# Copyright 2023 Rohan Taori, Ishaan Gulrajani, Tianyi Zhang, Yann Dubois, Xuechen Li

# 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 os

from os.path import exists, join, isdir

import gc

import json

import math

import random

import copy

from copy import deepcopy

import logging

from dataclasses import dataclass, field

from typing import Dict, Optional, Sequence, Callable, List, Tuple, Union, Any

import torch

from torch import nn

from torch.utils.data import Dataset

import bitsandbytes as bnb

import transformers
from transformers import Trainer, BitsAndBytesConfig, set_seed

from transformers.trainer_utils import PREFIX_CHECKPOINT_DIR

from peft import LoraConfig, get_peft_model, prepare_model_for_kbit_training

from peft.tuners.lora import LoraLayer

from datasets import load_dataset

from openrag.configuration_openrag import OpenRAGConfig

from openrag.modeling_openrag import LlamaForCausalLM

import warnings

warnings.filterwarnings("ignore")

from openrag.transformers_utils import (

get_keys_to_not_convert,

_load_pretrained_model,

import transformers.integrations

import transformers.modeling_utils

from ipdb import set_trace as stt

transformers.integrations.get_keys_to_not_convert = get_keys_to_not_convert

transformers.modeling_utils.PreTrainedModel._load_pretrained_model = (

_load_pretrained_model

import GPUtil

IGNORE_INDEX = -100

DEFAULT_PAD_TOKEN = "[PAD]"
context_markups = []

@dataclass

class ModelArguments:

model_name_or_path: Optional[str] = field(default="facebook/opt-125m")

lora_r: int = field(

default=64,

metadata={"help": "Rank of the LoRA update matrices"}

lora_alpha: int = field(

default=16,

metadata={"help": "Scaling factor for LoRA"}

adapter_dim: int = field(

default=64,

metadata={"help": "Dimension of the adapter layers"}

moe_scaling: float = field(

default=1.0,

metadata={"help": "Scaling factor for Mixture of Experts"}

num_experts: int = field(

default=8,

metadata={"help": "Number of experts in the Mixture of Experts layer"}

topk: int = field(

default=2,

metadata={"help": "Top-k value for routing or selection"}

)
@dataclass

class DataArguments:

data_path: str = field(

default=None, metadata={"help": "Path to the training data."}

data_subset: str = field(

default="default", metadata={"help": "Training data subset."}

@dataclass

class TrainingArguments(transformers.TrainingArguments):

report_to: str = field(default="none")

cache_dir: Optional[str] = field(default=None)

optim: str = field(

default="paged_adamw_32bit"

) # "paged_lion_8bit", "paged_adamw_8bit", "paged_lion_32bit", "paged_adamw_32bit"

lr_scheduler_type: str = field(

default="constant_with_warmup"

) # "constant", "constant_with_warmup", "cosine", "cosine_with_restarts", "linear"

model_max_length: int = field(

default=2048,

metadata={

"help": "Maximum sequence length. Sequences will be right padded (and possibly
truncated)."

},

)
def _tokenize_fn(

strings: Sequence[str], tokenizer: transformers.PreTrainedTokenizer

) -> Dict:

"""Tokenize a list of strings."""

tokenized_list = [

tokenizer(

text,

return_tensors="pt",

padding="longest",

max_length=tokenizer.model_max_length,

truncation=True,

for text in strings

input_ids = labels = [tokenized.input_ids[0] for tokenized in tokenized_list]

input_ids_lens = labels_lens = [

tokenized.input_ids.ne(tokenizer.pad_token_id).sum().item()

for tokenized in tokenized_list

return dict(

input_ids=input_ids,

labels=labels,

input_ids_lens=input_ids_lens,

labels_lens=labels_lens,

def preprocess(
 sources: Sequence[str],

targets: Sequence[str],

tokenizer: transformers.PreTrainedTokenizer,

) -> Dict:

"""Preprocess the data by tokenizing."""

examples = [

s+t

for s, t in zip(sources, targets)

examples_tokenized, sources_tokenized = [

_tokenize_fn(strings, tokenizer) for strings in (examples, sources)

input_ids = examples_tokenized["input_ids"]

labels = copy.deepcopy(input_ids)

for label, source_len in zip(labels, sources_tokenized["input_ids_lens"]):

label[:source_len] = IGNORE_INDEX

global context_markups

context_start = False

for j, orig_token in enumerate(label[source_len:]):

if context_start is False and orig_token == context_markups[0]:

context_start = True

assert label[source_len + j] == context_markups[0]

start_idx = j + source_len

end_idx = None

for k, orig_token_2 in enumerate(label[start_idx:]):

if orig_token_2 == context_markups[1]:

end_idx = start_idx + k

if end_idx is None:

end_idx = start_idx + k

else:
 assert label[end_idx] == context_markups[1]

label[start_idx + 1 : end_idx] = IGNORE_INDEX

context_start = False

return dict(input_ids=input_ids, labels=labels)

class SupervisedDataset(Dataset):

"""Dataset for supervised fine-tuning."""

def __init__(self, data_path: str, data_subset: str, tokenizer: transformers.PreTrainedTokenizer):

super(SupervisedDataset, self).__init__()

logging.warning("Loading data: {}".format(data_path))

dataset = load_dataset(data_path, data_subset)

data_list = dataset['train'].to_list()

# Preprocess Data

logging.warning("Processing data")

self.tokenizer = tokenizer

self.sources = []

self.targets = []

for idx in range(len(data_list)):

data = data_list[idx]

corpus = data["corpus"]

if corpus != "":

# pretrain mode

source = f"{tokenizer.bos_token}"

self.sources.append(source)
 target = f"{corpus}{tokenizer.eos_token}"

self.targets.append(target)

else:

# instruction mode

instruction = data["instruction"]

conversation = data["conversation"]

if len(conversation) == 1:

source = ""

source += (

f"### Instruction:\n{conversation[0]['input']}\n\n### Response:\n"

self.sources.append(source)

target = f"{conversation[0]['output']}{tokenizer.eos_token}"

self.targets.append(target)

del data_list

gc.collect()

logging.warning("there are {} data in dataset".format(len(self.sources)))

def __len__(self):

return len(self.sources)

def __getitem__(self, i):

source = [self.sources[i]]

target = [self.targets[i]]

data_dict = preprocess(source, target, self.tokenizer)

input_ids = data_dict["input_ids"][0]

labels = data_dict["labels"][0]
 return dict(input_ids=input_ids, labels=labels)

@dataclass

class DataCollatorForSupervisedDataset(object):

"""Collate examples for supervised fine-tuning."""

tokenizer: transformers.PreTrainedTokenizer

def __call__(self, instances: Sequence[Dict]) -> Dict[str, torch.Tensor]:

input_ids, labels = tuple(

[instance[key] for instance in instances] for key in ("input_ids", "labels")

input_ids = torch.nn.utils.rnn.pad_sequence(

input_ids, batch_first=True, padding_value=self.tokenizer.pad_token_id

labels = torch.nn.utils.rnn.pad_sequence(

labels, batch_first=True, padding_value=IGNORE_INDEX

return dict(

input_ids=input_ids,

labels=labels,

attention_mask=input_ids.ne(self.tokenizer.pad_token_id),

class SavePeftModelCallback(transformers.TrainerCallback):

def save_model(self, args, state, kwargs):

# print('Saving PEFT checkpoint...')

if state.best_model_checkpoint is not None:

checkpoint_folder = os.path.join(
 state.best_model_checkpoint, "adapter_model"

else:

checkpoint_folder = os.path.join(

args.output_dir, f"{PREFIX_CHECKPOINT_DIR}-{state.global_step}"

peft_model_path = os.path.join(checkpoint_folder, "adapter_model")

model = kwargs["model"]

model.save_pretrained(peft_model_path)

moe_state = {}

for param_tensor in model.state_dict():

if "adapter" in param_tensor:

moe_state.update({param_tensor: model.state_dict()[param_tensor]})

moe_model_path = os.path.join(checkpoint_folder, "moe_model.bin")

torch.save(moe_state, moe_model_path)

pytorch_model_path = os.path.join(checkpoint_folder, "pytorch_model.bin")

if os.path.exists(pytorch_model_path):

os.remove(pytorch_model_path)

def on_save(self, args, state, control, **kwargs):

self.save_model(args, state, kwargs)

return control

def on_train_end(self, args, state, control, **kwargs):

def touch(fname, times=None):

with open(fname, "a"):

os.utime(fname, times)
 touch(join(args.output_dir, "completed"))

self.save_model(args, state, kwargs)

def make_supervised_data_module(

tokenizer: transformers.PreTrainedTokenizer, data_args

) -> Dict:

"""Make dataset and collator for supervised fine-tuning."""

train_dataset = SupervisedDataset(

data_path=data_args.data_path, data_subset=data_args.data_subset,

tokenizer=tokenizer,

data_collator = DataCollatorForSupervisedDataset(tokenizer=tokenizer)

return dict(

train_dataset=train_dataset, eval_dataset=None, data_collator=data_collator

def find_all_linear_names(model, bits=4):

cls = (

bnb.nn.Linear4bit

if bits == 4

else (bnb.nn.Linear8bitLt if bits == 8 else torch.nn.Linear)

lora_module_names = set()

for name, module in model.named_modules():

if isinstance(module, cls):

names = name.split(".")

lora_module_names.add(names[0] if len(names) == 1 else names[-1])

 if "lm_head" in lora_module_names: # needed for 16-bit

lora_module_names.remove("lm_head")

return list(lora_module_names)

def print_trainable_parameters(model):

"""

Prints the number of trainable parameters in the model.

"""

trainable_params = 0

all_param = 0

for _, param in model.named_parameters():

all_param += param.numel()

if param.requires_grad:

trainable_params += param.numel()

print(

f"trainable params: {trainable_params} || all params: {all_param} || trainable%: {100 *

trainable_params / all_param}"

class PrinterCallback(transformers.TrainerCallback):

def on_step_end(self, args, state, control, **kwargs):

GPUtil.showUtilization(all=True, attrList=None, useOldCode=True)

def train():

parser = transformers.HfArgumentParser(

(ModelArguments, DataArguments, TrainingArguments)

model_args, data_args, training_args = parser.parse_args_into_dataclasses()

training_args.ddp_find_unused_parameters = False

set_seed(42)

model_config = OpenRAGConfig.from_pretrained(model_args.model_name_or_path)

model_config.pretraining_tp = 1 ## without tensor parallelism rank

# OpenRAG Config

model_config.moe_dtype = "bfloat16"

model_config.lora_r = model_args.lora_r

model_config.lora_alpha = model_args.lora_alpha

model_config.adapter_dim = model_args.adapter_dim

model_config.topk = model_args.topk

model_config.moe_scaling = model_args.moe_scaling

model_config.num_experts = model_args.num_experts

model_config.output_router_logits = False

# # Seq Length Extension

# model_config.rope_scaling = {

# "type": "dynamic",

# "factor": 2,

#}

model = LlamaForCausalLM.from_pretrained(

model_args.model_name_or_path,

config=model_config,

cache_dir=training_args.cache_dir,

load_in_4bit=True,

quantization_config=BitsAndBytesConfig(

load_in_4bit=True,

bnb_4bit_compute_dtype=torch.bfloat16,

bnb_4bit_use_double_quant=True,
 bnb_4bit_quant_type="nf4",

),

output_loading_info=False,

tokenizer = transformers.AutoTokenizer.from_pretrained(

model_args.model_name_or_path,

cache_dir=training_args.cache_dir,

model_max_length=training_args.model_max_length,

padding_side="right",

use_fast=False,

trust_remote_code=True,

special_token_dict = {

"additional_special_tokens": [

"[No Retrieval]",

"[Retrieval]",

"[Continue to Use Evidence]",

"[Irrelevant]",

"[Relevant]",

"<paragraph>",

"</paragraph>",

"[Utility:1]",

"[Utility:2]",

"[Utility:3]",

"[Utility:4]",

"[Utility:5]",

"[Fully supported]",

"[Partially supported]",

"[No support / Contradictory]",

 ]

special_token_dict["bos_token"] = "<s>"

special_token_dict["eos_token"] = "</s>"

special_token_dict["unk_token"] = "<unk>"

special_token_dict["pad_token"] = "<pad>"

num_added_tokens = tokenizer.add_special_tokens(special_token_dict)

global context_markups

context_markups = []

for token in ["<paragraph>", "</paragraph>"]:

context_markups.append(tokenizer.convert_tokens_to_ids(token))

assert (

num_added_tokens > 10

), "special tokens must be added to the original tokenizers."

embedding_size = model.get_input_embeddings().weight.shape[0]

if len(tokenizer) > embedding_size:

model.resize_token_embeddings(len(tokenizer))

model = prepare_model_for_kbit_training(model, use_gradient_checkpointing=True)

model.gradient_checkpointing_enable()

lora_modules = []

lora_modules += [

"embed_tokens",
 "lm_head",

lora_modules += [

"q_proj",

"k_proj",

"v_proj",

"o_proj",

"up_proj",

"gate_proj",

"down_proj",

config = LoraConfig(

r=model_config.lora_r,

lora_alpha=model_config.lora_alpha,

target_modules=lora_modules,

lora_dropout=0.1,

bias="none",

task_type="CAUSAL_LM",

model = get_peft_model(model, config)

# Zero Init

for n, p in model.named_parameters():

if "adapter_up" in n:

nn.init.zeros_(p)

if "adapter_down" in n:

nn.init.kaiming_uniform_(p, a=math.sqrt(5))

if "router" in n:
 nn.init.kaiming_uniform_(p, a=math.sqrt(5))

for name, module in model.named_modules():

if isinstance(module, LoraLayer):

if training_args.bf16:

module = module.to(torch.bfloat16)

if "norm" in name:

module = module.to(torch.float32)

if "lm_head" in name or "embed_tokens" in name:

if hasattr(module, "weight"):

if training_args.bf16 and module.weight.dtype == torch.float32:

module = module.to(torch.bfloat16)

if "adapter" in name:

if training_args.bf16:

module = module.to(torch.bfloat16)

else:

module = module.to(torch.float32)

for n, p in model.named_parameters():

if "adapter" in n:

p.requires_grad = True

model.config.use_cache = False

print_trainable_parameters(model)

data_module = make_supervised_data_module(tokenizer=tokenizer, data_args=data_args)

trainer = Trainer(

model=model, tokenizer=tokenizer, args=training_args, **data_module

trainer.add_callback(SavePeftModelCallback)

trainer.add_callback(PrinterCallback)
 trainer.train()

model.save_pretrained(training_args.output_dir)

if __name__ == "__main__":

train()