trainer.py

import copy
import functools
import os

import blobfile as bf
import numpy as np
import torch
import torch.distributed as dist
from torch.nn.parallel.distributed import DistributedDataParallel as DDP
from torch.optim import AdamW

from src.utils import dist_util, logger
from src.utils.fp16_util import (
    make_master_params,
    master_params_to_model_params,
    model_grads_to_master_grads,
    unflatten_master_params,
    zero_grad,
)

from src.modeling.diffusion.nn import update_ema
from src.modeling.diffusion.resample import LossAwareSampler, UniformSampler

# For ImageNet experiments, this was a good default value.
# We found that the lg_loss_scale quickly climbed to
# 20-21 within the first ~1K steps of training.
INITIAL_LOG_LOSS_SCALE = 20.0


class Trainer:
    def __init__(
        self,
        *,
        model,
        diffusion,
        data,
        batch_size,
        microbatch,
        lr,
        ema_rate,
        log_interval,
        save_interval,
        resume_checkpoint,
        use_fp16=False,
        fp16_scale_growth=1e-3,
        schedule_sampler=None,
        weight_decay=0.0,
        lr_anneal_steps=0,
        checkpoint_path='',
        gradient_clipping=-1.,
        eval_data=None,
        eval_interval=-1,
        warmup=None,
        dae=False,
        gamma_nll=0.01
    ):
        self.model = model
        self.diffusion = diffusion
        self.data = data
        self.eval_data = eval_data
        self.batch_size = batch_size
        self.microbatch = microbatch if microbatch > 0 else batch_size
        self.lr = lr
        self.gamma_nll = gamma_nll
        self.dae = dae
        self.ema_rate = (
            [ema_rate]
            if isinstance(ema_rate, float)
            else [float(x) for x in ema_rate.split(",")]
        )
        self.log_interval = log_interval
        self.eval_interval = eval_interval
        self.save_interval = save_interval
        self.resume_checkpoint = resume_checkpoint
        self.use_fp16 = use_fp16
        self.fp16_scale_growth = fp16_scale_growth
        self.schedule_sampler = schedule_sampler or UniformSampler(diffusion)
        self.weight_decay = weight_decay
        self.lr_anneal_steps = lr_anneal_steps
        self.gradient_clipping = gradient_clipping
        self.warmup = warmup

        self.step = 0
        self.resume_step = 0
        self.global_batch = self.batch_size * dist.get_world_size()

        self.model_params = list(self.model.parameters())
        self.master_params = self.model_params
        self.state_dict = self.model.state_dict()
        self.lg_loss_scale = INITIAL_LOG_LOSS_SCALE
        self.sync_cuda = torch.cuda.is_available()

        self.checkpoint_path = checkpoint_path  # DEBUG **

        self._load_and_sync_parameters()
        if self.use_fp16:
            self._setup_fp16()

        # self.opt = AdamW(self.master_params, lr=self.lr, weight_decay=self.weight_decay)
        self.opt = AdamW(filter(lambda p: p.requires_grad, self.master_params),
                         lr=self.lr, weight_decay=self.weight_decay)
        if self.resume_step:
            self._load_optimizer_state()
            self.ema_params = [
                self._load_ema_parameters(rate) for rate in self.ema_rate
            ]
        else:
            self.ema_params = [
                copy.deepcopy(self.master_params) for _ in range(len(self.ema_rate))
            ]

        if torch.cuda.is_available(): # DEBUG **
            self.use_ddp = True
            self.ddp_model = DDP(
                self.model,
                device_ids=[dist_util.dev()],
                output_device=dist_util.dev(),
                broadcast_buffers=False,
                bucket_cap_mb=128,
                find_unused_parameters=False,
            )
        else:
            if dist.get_world_size() > 1:
                logger.warn(
                    "Distributed training requires CUDA. "
                    "Gradients will not be synchronized properly!"
                )
            self.use_ddp = False
            self.ddp_model = self.model

    def _load_and_sync_parameters(self):
        resume_checkpoint = find_resume_checkpoint() or self.resume_checkpoint

        if resume_checkpoint:
            self.resume_step = parse_resume_step_from_filename(resume_checkpoint)
            if dist.get_rank() == 0:
                logger.log(f"loading model from checkpoint: {resume_checkpoint}...")
                self.model.load_state_dict(
                    dist_util.load_state_dict(
                        resume_checkpoint, map_location=dist_util.dev()
                    )
                )

        dist_util.sync_params(self.model.parameters())

    def _load_ema_parameters(self, rate):
        ema_params = copy.deepcopy(self.master_params)

        main_checkpoint = find_resume_checkpoint() or self.resume_checkpoint
        ema_checkpoint = find_ema_checkpoint(main_checkpoint, self.resume_step, rate)
        if ema_checkpoint:
            if dist.get_rank() == 0:
                logger.log(f"loading EMA from checkpoint: {ema_checkpoint}...")
                state_dict = dist_util.load_state_dict(
                    ema_checkpoint, map_location=dist_util.dev()
                )
                ema_params = self._state_dict_to_master_params(state_dict)

        dist_util.sync_params(ema_params)
        return ema_params

    def _load_optimizer_state(self):
        main_checkpoint = find_resume_checkpoint() or self.resume_checkpoint
        opt_checkpoint = bf.join(
            bf.dirname(main_checkpoint), f"opt{self.resume_step:06}.pt"
        )
        if bf.exists(opt_checkpoint):
            logger.log(f"loading optimizer state from checkpoint: {opt_checkpoint}")
            state_dict = dist_util.load_state_dict(
                opt_checkpoint, map_location=dist_util.dev()
            )
            self.opt.load_state_dict(state_dict)

    def _setup_fp16(self):
        self.master_params = make_master_params(self.model_params)
        self.model.convert_to_fp16()

    def run_loop(self):
        while (
            not self.lr_anneal_steps
            or self.step + self.resume_step < self.lr_anneal_steps
        ):  
            batch, cond = next(self.data)
            self.run_step(batch, cond)
            if self.step % self.log_interval == 0:
                logger.dumpkvs()
            if self.eval_data is not None and self.step % self.eval_interval == 1:
                batch_eval, cond_eval = next(self.eval_data)
                self.forward_only(batch_eval, cond_eval)
                print('diffu_eval on validation set')
                logger.dumpkvs()
            if self.step % self.save_interval == 0:
                self.save()
                if os.environ.get("DIFFUSION_TRAINING_TEST", "") and self.step > 0:
                    return
            self.step += 1

        # Save the last checkpoint if it wasn't already saved.
        if (self.step - 1) % self.save_interval != 0:
            self.save()

    def run_step(self, batch, cond):
        self.forward_backward(batch, cond)
        if self.use_fp16:
            self.optimize_fp16()
        else:
            self.optimize_normal()
        self.log_step()

    def forward_only(self, batch, cond):

        batch_size = batch['input_ids'].shape[0]
        with torch.no_grad():
            zero_grad(self.model_params)
            for i in range(0, batch_size, self.microbatch):
                micro = {key: item[i : i + self.microbatch].to(dist_util.dev()) for key, item in batch.items()}
                if cond == None:
                    micro_cond = None
                else:
                    micro_cond = {
                        k: v[i : i + self.microbatch].to(dist_util.dev())
                        for k, v in cond.items()
                    }
                last_batch = (i + self.microbatch) >= batch_size
                t, weights = self.schedule_sampler.sample(micro['input_ids'].shape[0], dist_util.dev())

                compute_losses = functools.partial(
                            self.diffusion.training_losses,
                            self.ddp_model,
                            self.step,
                            t,
                            model_kwargs=micro,
                            dae=self.dae
                        )

                if last_batch or not self.use_ddp:
                    losses, anchor_loss = compute_losses()
                else:
                    with self.ddp_model.no_sync():
                        losses, anchor_loss = compute_losses()

                # log_loss_dict(
                #     self.diffusion, t, {f"eval_{k}": v * weights for k, v in losses.items()}
                # )
                logger.logkv_mean("eval_anchor_loss", anchor_loss.item())
                logger.logkv_mean("eval_mse", (losses["mse"] * weights).mean().item())
                logger.logkv_mean("eval_loss_sds", (losses["loss"] * weights).mean().item())

    def forward_backward(self, batch, cond):
        # print(batch)
        batch_size = batch['input_ids'].shape[0]
        zero_grad(self.model_params)
        for i in range(0, batch_size, self.microbatch):
            micro = {key: item[i : i + self.microbatch].to(dist_util.dev()) for key, item in batch.items()}
            if cond == None:
                micro_cond = None
            else:
                micro_cond = {
                    k: v[i : i + self.microbatch].to(dist_util.dev())
                    for k, v in cond.items()
                }
            
            last_batch = (i + self.microbatch) >= batch_size
            t, weights = self.schedule_sampler.sample(micro['input_ids'].shape[0], dist_util.dev())

            compute_losses = functools.partial(
                self.diffusion.training_losses,
                self.ddp_model,
                self.step,
                t,
                model_kwargs=micro,
                dae=self.dae
            )

            if last_batch or not self.use_ddp:
                losses, anchor_loss = compute_losses()
            else:
                with self.ddp_model.no_sync():
                    losses, anchor_loss = compute_losses()

            # if isinstance(self.schedule_sampler, LossAwareSampler):
            #     self.schedule_sampler.update_with_local_losses(
            #         t, losses["loss"].detach()
            #     )
            # loss = (losses["loss"] * weights).mean()
            if isinstance(self.schedule_sampler, LossAwareSampler):
                self.schedule_sampler.update_with_local_losses(
                    t, losses["mse"].detach()
                )

            loss = (losses["mse_tT"] * weights).mean() + self.gamma_nll * anchor_loss

            # log_loss_dict(
            #     self.diffusion, t, {k: v * weights for k, v in losses.items()}
            # )
            logger.logkv_mean("anchor_loss", anchor_loss.item())
            logger.logkv_mean("mse", (losses["mse"] * weights).mean().item())
            logger.logkv_mean("loss_sds", (losses["loss"] * weights).mean().item())

            if self.use_fp16:
                loss_scale = 2 ** self.lg_loss_scale
                (loss * loss_scale).backward()
            else:
                loss.backward()

    def optimize_fp16(self):
        if any(not torch.isfinite(p.grad).all() for p in self.model_params):
            self.lg_loss_scale -= 1
            logger.log(f"Found NaN, decreased lg_loss_scale to {self.lg_loss_scale}")
            return

        model_grads_to_master_grads(self.model_params, self.master_params)
        self.master_params[0].grad.mul_(1.0 / (2 ** self.lg_loss_scale))
        self._log_grad_norm()
        self._anneal_lr()
        self.opt.step()
        for rate, params in zip(self.ema_rate, self.ema_params):
            update_ema(params, self.master_params, rate=rate)
        master_params_to_model_params(self.model_params, self.master_params)
        self.lg_loss_scale += self.fp16_scale_growth

    def grad_clip(self):
        # print('doing gradient clipping')
        max_grad_norm=self.gradient_clipping #3.0
        if hasattr(self.opt, "clip_grad_norm"):
            # Some optimizers (like the sharded optimizer) have a specific way to do gradient clipping
            self.opt.clip_grad_norm(max_grad_norm)
        else:
            # Revert to normal clipping otherwise, handling Apex or full precision
            torch.nn.utils.clip_grad_norm_(
                self.model.parameters(), #amp.master_params(self.opt) if self.use_apex else
                max_grad_norm,
            )

    def optimize_normal(self):
        if self.gradient_clipping > 0:
            self.grad_clip()
        self._log_grad_norm()
        self._anneal_lr()
        self.opt.step()
        for rate, params in zip(self.ema_rate, self.ema_params):
            update_ema(params, self.master_params, rate=rate)

    def _log_grad_norm(self):
        sqsum = 0.0
        for p in self.master_params:
            if p.grad is not None:
                sqsum += (p.grad ** 2).sum().item()
        logger.logkv_mean("grad_norm", np.sqrt(sqsum))

    def _anneal_lr(self):
        if not self.lr_anneal_steps:
            return
        
        if self.warmup is not None and self.warmup > 0:
            warmup_frac = (self.step + self.resume_step) / self.warmup
            frac_done = (self.step + self.resume_step  - self.warmup) / (self.lr_anneal_steps  - self.warmup)
            
            lr = self.lr * min(1, warmup_frac) * min(1-frac_done, 1)

        else:
            frac_done = (self.step + self.resume_step) / self.lr_anneal_steps
            lr = self.lr * (1 - frac_done)
        
        for param_group in self.opt.param_groups:
            param_group["lr"] = lr

    def log_step(self):
        logger.logkv("step", self.step + self.resume_step)
        logger.logkv("samples", (self.step + self.resume_step + 1) * self.global_batch)
        if self.use_fp16:
            logger.logkv("lg_loss_scale", self.lg_loss_scale)

    def save(self):
        def save_checkpoint(rate, params):
            self._master_params_to_state_dict(params)
            if dist.get_rank() == 0:
                logger.log(f"saving model {rate}...")
                if not rate:
                    filename = f"model{(self.step+self.resume_step):06d}.pt"
                else:
                    filename = f"ema_{rate}_{(self.step+self.resume_step):06d}.pt"
                print('writing to', bf.join(get_blob_logdir(), filename))
                print('writing to', bf.join(self.checkpoint_path, filename))
                # with bf.BlobFile(bf.join(get_blob_logdir(), filename), "wb") as f:
                #     torch.save(state_dict, f)
                with bf.BlobFile(bf.join(self.checkpoint_path, filename), "wb") as f: # DEBUG **
                    torch.save(self.state_dict, f)

        save_checkpoint(0, self.master_params)
        for rate, params in zip(self.ema_rate, self.ema_params):
            save_checkpoint(rate, params)

        dist.barrier()

    def _master_params_to_state_dict(self, master_params):
        if self.use_fp16:
            master_params = unflatten_master_params(
                list(self.model.parameters()), master_params # DEBUG **
            )
        self.state_dict = self.model.state_dict()
        for i, (name, _value) in enumerate(self.model.named_parameters()):
            assert name in self.state_dict
            self.state_dict[name] = master_params[i]
        # return self.state_dict

    def _state_dict_to_master_params(self, state_dict):
        params = [state_dict[name] for name, _ in self.model.named_parameters()]
        if self.use_fp16:
            return make_master_params(params)
        else:
            return params


def parse_resume_step_from_filename(filename):
    """
    Parse filenames of the form path/to/modelNNNNNN.pt, where NNNNNN is the
    checkpoint's number of steps.
    """
    split = filename.split("model")
    if len(split) < 2:
        return 0
    split1 = split[-1].split(".")[0]
    try:
        return int(split1)
    except ValueError:
        return 0


def get_blob_logdir():
    return os.environ.get("DIFFUSION_BLOB_LOGDIR", logger.get_dir())


def find_resume_checkpoint():
    # On your infrastructure, you may want to override this to automatically
    # discover the latest checkpoint on your blob storage, etc.
    return None


def find_ema_checkpoint(main_checkpoint, step, rate):
    if main_checkpoint is None:
        return None
    filename = f"ema_{rate}_{(step):06d}.pt"
    path = bf.join(bf.dirname(main_checkpoint), filename)
    if bf.exists(path):
        return path
    return None


def log_loss_dict(diffusion, ts, losses):
    for key, values in losses.items():
        logger.logkv_mean(key, values.mean().item())
        # Log the quantiles (four quartiles, in particular).
        for sub_t, sub_loss in zip(ts.cpu().numpy(), values.detach().cpu().numpy()):
            quartile = int(4 * sub_t / diffusion.num_timesteps)
            logger.logkv_mean(f"{key}_q{quartile}", sub_loss)


class Trainer_PTE:
    def __init__(
            self,
            *,
            model,
            data,
            batch_size,
            microbatch,
            lr,
            log_interval,
            save_interval,
            resume_checkpoint,
            use_fp16=False,
            fp16_scale_growth=1e-3,
            weight_decay=0.0,
            lr_anneal_steps=0,
            checkpoint_path='',
            gradient_clipping=-1.,
            eval_data=None,
            eval_interval=-1,
            warmup=None,
            dae=False
    ):
        self.model = model
        self.data = data
        self.eval_data = eval_data
        self.batch_size = batch_size
        self.microbatch = microbatch if microbatch > 0 else batch_size
        self.lr = lr
        self.dae = dae
        self.log_interval = log_interval
        self.eval_interval = eval_interval
        self.save_interval = save_interval
        self.resume_checkpoint = resume_checkpoint
        self.use_fp16 = use_fp16
        self.fp16_scale_growth = fp16_scale_growth
        self.weight_decay = weight_decay
        self.lr_anneal_steps = lr_anneal_steps
        self.gradient_clipping = gradient_clipping
        self.warmup = warmup

        self.step = 0
        self.resume_step = 0
        self.global_batch = self.batch_size * dist.get_world_size()

        self.model_params = list(self.model.parameters())
        self.master_params = self.model_params
        self.state_dict = self.model.state_dict()
        self.lg_loss_scale = INITIAL_LOG_LOSS_SCALE
        self.sync_cuda = torch.cuda.is_available()

        self.checkpoint_path = checkpoint_path  # DEBUG **

        self._load_and_sync_parameters()
        if self.use_fp16:
            self._setup_fp16()

        self.opt = AdamW(self.master_params, lr=self.lr, weight_decay=self.weight_decay)

        if self.resume_step:
            self._load_optimizer_state()

        if torch.cuda.is_available():  # DEBUG **
            self.use_ddp = True
            self.ddp_model = DDP(
                self.model,
                device_ids=[dist_util.dev()],
                output_device=dist_util.dev(),
                broadcast_buffers=False,
                bucket_cap_mb=128,
                find_unused_parameters=False,
            )
        else:
            if dist.get_world_size() > 1:
                logger.warn(
                    "Distributed training requires CUDA. "
                    "Gradients will not be synchronized properly!"
                )
            self.use_ddp = False
            self.ddp_model = self.model

    def _load_and_sync_parameters(self):
        resume_checkpoint = find_resume_checkpoint() or self.resume_checkpoint
        if resume_checkpoint:
            self.resume_step = parse_resume_step_from_filename(resume_checkpoint)
            if dist.get_rank() == 0:
                logger.log(f"loading model from checkpoint: {resume_checkpoint}...")
                self.model.load_state_dict(
                    dist_util.load_state_dict(
                        resume_checkpoint, map_location=dist_util.dev()
                    )
                )
        dist_util.sync_params(self.model.parameters())

    def _load_optimizer_state(self):
        main_checkpoint = find_resume_checkpoint() or self.resume_checkpoint
        opt_checkpoint = bf.join(
            bf.dirname(main_checkpoint), f"opt{self.resume_step:06}.pt"
        )
        if bf.exists(opt_checkpoint):
            logger.log(f"loading optimizer state from checkpoint: {opt_checkpoint}")
            state_dict = dist_util.load_state_dict(
                opt_checkpoint, map_location=dist_util.dev()
            )
            self.opt.load_state_dict(state_dict)

    def _setup_fp16(self):
        self.master_params = make_master_params(self.model_params)
        self.model.convert_to_fp16()

    def run_loop(self):
        while (
                not self.lr_anneal_steps
                or self.step + self.resume_step < self.lr_anneal_steps
        ):
            batch, _ = next(self.data)
            self.run_step(batch)
            if self.step % self.log_interval == 0:
                logger.dumpkvs()
            if self.eval_data is not None and self.step % self.eval_interval == 1:
                batch_eval, _ = next(self.eval_data)
                self.forward_only(batch_eval)
                print('diffu_eval on validation set')
                logger.dumpkvs()
            if self.step % self.save_interval == 0:
                self.save()
                if os.environ.get("DIFFUSION_TRAINING_TEST", "") and self.step > 0:
                    return
            self.step += 1

        # Save the last checkpoint if it wasn't already saved.
        if (self.step - 1) % self.save_interval != 0:
            self.save()

    def run_step(self, batch):
        self.forward_backward(batch)
        if self.use_fp16:
            self.optimize_fp16()
        else:
            self.optimize_normal()
        self.log_step()

    def forward_only(self, batch):

        batch_size = batch['input_ids'].shape[0]
        with torch.no_grad():
            zero_grad(self.model_params)
            for i in range(0, batch_size, self.microbatch):
                micro = {key: item[i: i + self.microbatch].to(dist_util.dev()) for key, item in batch.items()}
                last_batch = (i + self.microbatch) >= batch_size
                if last_batch or not self.use_ddp:
                    loss, _ = self.ddp_model.module.forward(input_ids=micro["input_ids"],
                                                            attention_mask=micro["attention_mask"],
                                                            decoder_input_ids=micro["decoder_input_ids"],
                                                            labels=micro["labels"],
                                                            dae=self.dae)
                else:
                    with self.ddp_model.no_sync():
                        loss, _ = self.ddp_model.module.forward(input_ids=micro["input_ids"],
                                                                attention_mask=micro["attention_mask"],
                                                                decoder_input_ids=micro["decoder_input_ids"],
                                                                labels=micro["labels"],
                                                                dae=self.dae)
                logger.logkv_mean("loss", loss.item())

    def forward_backward(self, batch):
        # print(batch)
        batch_size = batch['input_ids'].shape[0]
        zero_grad(self.model_params)
        for i in range(0, batch_size, self.microbatch):
            micro = {key: item[i: i + self.microbatch].to(dist_util.dev()) for key, item in batch.items()}
            last_batch = (i + self.microbatch) >= batch_size
            if last_batch or not self.use_ddp:
                loss, _ = self.ddp_model.module.forward(input_ids=micro["input_ids"],
                                                        attention_mask=micro["attention_mask"],
                                                        decoder_input_ids=micro["decoder_input_ids"],
                                                        labels=micro["labels"],
                                                        dae=self.dae)
            else:
                with self.ddp_model.no_sync():
                    loss, _ = self.ddp_model.module.forward(input_ids=micro["input_ids"],
                                                            attention_mask=micro["attention_mask"],
                                                            decoder_input_ids=micro["decoder_input_ids"],
                                                            labels=micro["labels"],
                                                            dae=self.dae)
            logger.logkv_mean("loss", loss.item())
            if self.use_fp16:
                loss_scale = 2 ** self.lg_loss_scale
                (loss * loss_scale).backward()
            else:
                loss.backward()

    def optimize_fp16(self):
        if any(not torch.isfinite(p.grad).all() for p in self.model_params):
            self.lg_loss_scale -= 1
            logger.log(f"Found NaN, decreased lg_loss_scale to {self.lg_loss_scale}")
            return

        model_grads_to_master_grads(self.model_params, self.master_params)
        self.master_params[0].grad.mul_(1.0 / (2 ** self.lg_loss_scale))
        self._anneal_lr()
        self.opt.step()
        master_params_to_model_params(self.model_params, self.master_params)
        self.lg_loss_scale += self.fp16_scale_growth

    def grad_clip(self):
        # print('doing gradient clipping')
        max_grad_norm = self.gradient_clipping  # 3.0
        if hasattr(self.opt, "clip_grad_norm"):
            # Some optimizers (like the sharded optimizer) have a specific way to do gradient clipping
            self.opt.clip_grad_norm(max_grad_norm)
        else:
            # Revert to normal clipping otherwise, handling Apex or full precision
            torch.nn.utils.clip_grad_norm_(
                self.model.parameters(),  # amp.master_params(self.opt) if self.use_apex else
                max_grad_norm,
            )

    def optimize_normal(self):
        if self.gradient_clipping > 0:
            self.grad_clip()
        self._anneal_lr()
        self.opt.step()

    def _anneal_lr(self):
        if not self.lr_anneal_steps:
            return

        if self.warmup is not None and self.warmup > 0:
            warmup_frac = (self.step + self.resume_step) / self.warmup
            frac_done = (self.step + self.resume_step - self.warmup) / (self.lr_anneal_steps - self.warmup)

            lr = self.lr * min(1, warmup_frac) * min(1 - frac_done, 1)

        else:
            frac_done = (self.step + self.resume_step) / self.lr_anneal_steps
            lr = self.lr * (1 - frac_done)

        for param_group in self.opt.param_groups:
            param_group["lr"] = lr

    def log_step(self):
        logger.logkv("step", self.step + self.resume_step)
        logger.logkv("samples", (self.step + self.resume_step + 1) * self.global_batch)
        if self.use_fp16:
            logger.logkv("lg_loss_scale", self.lg_loss_scale)

    def save(self):

        self._master_params_to_state_dict(self.master_params)

        if dist.get_rank() == 0:

            logger.log(f"saving model state dict...")
            filename = f"model{(self.step + self.resume_step):06d}.pt"
            with bf.BlobFile(bf.join(self.checkpoint_path, filename), "wb") as f:  # DEBUG **
                torch.save(self.state_dict, f)

            logger.log(f"saving pretrained model...")
            model_to_save = self.ddp_model.module if hasattr(self.ddp_model, "module") else self.ddp_model
            dict_name = f"model{(self.step + self.resume_step):06d}"
            model_to_save.embedder.save_pretrained(os.path.join(self.checkpoint_path, dict_name))

        dist.barrier()

    def _master_params_to_state_dict(self, master_params):
        if self.use_fp16:
            master_params = unflatten_master_params(
                list(self.model.parameters()), master_params  # DEBUG **
            )
        self.state_dict = self.model.state_dict()
        for i, (name, _value) in enumerate(self.model.named_parameters()):
            assert name in self.state_dict
            self.state_dict[name] = master_params[i]
        # return self.state_dict

    def _state_dict_to_master_params(self, state_dict):
        params = [state_dict[name] for name, _ in self.model.named_parameters()]
        if self.use_fp16:
            return make_master_params(params)
        else:
            return params