensor
+                    slow_p.add_(self.alpha, p.data - slow_p)  # (fast weights - slow weights) * alpha
+                    p.data.copy_(slow_p)  # copy interpolated weights to RAdam param tensor
+
+        return loss