riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-07-31 02:02:21 +06:00
Code Issues Actions 23 Packages Projects Releases Wiki Activity

HF <-> megatron checkpoint reshaping and conversion for GPT (#19317)

Browse Source 
* HF <-> megatron checkpoint conversion handling reshaping from different tensor and parallel sizes

* Apply suggestions from code review

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* addressing comments

* add doc strings and  🐛 fixes

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
This commit is contained in:
Sourab Mangrulkar 2022-10-07 19:16:55 +05:30 committed by GitHub
parent 41ec5d0ced
commit 56af8df359
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 900 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

900
src/transformers/models/megatron_gpt2/checkpoint_reshaping_and_interoperability.py Normal file
View File

@ -0,0 +1,900 @@
# Copyright 2022 The HuggingFace 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 json
import os
import re
import sys
import types
import torch
from transformers import AutoTokenizer, GPT2Config
from transformers.modeling_utils import WEIGHTS_INDEX_NAME, WEIGHTS_NAME, shard_checkpoint
def add_checkpointing_args(parser):
parser.add_argument("--megatron-path", type=str, default=None, help="Base directory of Megatron repository")
parser.add_argument(
"--convert_checkpoint_from_megatron_to_transformers",
action="store_true",
help=(
"If True, convert a Megatron checkpoint to a Transformers checkpoint. "
"If False, convert a Transformers checkpoint to a Megatron checkpoint."
),
)
parser.add_argument(
"--load_path",
type=str,
required=True,
help="Path to the checkpoint to convert.",
)
parser.add_argument(
"--save_path",
type=str,
required=True,
help="Path to the converted checkpoint.",
)
parser.add_argument("--print-checkpoint-structure", action="store_true")
return parser
def add_megatron_checkpoint_args(parser):
parser.add_argument(
"--target_tensor_model_parallel_size",
type=int,
default=1,
help=(
"The tensor model parallel size of the converted checkpoint. "
"Only used when converting a Transformers checkpoint to a Megatron checkpoint."
),
)
parser.add_argument(
"--target_pipeline_model_parallel_size",
type=int,
default=1,
help=(
"The pipeline model parallel size of the converted checkpoint. "
"Only used when converting a Transformers checkpoint to a Megatron checkpoint."
),
)
parser.add_argument(
"--target_data_parallel_size",
type=int,
default=1,
help=(
"The data parallel size of the converted checkpoint. "
"Only used when converting a Transformers checkpoint to a Megatron checkpoint."
),
)
parser.add_argument(
"--target_params_dtype",
type=str,
default="fp32",
help=(
"The dtype of the converted checkpoint. "
"Only used when converting a Transformers checkpoint to a Megatron checkpoint."
),
)
parser.add_argument(
"--make_vocab_size_divisible_by",
type=int,
default=128,
help=(
"Pad the vocab size to be divisible by this value. "
"This is added for computational efficieny reasons. "
"Only used when converting a Transformers checkpoint to a Megatron checkpoint."
),
)
parser.add_argument(
"--use_distributed_optimizer",
action="store_true",
help=(
"If True, use the distributed optimizer. "
"Only used when converting a Transformers checkpoint to a Megatron checkpoint."
),
)
return parser
def add_transformers_checkpoint_args(parser):
parser.add_argument(
"--tokenizer_name",
type=str,
default=None,
help=(
"The name of the pre-trained tokenizer to save. "
"If not None, the tokenizer will be saved. "
"Only used when converting a Megatron checkpoint to a Transformers checkpoint."
),
)
parser.add_argument(
"--max_shard_size",
type=str,
default="10GB",
help=(
"The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size "
"lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`). "
"Only used when converting a Megatron checkpoint to a Transformers checkpoint."
),
)
return parser
# The simple map of names for "automated" rules.
megatron_to_transformers = {
"attention.dense": ".attn.c_proj.",
"self_attention.dense": ".attn.c_proj.",
"mlp.dense_h_to_4h": ".mlp.c_fc.",
"mlp.dense_4h_to_h": ".mlp.c_proj.",
}
transformers_to_megatron = {v[1:-1]: k for k, v in megatron_to_transformers.items()}
tensor_parallel_params = [
# megatron-lm layers to merge across tp ranks
"self_attention.query_key_value.weight",
"self_attention.query_key_value.bias",
"self_attention.dense.weight",
"mlp.dense_h_to_4h.weight",
"mlp.dense_h_to_4h.bias",
"mlp.dense_4h_to_h.weight",
# deprecated
"attention.query_key_value.weight",
"attention.query_key_value.bias",
"attention.dense.weight",
# transformers layers to split across tp ranks
"attn.c_attn.weight",
"attn.c_attn.bias",
"attn.c_proj.weight",
"mlp.c_fc.weight",
"mlp.c_fc.bias",
"mlp.c_proj.weight",
]
def recursive_print(name, val, spaces=0):
"""
Recursively print the structure of a checkpoint. This function is taken from `convert_megatron_gpt2_checkpoint.py`
Args:
name (str): the name of the current tensor parameter
val (Tuple(int)): the shape of the current tensor parameter
spaces (int): the number of spaces to print before the output for a nested structure
"""
# Format the message.
if name is None:
msg = None
else:
fmt = "." * max(0, spaces - 2) + "# {:" + str(50 - spaces) + "s}"
msg = fmt.format(name)
# Print and recurse (if needed).
if isinstance(val, dict):
if msg is not None:
print(msg)
for k in val.keys():
recursive_print(k, val[k], spaces + 2)
elif isinstance(val, torch.Tensor):
print(msg, ":", val.size())
else:
print(msg, ":", val)
def megatron_to_transformers_fix_query_key_value_ordering(
param, checkpoint_version, num_splits, num_heads, hidden_size
):
"""
Permutes layout of param tensor to [num_splits * num_heads * hidden_size, :] for compatibility with later versions
of NVIDIA Megatron-LM. The inverse operation is performed inside Megatron-LM to read checkpoints:
https://github.com/NVIDIA/Megatron-LM/blob/v2.4/megatron/checkpointing.py#L209 If param is the weight tensor of the
self-attention block, the returned tensor will have to be transposed one more time to be read by HuggingFace GPT2.
This function is taken from `convert_megatron_gpt2_checkpoint.py`
Args:
param (torch.Tensor): the tensor to permute
checkpoint_version (int): the version of the checkpoint.
num_splits (int): the number of projections, usually 3 for (Query, Key, Value)
num_heads (int): the number of attention heads
hidden_size (int): the hidden size per head
"""
input_shape = param.size()
if checkpoint_version == 1.0:
# version 1.0 stores [num_heads * hidden_size * num_splits, :]
saved_shape = (num_heads, hidden_size, num_splits) + input_shape[1:]
param = param.view(*saved_shape)
param = param.transpose(0, 2)
param = param.transpose(1, 2).contiguous()
elif checkpoint_version >= 2.0:
# other versions store [num_heads * num_splits * hidden_size, :]
saved_shape = (num_heads, num_splits, hidden_size) + input_shape[1:]
param = param.view(*saved_shape)
param = param.transpose(0, 1).contiguous()
param = param.view(*input_shape)
return param
def transformers_to_megatron_fix_query_key_value_ordering(
param, checkpoint_version, num_splits, num_heads, hidden_size
):
"""
Permutes layout of param tensor to the one compatible with respective NVIDIA Megatron-LM chekpoint versions. Input
is [num_splits * num_heads * hidden_size, :] and output is [num_heads * hidden_size * num_splits, :] for version
1.0 and [num_heads * num_splits * hidden_size, :] for version 2.0 and later. If param is the weight tensor of the
self-attention block, the param needs to be already transposed before calling this function.
Args:
param (torch.Tensor): the tensor to permute
checkpoint_version (int): the version of the checkpoint.
num_splits (int): the number of projections, usually 3 for (Query, Key, Value)
num_heads (int): the number of attention heads
hidden_size (int): the hidden size per head
"""
# Input is [num_splits * num_heads * hidden_size, :]
input_shape = param.size()
if checkpoint_version == 1.0:
# version 1.0 stores [num_heads * hidden_size * num_splits, :]
current_shape = (num_splits, num_heads, hidden_size) + input_shape[1:]
param = param.view(*current_shape)
param = param.transpose(0, 2)
param = param.transpose(1, 2).contiguous()
elif checkpoint_version >= 2.0:
# other versions store [num_heads * num_splits * hidden_size, :]
current_shape = (num_splits, num_heads, hidden_size) + input_shape[1:]
param = param.view(*current_shape)
param = param.transpose(0, 1).contiguous()
param = param.view(*input_shape)
return param
def merge_transformers_sharded_states(path, num_checkpoints):
"""
Merge sharded checkpoints from transformers into a single checkpoint.
Args:
path (str): the path to the sharded checkpoints
num_checkpoints (int): the number of checkpoints to merge
"""
state_dict = {}
for i in range(1, num_checkpoints + 1):
checkpoint_path = os.path.join(path, f"pytorch_model-{i:05d}-of-{num_checkpoints:05d}.bin")
current_chunk = torch.load(checkpoint_path, map_location="cpu")
state_dict.update(current_chunk)
return state_dict
def get_megatron_sharded_states(args, tp_size, pp_size, pp_rank):
"""
Get sharded checkpoints from NVIDIA Megatron-LM checkpoint based on the provided tensor parallel size, pipeline
parallel size and pipeline parallel rank.
Args:
args (argparse.Namespace): the arguments to the script
tp_size (int): the tensor parallel size
pp_size (int): the pipeline parallel size
pp_rank (int): the pipeline parallel rank
"""
tp_state_dicts = []
for i in range(tp_size):
sub_dir_name = f"mp_rank_{i:02d}" if pp_size == 1 else f"mp_rank_{i:02d}_{pp_rank:03d}"
checkpoint_name = os.listdir(os.path.join(args.load_path, sub_dir_name))[0]
checkpoint_path = os.path.join(args.load_path, sub_dir_name, checkpoint_name)
state_dict = torch.load(checkpoint_path, map_location="cpu")
tp_state_dicts.append(state_dict)
return tp_state_dicts
def get_element_from_dict_by_path(d, path):
"""
Get element from dictionary by path. If element is not present, recursively add empty dictionaries.
Args:
d (dict): the dictionary to get the element from
path (list): the path to the element which is delimited by "."
"""
path = path.split(".")
for k in path:
if k not in d:
d[k] = {}
d = d[k]
return d
def convert_checkpoint_from_megatron_to_transformers(args):
"""
Convert NVIDIA Megatron-LM checkpoint to HuggingFace Transformers checkpoint. This handles Megatron checkpoints
with different tensor parallelism and pipeline parallelism sizes. It saves the converted checkpoint into shards
using HuggingFace Transformers checkpoint sharding functionality. This greatly extends the functionality of
`convert_megatron_gpt2_checkpoint.py`
Args:
args (argparse.Namespace): the arguments to the script
"""
# Load Megatron-LM checkpoint arguments from the state dict
sub_dirs = os.listdir(args.load_path)
possible_sub_dirs = ["mp_rank_00", "mp_rank_00_000"]
for sub_dir in possible_sub_dirs:
if sub_dir in sub_dirs:
rank0_checkpoint_name = os.listdir(os.path.join(args.load_path, sub_dir))[0]
rank0_checkpoint_path = os.path.join(args.load_path, sub_dir, rank0_checkpoint_name)
break
print(f"Loading Megatron-LM checkpoint arguments from: {rank0_checkpoint_path}")
state_dict = torch.load(rank0_checkpoint_path, map_location="cpu")
megatron_args = state_dict.get("args", None)
if megatron_args is None:
raise ValueError(
"Megatron-LM checkpoint does not contain arguments. This utility only supports Megatron-LM checkpoints"
" containing all the megatron arguments. This is because it loads all config related to model"
" architecture, the tensor and pipeline model parallel size from the checkpoint insead of user having to"
" manually specify all the details. Please save Megatron-LM checkpoint along with all the megatron"
" arguments to use this utility."
)
# Create Transformers GPT2 config from Megatron-LM arguments
if megatron_args is not None:
if megatron_args.bias_gelu_fusion:
activation_function = "gelu_fast"
elif megatron_args.openai_gelu:
activation_function = "gelu_new"
else:
activation_function = "gelu"
else:
# in the very early days this used to be "gelu_new"
activation_function = "gelu_new"
vocab_size = (
megatron_args.padded_vocab_size
if getattr(megatron_args, "orig_vocab_size", None) is None
else megatron_args.orig_vocab_size
)
print(vocab_size)
config = GPT2Config(
vocab_size=vocab_size,
n_positions=megatron_args.max_position_embeddings,
n_embd=megatron_args.hidden_size,
n_layer=megatron_args.num_layers,
n_head=megatron_args.num_attention_heads,
n_inner=megatron_args.ffn_hidden_size,
activation_function=activation_function,
resid_pdrop=0.1,
embd_pdrop=0.1,
attn_pdrop=0.1,
layer_norm_epsilon=1e-5,
initializer_range=0.02,
summary_type="cls_index",
summary_use_proj=True,
summary_activation=None,
summary_proj_to_labels=True,
summary_first_dropout=0.1,
scale_attn_weights=True,
use_cache=True,
bos_token_id=vocab_size - 1,
eos_token_id=vocab_size - 1,
architectures=["GPT2LMHeadModel"],
)
output_state_dict = {}
checkpoint_version = state_dict.get("checkpoint_version", 0.0)
tp_size = megatron_args.tensor_model_parallel_size
pp_size = megatron_args.pipeline_model_parallel_size
dtype = torch.float32
# The regex to extract layer names.
layer_re = re.compile("layers\.(\d+)\.([a-z0-9_.]+)\.([a-z]+)")
# Convert.
print("Converting")
# Embeddings
print("Converting embeddings")
tp_state_dicts = get_megatron_sharded_states(args, tp_size, pp_size, 0)
# Convert and store the position embeddings.
position_embeddings = get_element_from_dict_by_path(
tp_state_dicts[0], "model.language_model.embedding.position_embeddings.weight"
)
output_state_dict["transformer.wpe.weight"] = position_embeddings.to(dtype)
# Convert and store the word embeddings.
word_embeddings = torch.cat(
[
get_element_from_dict_by_path(
tp_state_dicts[tp_rank], "model.language_model.embedding.word_embeddings.weight"
)
for tp_rank in range(tp_size)
],
dim=0,
)
word_embeddings = word_embeddings[:vocab_size].to(dtype)
output_state_dict["transformer.wte.weight"] = word_embeddings
# Transformer Layers
print("Converting transformer layers")
# The number of heads.
heads = config.n_head
# The hidden_size per head.
hidden_size_per_head = config.n_embd // config.n_head
n_positions = config.n_positions
num_layers = config.num_hidden_layers // pp_size
for pp_rank in range(pp_size):
if pp_size > 0:
print(f"Converting pipeline parallel rank {pp_rank}")
tp_state_dicts = get_megatron_sharded_states(args, tp_size, pp_size, pp_rank)
# The transformer.
path = (
"model.language_model.transformer"
if "transformer" in get_element_from_dict_by_path(tp_state_dicts[0], "model.language_model").keys()
else "model.language_model.encoder"
)
# Extract the layers.
for key, val in get_element_from_dict_by_path(tp_state_dicts[0], path).items():
# Match the name.
m = layer_re.match(key)
# Stop if that's not a layer
if m is None:
break
# The index of the layer.
layer_idx = int(m.group(1)) + pp_rank * num_layers
# The name of the operation.
op_name = m.group(2)
# Is it a weight or a bias?
weight_or_bias = m.group(3)
# The name of the layer.
layer_name = f"transformer.h.{layer_idx}"
if op_name + "." + weight_or_bias not in tensor_parallel_params:
params = val.to(dtype)
else:
dim = 1 if op_name in ["self_attention.dense", "mlp.dense_4h_to_h", "attention.dense"] else 0
params = torch.cat(
[val]
+ [
get_element_from_dict_by_path(tp_state_dicts[tp_rank], f"{path}")[key]
for tp_rank in range(1, tp_size)
],
dim=dim,
).to(dtype)
# For layernorm(s), simply store the layer norm.
if op_name.endswith("layernorm"):
ln_name = "ln_1" if op_name.startswith("input") else "ln_2"
output_state_dict[layer_name + "." + ln_name + "." + weight_or_bias] = params
# Transpose the QKV matrix.
elif (
op_name == "attention.query_key_value" or op_name == "self_attention.query_key_value"
) and weight_or_bias == "weight":
# Insert a tensor of 1x1xDxD bias.
causal_mask = torch.tril(torch.ones((n_positions, n_positions), dtype=dtype)).view(
1, 1, n_positions, n_positions
)
output_state_dict[layer_name + ".attn.bias"] = causal_mask
# Insert a "dummy" tensor for masked_bias.
masked_bias = torch.tensor(-1e4, dtype=dtype)
output_state_dict[layer_name + ".attn.masked_bias"] = masked_bias
out_val = megatron_to_transformers_fix_query_key_value_ordering(
params,
checkpoint_version,
3,
heads,
hidden_size_per_head,
)
# Megatron stores (3*D) x D but transformers-GPT2 expects D x 3*D.
out_val = out_val.transpose(0, 1).contiguous()
# Store.
output_state_dict[layer_name + ".attn.c_attn.weight"] = out_val
# Transpose the bias.
elif (
op_name == "attention.query_key_value" or op_name == "self_attention.query_key_value"
) and weight_or_bias == "bias":
out_val = megatron_to_transformers_fix_query_key_value_ordering(
params, checkpoint_version, 3, heads, hidden_size_per_head
)
# Store. No change of shape.
output_state_dict[layer_name + ".attn.c_attn.bias"] = out_val
# Transpose the weights.
elif weight_or_bias == "weight":
out_name = megatron_to_transformers[op_name]
output_state_dict[layer_name + out_name + "weight"] = params.transpose(0, 1)
# Copy the bias.
elif weight_or_bias == "bias":
out_name = megatron_to_transformers[op_name]
output_state_dict[layer_name + out_name + "bias"] = params
if config.n_layer != (layer_idx + 1):
raise ValueError(f"Expected {config.n_layer} layers but found {layer_idx + 1}")
# The final layernorm.
print("Converting final layernorm")
params = get_element_from_dict_by_path(tp_state_dicts[0], str(path))
output_state_dict["transformer.ln_f.weight"] = params["final_layernorm.weight"].to(dtype)
output_state_dict["transformer.ln_f.bias"] = params["final_layernorm.bias"].to(dtype)
# For LM head, transformers' wants the matrix to weight embeddings.
print("Converting LM head")
output_state_dict["lm_head.weight"] = word_embeddings.to(dtype)
# It should be done!
print("Conversion from Megatron-LM to Transformers is done!")
# Print the structure of converted state dict.
if args.print_checkpoint_structure:
recursive_print(None, output_state_dict)
# Add tokenizer class info to config
# see https://github.com/huggingface/transformers/issues/13906)
if args.tokenizer_name is None:
tokenizer_name = "gpt2"
else:
tokenizer_name = args.tokenizer_name
tokenizer = AutoTokenizer.from_pretrained(tokenizer_name)
tokenizer_class = type(tokenizer).__name__
config.tokenizer_class = tokenizer_class
# Store the config to file.
print("Saving config")
config.save_pretrained(args.save_path)
# Save tokenizer based on args
if args.tokenizer_name is not None:
print(f"Adding {tokenizer_class} tokenizer files")
tokenizer.save_pretrained(args.save_path)
# Store the state_dict to file.
max_shard_size = int(args.max_shard_size) if args.max_shard_size.isdigit() else args.max_shard_size
shards, index = shard_checkpoint(output_state_dict, max_shard_size=max_shard_size)
# Save the model
for shard_file, shard in shards.items():
torch.save(shard, os.path.join(args.save_path, shard_file))
if index is None:
print(f"Model weights saved in {os.path.join(args.save_path, WEIGHTS_NAME)}")
else:
save_index_file = os.path.join(args.save_path, WEIGHTS_INDEX_NAME)
# Save the index as well
with open(save_index_file, "w", encoding="utf-8") as f:
content = json.dumps(index, indent=2, sort_keys=True) + "\n"
f.write(content)
print(
f"The model is bigger than the maximum size per checkpoint ({args.max_shard_size}) and is going to be "
f"split in {len(shards)} checkpoint shards. You can find where each parameters has been saved in the "
f"index located at {save_index_file}."
)
def convert_checkpoint_from_transformers_to_megatron(args):
"""
Convert a checkpoint from HuggingFace Transformers to Megatron-LM. This allows converted checkpoints with variable
tensor parallelism and pipeline parallelism sizes. It takes as input a checkpoint from HuggingFace Transformers
which can have multiple shards.
Args:
args (argparse.Namespace): the arguments to the script
"""
os.makedirs(args.save_path, exist_ok=True)
# Search in directory above this
sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), os.path.pardir)))
if args.megatron_path is not None:
sys.path.insert(0, args.megatron_path)
try:
from megatron.tokenizer.tokenizer import _vocab_size_with_padding
except ModuleNotFoundError:
print("Unable to import Megatron, please specify the path to Megatron using --megatron-path. Exiting.")
exit(1)
# load the transformers model state dict and config
sub_dirs = [x for x in os.listdir(args.load_path) if x.startswith("pytorch_model")]
if len(sub_dirs) == 1:
checkpoint_name = "pytorch_model.bin"
state_dict = torch.load(os.path.join(args.load_path, checkpoint_name), map_location="cpu")
else:
num_checkpoints = len(sub_dirs) - 1
state_dict = merge_transformers_sharded_states(args.load_path, num_checkpoints)
config = GPT2Config.from_pretrained(args.load_path)
# Saving the tracker file
tracker_filepath = os.path.join(args.save_path, "latest_checkpointed_iteration.txt")
with open(tracker_filepath, "w") as f:
f.write("release")
# create `release` dir in args.load_path
release_dir = os.path.join(args.save_path, "release")
os.makedirs(release_dir, exist_ok=True)
# megatron args
megatron_args = {
"orig_vocab_size": config.vocab_size,
"max_position_embeddings": config.n_positions,
"hidden_size": config.n_embd,
"num_layers": config.n_layer,
"num_attention_heads": config.n_head,
"ffn_hidden_size": config.n_inner,
"tensor_model_parallel_size": args.target_tensor_model_parallel_size,
"pipeline_model_parallel_size": args.target_pipeline_model_parallel_size,
"data_parallel_size": args.target_data_parallel_size,
"make_vocab_size_divisible_by": args.make_vocab_size_divisible_by,
"rank": 0,
"tokenizer_type": None,
}
if config.activation_function == "gelu":
megatron_args["bias_gelu_fusion"] = False
megatron_args["openai_gelu"] = False
elif config.activation_function == "gelu_fast":
megatron_args["bias_gelu_fusion"] = True
megatron_args["openai_gelu"] = False
elif config.activation_function == "gelu_new":
megatron_args["bias_gelu_fusion"] = False
megatron_args["openai_gelu"] = True
margs = types.SimpleNamespace()
for k, v in megatron_args.items():
setattr(margs, k, v)
# params dtype
if args.target_params_dtype == "fp16":
dtype = torch.float16
elif args.target_params_dtype == "bf16":
dtype = torch.bfloat16
else:
dtype = torch.float32
setattr(margs, "params_dtype", dtype)
# save dummy optim state dict
dummy_optim_state_dict = {}
dummy_optim_state_dict["optimizer"] = {
"step": 0,
"param_groups": [
{
"lr": 0.0,
"beta1": 0.0,
"beta2": 0.0,
"eps": 0.0,
"weight_decay": 0.0,
"correct_bias": False,
"params": [],
}
],
}
if args.use_distributed_optimizer:
for i in range(args.target_pipeline_model_parallel_size):
for j in range(args.target_tensor_model_parallel_size):
for k in range(args.target_data_parallel_size):
if args.target_pipeline_model_parallel_size == 1:
checkpoint_dir = f"mp_rank_{i:02d}_{k:03d}"
else:
checkpoint_dir = f"mp_rank_{i:02d}_{j:03d}_{k:03d}"
checkpoint_dir = os.path.join(release_dir, checkpoint_dir)
os.makedirs(checkpoint_dir, exist_ok=True)
torch.save(
dummy_optim_state_dict,
os.path.join(checkpoint_dir, "optim.pt"),
)
# Convert.
print("Converting")
output_state_dict = []
for i in range(args.target_tensor_model_parallel_size):
output_state_dict.append({})
# Embedding layer
print("converting embedding layer")
pos_embedding = state_dict["transformer.wpe.weight"].to(dtype)
word_embedding = state_dict["transformer.wte.weight"].to(dtype)
orig_vocab_size = config.vocab_size
padded_vocab_size = _vocab_size_with_padding(orig_vocab_size, margs)
setattr(margs, "padded_vocab_size", padded_vocab_size)
# Cut out extra padding we don't need
if orig_vocab_size > padded_vocab_size:
full_word_embed = word_embedding[0:padded_vocab_size, :]
# Expanding embedding to larger size by replicating final entry
elif orig_vocab_size < padded_vocab_size:
padding_size = padded_vocab_size - orig_vocab_size
full_word_embed = torch.cat((word_embedding, word_embedding[-1].unsqueeze(0).expand(padding_size, -1)))
# Same size!
else:
full_word_embed = word_embedding
# Split into new tensor model parallel sizes
out_word_embed = torch.chunk(full_word_embed, args.target_tensor_model_parallel_size, dim=0)
for i in range(args.target_tensor_model_parallel_size):
pos_emb_dict = get_element_from_dict_by_path(
output_state_dict[i], "model.language_model.embedding.position_embeddings"
)
pos_emb_dict["weight"] = pos_embedding
word_emb_dict = get_element_from_dict_by_path(
output_state_dict[i], "model.language_model.embedding.word_embeddings"
)
word_emb_dict["weight"] = out_word_embed[i]
# Transformer layers
print("converting transformer layers")
if config.num_hidden_layers % args.target_tensor_model_parallel_size != 0:
raise ValueError(
f"Number of layers ({config.num_hidden_layers}) must be divisible by number of tensor parallelism"
f" ({args.target_tensor_model_parallel_size})"
)
num_layers = config.num_hidden_layers // args.target_pipeline_model_parallel_size
layer_re = re.compile("transformer.h\.(\d+)\.([a-z0-9_.]+)\.([a-z]+)")
# The number of heads.
heads = config.n_head
# The hidden_size per head.
hidden_size_per_head = config.n_embd // config.n_head
for pp_rank in range(args.target_pipeline_model_parallel_size):
layer_offset = pp_rank * num_layers
if pp_rank > 0:
output_state_dict = []
for i in range(args.target_tensor_model_parallel_size):
output_state_dict.append({})
for layer in range(num_layers):
pp_layer_id = layer + layer_offset
layers_to_copy = [
layer_name
for layer_name in state_dict.keys()
if layer_name.startswith(f"transformer.h.{pp_layer_id}.")
]
for layer_name in layers_to_copy:
m = layer_re.match(layer_name)
# Stop if that's not a layer
if m is None:
break
# The index of the layer.
_ = int(m.group(1))
# The name of the operation.
op_name = m.group(2)
# Is it a weight or a bias?
weight_or_bias = m.group(3)
params = state_dict[layer_name].to(dtype)
# handle layernorm
if op_name.startswith("ln"):
out_name = "input_layernorm" if op_name.endswith("1") else "post_attention_layernorm"
layer_name = f"layers.{layer}.{out_name}.{weight_or_bias}"
# handle attention K, V, Q weights
elif op_name.startswith("attn.c_attn") and weight_or_bias == "weight":
# transformers stores D X (3*D) but Megatron-LM expects (3*D) X D.
params = params.transpose(0, 1).contiguous()
params = transformers_to_megatron_fix_query_key_value_ordering(
params,
3.0,
3,
heads,
hidden_size_per_head,
)
layer_name = f"layers.{layer}.self_attention.query_key_value.{weight_or_bias}"
# handle attention K, V, Q bias
elif op_name.startswith("attn.c_attn") and weight_or_bias == "bias":
params = transformers_to_megatron_fix_query_key_value_ordering(
params,
3.0,
3,
heads,
hidden_size_per_head,
)
layer_name = f"layers.{layer}.self_attention.query_key_value.{weight_or_bias}"
# handle attention and mlp weights
elif weight_or_bias == "weight":
out_name = transformers_to_megatron.get(op_name, None)
if out_name is None:
continue
params = params.transpose(0, 1)
layer_name = f"layers.{layer}.{out_name}.{weight_or_bias}"
# handle attention and mlp bias
elif weight_or_bias == "bias":
out_name = transformers_to_megatron.get(op_name, None)
if out_name is None:
continue
layer_name = f"layers.{layer}.{out_name}.{weight_or_bias}"
# skip
else:
continue
if op_name + "." + weight_or_bias in tensor_parallel_params:
dim = 1 if op_name in ["attn.c_proj", "mlp.c_proj"] else 0
params = torch.chunk(params, args.target_tensor_model_parallel_size, dim=dim)
for i in range(args.target_tensor_model_parallel_size):
params_dict = get_element_from_dict_by_path(output_state_dict[i], "model.language_model.encoder")
params_dict[layer_name] = (
params[i] if (op_name + "." + weight_or_bias in tensor_parallel_params) else params
)
if pp_rank == args.target_pipeline_model_parallel_size - 1:
# handle final layernorm
for weight_or_bias in ["weight", "bias"]:
params = state_dict[f"transformer.ln_f.{weight_or_bias}"].to(dtype)
layer_name = f"final_layernorm.{weight_or_bias}"
for i in range(args.target_tensor_model_parallel_size):
params_dict = get_element_from_dict_by_path(output_state_dict[i], "model.language_model.encoder")
params_dict[layer_name] = params
# add the LM head
for i in range(args.target_tensor_model_parallel_size):
params_dict = get_element_from_dict_by_path(output_state_dict[i], "model.word_embeddings_for_head")
params_dict["weight"] = out_word_embed[i]
# saving the state dict as per the tp_rank and pp_rank
for tp_rank in range(args.target_tensor_model_parallel_size):
output_state_dict[tp_rank]["checkpoint_version"] = 3.0
output_state_dict[tp_rank]["args"] = margs
checkpoint_dir = (
f"mp_rank_{tp_rank:02d}"
if args.target_pipeline_model_parallel_size == 1
else f"mp_rank_{tp_rank:02d}_{pp_rank:03d}"
)
if args.use_distributed_optimizer:
checkpoint_name = "model_rng.pt"
else:
checkpoint_name = "model_optim_rng.pt"
output_state_dict[tp_rank]["optimizer"] = dummy_optim_state_dict["optimizer"]
checkpoint_dir = os.path.join(release_dir, checkpoint_dir)
os.makedirs(checkpoint_dir, exist_ok=True)
checkpoint_path = os.path.join(checkpoint_dir, checkpoint_name)
if args.print_checkpoint_structure:
print(
f"Checkpoint structure of model state dict shard belonging to TP rank {tp_rank} and PP rank"
f" {pp_rank}:"
)
recursive_print(None, output_state_dict[tp_rank])
torch.save(output_state_dict[tp_rank], checkpoint_path)
def main():
parser = argparse.ArgumentParser()
parser = add_checkpointing_args(parser)
parser = add_megatron_checkpoint_args(parser)
parser = add_transformers_checkpoint_args(parser)
args = parser.parse_args()
if args.convert_checkpoint_from_megatron_to_transformers:
convert_checkpoint_from_megatron_to_transformers(args)
else:
convert_checkpoint_from_transformers_to_megatron(args)
if __name__ == "__main__":
main()