Hello, (Initially posted in beginners section but moved)
I am trying to be able to train the transformer model GPT-JT-6B-v1. Although i have a couple of CPU servers with 128 gb ram and a couple GPU servers with 48 gb ram I cant really seem to get the accelerate library to work.
I started a simple test where I have two CPU servers on the same local network, I run accelerate config and answer according to the included picture. I do the same on both servers. The code i want to run by multi node cpu is a computer vision example from accelerate/cv_example.py at main · huggingface/accelerate · GitHub
which you can see below, I added a few lines to print out a message just at the main process but when I run accelerate launch ./cv_example.py --data_dir ./images --cpu
on both servers they both print out that they are the main process. Is this not the way to do it?
Thanks for the help!
Best Regards
Heigke
# coding=utf-8
# Copyright 2021 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import argparse
import os
import re
import numpy as np
import PIL
import torch
from timm import create_model
from torch.optim.lr_scheduler import OneCycleLR
from torch.utils.data import DataLoader, Dataset
from torchvision.transforms import Compose, RandomResizedCrop, Resize, ToTensor
from accelerate import Accelerator
########################################################################
# This is a fully working simple example to use Accelerate
#
# This example trains a ResNet50 on the Oxford-IIT Pet Dataset
# in any of the following settings (with the same script):
# - single CPU or single GPU
# - multi GPUS (using PyTorch distributed mode)
# - (multi) TPUs
# - fp16 (mixed-precision) or fp32 (normal precision)
#
# To run it in each of these various modes, follow the instructions
# in the readme for examples:
# https://github.com/huggingface/accelerate/tree/main/examples
#
########################################################################
# Function to get the label from the filename
def extract_label(fname):
stem = fname.split(os.path.sep)[-1]
return re.search(r"^(.*)_\d+\.jpg$", stem).groups()[0]
class PetsDataset(Dataset):
def __init__(self, file_names, image_transform=None, label_to_id=None):
self.file_names = file_names
self.image_transform = image_transform
self.label_to_id = label_to_id
def __len__(self):
return len(self.file_names)
def __getitem__(self, idx):
fname = self.file_names[idx]
raw_image = PIL.Image.open(fname)
image = raw_image.convert("RGB")
if self.image_transform is not None:
image = self.image_transform(image)
label = extract_label(fname)
if self.label_to_id is not None:
label = self.label_to_id[label]
return {"image": image, "label": label}
def training_function(config, args):
# Initialize accelerator
accelerator = Accelerator(cpu=args.cpu, mixed_precision=args.mixed_precision)
# Sample hyper-parameters for learning rate, batch size, seed and a few other HPs
lr = config["lr"]
num_epochs = int(config["num_epochs"])
seed = int(config["seed"])
batch_size = int(config["batch_size"])
image_size = config["image_size"]
if not isinstance(image_size, (list, tuple)):
image_size = (image_size, image_size)
# Grab all the image filenames
file_names = [os.path.join(args.data_dir, fname) for fname in os.listdir(args.data_dir) if fname.endswith(".jpg")]
# Build the label correspondences
all_labels = [extract_label(fname) for fname in file_names]
id_to_label = list(set(all_labels))
id_to_label.sort()
label_to_id = {lbl: i for i, lbl in enumerate(id_to_label)}
# Set the seed before splitting the data.
np.random.seed(seed)
torch.manual_seed(seed)
torch.cuda.manual_seed_all(seed)
# Split our filenames between train and validation
random_perm = np.random.permutation(len(file_names))
cut = int(0.8 * len(file_names))
train_split = random_perm[:cut]
eval_split = random_perm[cut:]
# For training we use a simple RandomResizedCrop
train_tfm = Compose([RandomResizedCrop(image_size, scale=(0.5, 1.0)), ToTensor()])
train_dataset = PetsDataset(
[file_names[i] for i in train_split], image_transform=train_tfm, label_to_id=label_to_id
)
# For evaluation, we use a deterministic Resize
eval_tfm = Compose([Resize(image_size), ToTensor()])
eval_dataset = PetsDataset([file_names[i] for i in eval_split], image_transform=eval_tfm, label_to_id=label_to_id)
# Instantiate dataloaders.
train_dataloader = DataLoader(train_dataset, shuffle=True, batch_size=batch_size, num_workers=4)
eval_dataloader = DataLoader(eval_dataset, shuffle=False, batch_size=batch_size, num_workers=4)
# Instantiate the model (we build the model here so that the seed also control new weights initialization)
model = create_model("resnet50d", pretrained=True, num_classes=len(label_to_id))
# We could avoid this line since the accelerator is set with `device_placement=True` (default value).
# Note that if you are placing tensors on devices manually, this line absolutely needs to be before the optimizer
# creation otherwise training will not work on TPU (`accelerate` will kindly throw an error to make us aware of that).
model = model.to(accelerator.device)
# Freezing the base model
for param in model.parameters():
param.requires_grad = False
for param in model.get_classifier().parameters():
param.requires_grad = True
# We normalize the batches of images to be a bit faster.
mean = torch.tensor(model.default_cfg["mean"])[None, :, None, None].to(accelerator.device)
std = torch.tensor(model.default_cfg["std"])[None, :, None, None].to(accelerator.device)
# Instantiate optimizer
optimizer = torch.optim.Adam(params=model.parameters(), lr=lr / 25)
# Instantiate learning rate scheduler
lr_scheduler = OneCycleLR(optimizer=optimizer, max_lr=lr, epochs=num_epochs, steps_per_epoch=len(train_dataloader))
# Prepare everything
# There is no specific order to remember, we just need to unpack the objects in the same order we gave them to the
# prepare method.
model, optimizer, train_dataloader, eval_dataloader, lr_scheduler = accelerator.prepare(
model, optimizer, train_dataloader, eval_dataloader, lr_scheduler
)
# HEIGKE ADDED THESE LINES TO SEE THAT IT WORKS
if accelerator.is_main_process:
print("I believe that I am the main process")
# Now we train the model
for epoch in range(num_epochs):
model.train()
for step, batch in enumerate(train_dataloader):
# We could avoid this line since we set the accelerator with `device_placement=True`.
batch = {k: v.to(accelerator.device) for k, v in batch.items()}
inputs = (batch["image"] - mean) / std
outputs = model(inputs)
loss = torch.nn.functional.cross_entropy(outputs, batch["label"])
accelerator.backward(loss)
optimizer.step()
lr_scheduler.step()
optimizer.zero_grad()
model.eval()
accurate = 0
num_elems = 0
for _, batch in enumerate(eval_dataloader):
# We could avoid this line since we set the accelerator with `device_placement=True`.
batch = {k: v.to(accelerator.device) for k, v in batch.items()}
inputs = (batch["image"] - mean) / std
with torch.no_grad():
outputs = model(inputs)
predictions = outputs.argmax(dim=-1)
predictions, references = accelerator.gather_for_metrics((predictions, batch["label"]))
accurate_preds = predictions == references
num_elems += accurate_preds.shape[0]
accurate += accurate_preds.long().sum()
eval_metric = accurate.item() / num_elems
# Use accelerator.print to print only on the main process.
accelerator.print(f"epoch {epoch}: {100 * eval_metric:.2f}")
def main():
parser = argparse.ArgumentParser(description="Simple example of training script.")
parser.add_argument("--data_dir", required=True, help="The data folder on disk.")
parser.add_argument(
"--mixed_precision",
type=str,
default=None,
choices=["no", "fp16", "bf16"],
help="Whether to use mixed precision. Choose"
"between fp16 and bf16 (bfloat16). Bf16 requires PyTorch >= 1.10."
"and an Nvidia Ampere GPU.",
)
parser.add_argument(
"--checkpointing_steps",
type=str,
default=None,
help="Whether the various states should be saved at the end of every n steps, or 'epoch' for each epoch.",
)
parser.add_argument("--cpu", action="store_true", help="If passed, will train on the CPU.")
args = parser.parse_args()
config = {"lr": 3e-2, "num_epochs": 3, "seed": 42, "batch_size": 64, "image_size": 224}
training_function(config, args)
if __name__ == "__main__":
main()
And also what I actually want to get to work is a multi node version of this GPT-JT-6B-V1 test code
accelerator = Accelerator(gradient_accumulation_steps=2)
print("loadin dataset")
dataset = load_dataset("yelp_review_full")
dataset["train"][100]
print("done loading dataset")
print("loading model")
model = AutoModelForCausalLMWithValueHead.from_pretrained("togethercomputer/GPT-JT-6B-v1")
model = accelerator.prepare(model)
#model_ref = AutoModelForCausalLMWithValueHead.from_pretrained("togethercomputer/GPT-JT-6B-v1")
print("MODEL LOADED")
#model = AutoModelForCausalLMWithValueHead.from_pretrained('gpt2')
#model_ref = AutoModelForCausalLMWithValueHead.from_pretrained('gpt2')
#tokenizer = GPT2Tokenizer.from_pretrained('gpt2')
tokenizer = AutoTokenizer.from_pretrained("togethercomputer/GPT-JT-6B-v1")
tokenizer.pad_token = tokenizer.eos_token
#tokenizer.add_special_tokens({'pad_token': '[PAD]'})
print("TOKENIZER LOADED")
#tokenizer.add_special_tokens({'pad_token': '[PAD]'})
metric = evaluate.load("accuracy")
print("done loading accuracy")
def compute_metrics(eval_pred):
logits, labels = eval_pred
predictions = np.argmax(logits, axis=-1)
return metric.compute(predictions=predictions, references=labels)
training_args = TrainingArguments(output_dir="test_trainer", evaluation_strategy="epoch")
print("trainingargs created")
def tokenize_function(examples):
return tokenizer(examples["text"], padding="max_length", truncation=True)
print("starting to tokenize")
tokenized_datasets = dataset.map(tokenize_function, batched=True)
print("done tokenizing")
small_train_dataset = tokenized_datasets["train"].shuffle(seed=42).select(range(1000))
small_eval_dataset = tokenized_datasets["test"].shuffle(seed=42).select(range(1000))
print(small_train_dataset)
trainer = Trainer(
model=model,
args=training_args,
train_dataset=small_train_dataset,
eval_dataset=small_eval_dataset,
compute_metrics=compute_metrics,
)
print("Training on Yelp review")
trainer.train()