def main():
"""Entrypoint for standalone HF causal LM fine-tuning without Trainer."""
rank = ezpz.setup_torch()
args = parse_args()
# Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
# information sent is the one passed as arguments along with your Python/PyTorch versions.
# send_example_telemetry("run_clm_no_trainer", args)
# Initialize the accelerator. We will let the accelerator handle device placement for us in this example.
# If we're using tracking, we also need to initialize it here and it will by default pick up all supported trackers
# in the environment
accelerator_log_kwargs = {}
if args.with_tracking:
accelerator_log_kwargs["log_with"] = args.report_to
accelerator_log_kwargs["project_dir"] = args.output_dir
accelerator = Accelerator(
gradient_accumulation_steps=args.gradient_accumulation_steps,
**accelerator_log_kwargs,
)
# Make one log on every process with the configuration for debugging.
# logging.basicConfig(
# format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
# datefmt="%m/%d/%Y %H:%M:%S",
# level=logging.INFO,
# )
logger.warning(accelerator.state)
if accelerator.is_local_main_process:
datasets.utils.logging.set_verbosity_warning()
transformers.utils.logging.set_verbosity_info()
else:
datasets.utils.logging.set_verbosity_error()
transformers.utils.logging.set_verbosity_error()
# If passed along, set the training seed now.
if args.seed is not None:
set_seed(args.seed)
# Handle the repository creation
if accelerator.is_main_process:
if args.push_to_hub:
# Retrieve of infer repo_name
repo_name = args.hub_model_id
if repo_name is None:
repo_name = Path(args.output_dir).absolute().name
# Create repo and retrieve repo_id
api = HfApi()
repo_id = api.create_repo(
repo_name, exist_ok=True, token=args.hub_token
).repo_id
with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
if "step_*" not in gitignore:
gitignore.write("step_*\n")
if "epoch_*" not in gitignore:
gitignore.write("epoch_*\n")
elif args.output_dir is not None:
os.makedirs(args.output_dir, exist_ok=True)
accelerator.wait_for_everyone()
# Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below)
# or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/
# (the dataset will be downloaded automatically from the datasets Hub).
#
# For CSV/JSON files, this script will use the column called 'text' or the first column if no column called
# 'text' is found. You can easily tweak this behavior (see below).
#
# In distributed training, the load_dataset function guarantee that only one local process can concurrently
# download the dataset.
if args.dataset_name is not None:
# Downloading and loading a dataset from the hub.
raw_datasets = load_dataset(
args.dataset_name,
args.dataset_config_name,
trust_remote_code=args.trust_remote_code,
)
rdk = getattr(raw_datasets, "keys", lambda: raw_datasets.keys())
if "validation" not in rdk():
# if "validation" not in raw_datasets.keys():
raw_datasets["validation"] = load_dataset(
args.dataset_name,
args.dataset_config_name,
split=f"train[:{args.validation_split_percentage}%]",
trust_remote_code=args.trust_remote_code,
)
raw_datasets["train"] = load_dataset(
args.dataset_name,
args.dataset_config_name,
split=f"train[{args.validation_split_percentage}%:]",
trust_remote_code=args.trust_remote_code,
)
else:
data_files = {}
dataset_args = {}
if args.train_file is not None:
data_files["train"] = args.train_file
extension = args.train_file.split(".")[-1]
if args.validation_file is not None:
data_files["validation"] = args.validation_file
extension = args.validation_file.split(".")[-1]
if extension == "txt":
extension = "text"
dataset_args["keep_linebreaks"] = not args.no_keep_linebreaks
raw_datasets = load_dataset(extension, data_files=data_files, **dataset_args)
# If no validation data is there, validation_split_percentage will be used to divide the dataset.
if "validation" not in raw_datasets.keys():
raw_datasets["validation"] = load_dataset(
extension,
data_files=data_files,
split=f"train[:{args.validation_split_percentage}%]",
**dataset_args,
)
raw_datasets["train"] = load_dataset(
extension,
data_files=data_files,
split=f"train[{args.validation_split_percentage}%:]",
**dataset_args,
)
# See more about loading any type of standard or custom dataset (from
# files, python dict, pandas DataFrame, etc) at
# https://huggingface.co/docs/datasets/loading_datasets.
# Load pretrained model and tokenizer
#
# In distributed training, the .from_pretrained methods guarantee that only
# one local process can concurrently
# download model & vocab.
if args.config_name:
config = AutoConfig.from_pretrained(
args.config_name,
trust_remote_code=args.trust_remote_code,
)
elif args.model_name_or_path:
config = AutoConfig.from_pretrained(
args.model_name_or_path,
trust_remote_code=args.trust_remote_code,
)
else:
config = CONFIG_MAPPING[args.model_type]()
logger.warning("You are instantiating a new config instance from scratch.")
if args.tokenizer_name:
tokenizer = AutoTokenizer.from_pretrained(
args.tokenizer_name,
use_fast=not args.use_slow_tokenizer,
trust_remote_code=args.trust_remote_code,
)
elif args.model_name_or_path:
tokenizer = AutoTokenizer.from_pretrained(
args.model_name_or_path,
use_fast=not args.use_slow_tokenizer,
trust_remote_code=args.trust_remote_code,
)
else:
raise ValueError(
"You are instantiating a new tokenizer from scratch. This is not supported by this script. "
"You can do it from another script, save it, and load it from here, using --tokenizer_name."
)
if args.model_name_or_path:
model = AutoModelForCausalLM.from_pretrained(
args.model_name_or_path,
from_tf=bool(".ckpt" in args.model_name_or_path),
config=config,
low_cpu_mem_usage=args.low_cpu_mem_usage,
trust_remote_code=args.trust_remote_code,
)
else:
logger.info("Training new model from scratch")
model = AutoModelForCausalLM.from_config(
config, trust_remote_code=args.trust_remote_code
)
# We resize the embeddings only when necessary to avoid index errors. If you are creating a model from scratch
# on a small vocab and want a smaller embedding size, remove this test.
embedding_size = model.get_input_embeddings().weight.shape[0]
if len(tokenizer) > embedding_size:
model.resize_token_embeddings(len(tokenizer))
# Preprocessing the datasets.
# First we tokenize all the texts.
column_names = raw_datasets["train"].column_names
text_column_name = "text" if "text" in column_names else column_names[0]
def tokenize_function(examples):
"""Tokenize raw text from the selected column."""
return tokenizer(examples[text_column_name])
with accelerator.main_process_first():
tokenized_datasets = raw_datasets.map(
tokenize_function,
batched=True,
num_proc=args.preprocessing_num_workers,
remove_columns=column_names,
load_from_cache_file=not args.overwrite_cache,
desc="Running tokenizer on dataset",
)
if args.block_size is None:
block_size = tokenizer.model_max_length
if block_size > config.max_position_embeddings:
logger.warning(
f"The tokenizer picked seems to have a very large `model_max_length` ({tokenizer.model_max_length}). "
f"Using block_size={min(1024, config.max_position_embeddings)} instead. You can change that default value by passing --block_size xxx."
)
block_size = min(1024, config.max_position_embeddings)
else:
if args.block_size > tokenizer.model_max_length:
logger.warning(
f"The block_size passed ({args.block_size}) is larger than the maximum length for the model "
f"({tokenizer.model_max_length}). Using block_size={tokenizer.model_max_length}."
)
block_size = min(args.block_size, tokenizer.model_max_length)
# Main data processing function that will concatenate all texts from our dataset and generate chunks of block_size.
def group_texts(examples):
"""Concatenate and chunk tokenized text into fixed-size blocks."""
# Concatenate all texts.
concatenated_examples = {k: list(chain(*examples[k])) for k in examples.keys()}
total_length = len(concatenated_examples[list(examples.keys())[0]])
# We drop the small remainder, and if the total_length < block_size we exclude this batch and return an empty dict.
# We could add padding if the model supported it instead of this drop, you can customize this part to your needs.
total_length = (total_length // block_size) * block_size
# Split by chunks of max_len.
result = {
k: [t[i : i + block_size] for i in range(0, total_length, block_size)]
for k, t in concatenated_examples.items()
}
result["labels"] = result["input_ids"].copy()
return result
# Note that with `batched=True`, this map processes 1,000 texts together, so group_texts throws away a remainder
# for each of those groups of 1,000 texts. You can adjust that batch_size here but a higher value might be slower
# to preprocess.
#
# To speed up this part, we use multiprocessing. See the documentation of the map method for more information:
# https://huggingface.co/docs/datasets/process#map
with accelerator.main_process_first():
lm_datasets = tokenized_datasets.map(
group_texts,
batched=True,
num_proc=args.preprocessing_num_workers,
load_from_cache_file=not args.overwrite_cache,
desc=f"Grouping texts in chunks of {block_size}",
)
train_dataset = lm_datasets["train"]
eval_dataset = lm_datasets["validation"]
# Log a few random samples from the training set:
for index in random.sample(range(len(train_dataset)), 3):
logger.debug(f"Sample {index} of the training set: {train_dataset[index]}.")
# DataLoaders creation:
train_dataloader = DataLoader(
train_dataset,
shuffle=True,
collate_fn=default_data_collator,
batch_size=args.per_device_train_batch_size,
)
eval_dataloader = DataLoader(
eval_dataset,
collate_fn=default_data_collator,
batch_size=args.per_device_eval_batch_size,
)
# Optimizer
# Split weights in two groups, one with weight decay and the other not.
no_decay = ["bias", "layer_norm.weight"]
optimizer_grouped_parameters = [
{
"params": [
p
for n, p in model.named_parameters()
if not any(nd in n for nd in no_decay)
],
"weight_decay": args.weight_decay,
},
{
"params": [
p
for n, p in model.named_parameters()
if any(nd in n for nd in no_decay)
],
"weight_decay": 0.0,
},
]
optimizer = torch.optim.AdamW(optimizer_grouped_parameters, lr=args.learning_rate)
# Scheduler and math around the number of training steps.
overrode_max_train_steps = False
num_update_steps_per_epoch = math.ceil(
len(train_dataloader) / args.gradient_accumulation_steps
)
if args.max_train_steps is None:
args.max_train_steps = args.num_train_epochs * num_update_steps_per_epoch
overrode_max_train_steps = True
lr_scheduler = get_scheduler(
name=args.lr_scheduler_type,
optimizer=optimizer,
num_warmup_steps=args.num_warmup_steps * accelerator.num_processes,
num_training_steps=args.max_train_steps
if overrode_max_train_steps
else args.max_train_steps * accelerator.num_processes,
)
# Prepare everything with our `accelerator`.
(
model,
optimizer,
train_dataloader,
eval_dataloader,
lr_scheduler,
) = accelerator.prepare(
model, optimizer, train_dataloader, eval_dataloader, lr_scheduler
)
# On TPU, the tie weights in our model have been disconnected, so we need to restore the ties.
# if accelerator.distributed_type == DistributedType.TPU:
# model.tie_weights()
# We need to recalculate our total training steps as the size of the training dataloader may have changed.
num_update_steps_per_epoch = math.ceil(
len(train_dataloader) / args.gradient_accumulation_steps
)
if overrode_max_train_steps:
args.max_train_steps = args.num_train_epochs * num_update_steps_per_epoch
# Afterwards we recalculate our number of training epochs
args.num_train_epochs = math.ceil(args.max_train_steps / num_update_steps_per_epoch)
# Figure out how many steps we should save the Accelerator states
checkpointing_steps = args.checkpointing_steps
if checkpointing_steps is not None and checkpointing_steps.isdigit():
checkpointing_steps = int(checkpointing_steps)
# We need to initialize the trackers we use, and also store our configuration.
# The trackers initializes automatically on the main process.
if args.with_tracking:
experiment_config = vars(args)
# TensorBoard cannot log Enums, need the raw value
experiment_config["lr_scheduler_type"] = experiment_config[
"lr_scheduler_type"
].value
try:
import wandb
except (ImportError, ModuleNotFoundError):
wandb = None
wbproj_name = (
args.wandb_project_name
if args.wandb_project_name is not None
else args.model_name_or_path
)
wbproj_name = f"ezpz-hf-{wbproj_name}".replace("/", "-")
if (
wandb is not None
and rank == 0
and not os.environ.get("WANDB_DISABLED", False)
):
run = ezpz.setup_wandb(project_name=wbproj_name)
wandb.define_metric(
"num_input_tokens_seen"
) # Allow us to track the number of tokens seen during training
if run is not None:
# assert wandb is not None and run is wandb.run and run is not None
run.config.update(ezpz.get_dist_info())
# try:
run.config.update(args)
# except Exception:
# ezpz.breakpoint(0)
accelerator.init_trackers(wbproj_name, experiment_config)
# Train!
total_batch_size = (
args.per_device_train_batch_size
* accelerator.num_processes
* args.gradient_accumulation_steps
)
logger.info("***** Model *****")
logger.info(f" Model = {model}")
logger.info("***** Args *****")
logger.info(json.dumps(vars(args), indent=4, sort_keys=True))
logger.info("***** Running training *****")
logger.info(f" Num processes = {accelerator.num_processes}")
logger.info(f" Num examples = {len(train_dataset)}")
logger.info(f" Num Epochs = {args.num_train_epochs}")
logger.info(
f" Instantaneous batch size per device = {args.per_device_train_batch_size}"
)
logger.info(
f" Total train batch size (w. parallel, distributed & accumulation) = {total_batch_size}"
)
logger.info(f" Gradient Accumulation steps = {args.gradient_accumulation_steps}")
logger.info(f" Total optimization steps = {args.max_train_steps}")
# Only show the progress bar once on each machine.
progress_bar = tqdm(
range(args.max_train_steps),
disable=not accelerator.is_local_main_process,
)
completed_steps = 0
starting_epoch = 0
# Potentially load in the weights and states from a previous save
if args.resume_from_checkpoint:
if args.resume_from_checkpoint is not None or args.resume_from_checkpoint != "":
checkpoint_path = args.resume_from_checkpoint
path = os.path.basename(args.resume_from_checkpoint)
else:
# Get the most recent checkpoint
dirs = [f.name for f in os.scandir(os.getcwd()) if f.is_dir()]
dirs.sort(key=os.path.getctime)
path = dirs[
-1
] # Sorts folders by date modified, most recent checkpoint is the last
checkpoint_path = path
path = os.path.basename(checkpoint_path)
accelerator.print(f"Resumed from checkpoint: {checkpoint_path}")
accelerator.load_state(checkpoint_path)
# Extract `epoch_{i}` or `step_{i}`
training_difference = os.path.splitext(path)[0]
if "epoch" in training_difference:
starting_epoch = int(training_difference.replace("epoch_", "")) + 1
resume_step = None
completed_steps = starting_epoch * num_update_steps_per_epoch
else:
# need to multiply `gradient_accumulation_steps` to reflect real steps
resume_step = (
int(training_difference.replace("step_", ""))
* args.gradient_accumulation_steps
)
starting_epoch = resume_step // len(train_dataloader)
completed_steps = resume_step // args.gradient_accumulation_steps
resume_step -= starting_epoch * len(train_dataloader)
else:
resume_step = None
# update the progress_bar if load from checkpoint
progress_bar.update(completed_steps)
import time
total_loss = 0
metrics = {}
for epoch in range(starting_epoch, args.num_train_epochs):
model.train()
if args.with_tracking:
total_loss = 0
if (
args.resume_from_checkpoint
and epoch == starting_epoch
and resume_step is not None
):
# We skip the first `n` batches in the dataloader when resuming from a checkpoint
active_dataloader = accelerator.skip_first_batches(
train_dataloader, resume_step
)
else:
active_dataloader = train_dataloader
for step, batch in enumerate(active_dataloader):
t0step = time.perf_counter()
with accelerator.accumulate(model):
outputs = model(**batch)
loss = outputs.loss
accelerator.backward(loss)
optimizer.step()
lr_scheduler.step()
optimizer.zero_grad()
# Checks if the accelerator has performed an optimization step behind the scenes
if accelerator.sync_gradients:
t1step = time.perf_counter() - t0step
progress_bar.update(1)
completed_steps += 1
if isinstance(checkpointing_steps, int):
if (
completed_steps % checkpointing_steps == 0
and accelerator.sync_gradients
):
output_dir = f"step_{completed_steps}"
if args.output_dir is not None:
output_dir = os.path.join(args.output_dir, output_dir)
accelerator.save_state(output_dir)
if completed_steps % args.logging_steps == 0:
# We keep track of the loss at each epoch
if args.with_tracking:
logs = {
"epoch": epoch,
"step": completed_steps,
"loss": loss.detach().float(),
"learning_rate": lr_scheduler.get_last_lr()[0],
"dts": t1step,
}
accelerator.log(logs, step=completed_steps)
if wandb is not None and wandb.run is not None:
wandb.log(logs)
# logger.info(
# f"epoch {epoch}: step {completed_steps}: loss: {loss.detach().float()}"
# )
if completed_steps >= args.max_train_steps:
break
model.eval()
losses = []
for step, batch in enumerate(eval_dataloader):
with torch.no_grad():
outputs = model(**batch)
loss = outputs.loss
losses.append(
accelerator.gather_for_metrics(
loss.repeat(args.per_device_eval_batch_size)
)
)
losses = torch.cat(losses)
eval_loss = torch.mean(losses)
try:
perplexity = math.exp(eval_loss)
except OverflowError:
perplexity = float("inf")
logger.info(f"epoch {epoch}: perplexity: {perplexity} eval_loss: {eval_loss}")
if args.with_tracking:
accelerator.log(
{
"perplexity": perplexity,
"eval_loss": eval_loss,
"train_loss": total_loss.item() / len(train_dataloader),
"epoch": epoch,
"step": completed_steps,
},
step=completed_steps,
)
if args.push_to_hub and epoch < args.num_train_epochs - 1:
accelerator.wait_for_everyone()
unwrapped_model = accelerator.unwrap_model(model)
unwrapped_model.save_pretrained(
args.output_dir,
is_main_process=accelerator.is_main_process,
save_function=accelerator.save,
)
if accelerator.is_main_process:
tokenizer.save_pretrained(args.output_dir)
api.upload_folder(
commit_message=f"Training in progress epoch {epoch}",
folder_path=args.output_dir,
repo_id=repo_id,
repo_type="model",
token=args.hub_token,
)
if args.checkpointing_steps == "epoch":
output_dir = f"epoch_{epoch}"
if args.output_dir is not None:
output_dir = os.path.join(args.output_dir, output_dir)
accelerator.save_state(output_dir)
if args.with_tracking:
accelerator.end_training()
if args.output_dir is not None:
accelerator.wait_for_everyone()
unwrapped_model = accelerator.unwrap_model(model)
unwrapped_model.save_pretrained(
args.output_dir,
is_main_process=accelerator.is_main_process,
save_function=accelerator.save,
)
if accelerator.is_main_process:
tokenizer.save_pretrained(args.output_dir)
if args.push_to_hub:
api.upload_folder(
commit_message="End of training",
folder_path=args.output_dir,
repo_id=repo_id,
repo_type="model",
token=args.hub_token,
)
with open(os.path.join(args.output_dir, "all_results.json"), "w") as f:
json.dump({"perplexity": perplexity}, f)