# Adopted from https://github.com/lm-sys/FastChat. Below is the original copyright:
# Adopted from tatsu-lab@stanford_alpaca. Below is the original copyright:
# 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
import copy
import deepspeed
from dataclasses import dataclass, field
import json
import logging
import pathlib
from typing import Dict, Optional, Sequence, List
import ast
import yaml
import time
import random
import yaml
import math
import re
import torch
import transformers
import tokenizers
from llava.constants import IGNORE_INDEX, DEFAULT_IMAGE_TOKEN, DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN, IMAGE_TOKEN_INDEX
from torch.utils.data import Dataset
from llava.train.llava_trainer import LLaVADPOTrainer
from data_processing.utils import load_jsonl, load_json
from llava import conversation as conversation_lib
from llava.model import *
from llava.model.language_model.llava_qwen import LlavaQwenConfig
from llava.model.language_model.llava_llama import LlavaConfig
from llava.model.language_model.llava_mistral import LlavaMistralConfig
from llava.mm_utils import process_highres_image, process_anyres_image, process_highres_image_crop_split, tokenizer_image_token
from llava.utils import rank0_print
from transformers import AutoConfig
import pickle
from trl.trainer.utils import DPODataCollatorWithPadding
from PIL import Image, ImageFile
from decord import VideoReader, cpu
ImageFile.LOAD_TRUNCATED_IMAGES = True
from packaging import version
from typing import Any
local_rank = None
import numpy as np
IS_TOKENIZER_GREATER_THAN_0_14 = version.parse(tokenizers.__version__) >= version.parse("0.14")
@dataclass
class ModelArguments:
model_name_or_path: Optional[str] = field(default="facebook/opt-125m")
model_class_name: Optional[str] = field(default=None, metadata={"help": "Used to init model class, format is XXXXForCausalLM. e.g. currently XXXX is chosen from LlavaLlama, LlavaMixtral, LlavaMistral, Llama"})
mm_tunable_parts: Optional[str] = field(
default=None, metadata={"help": 'Could be "mm_mlp_adapter", "mm_vision_resampler", "mm_vision_tower,mm_mlp_adapter,mm_language_model", "mm_vision_tower,mm_mlp_adapter,mm_language_model", "mm_mlp_adapter,mm_language_model"'}
)
# deciding which part of the multimodal model to tune, will overwrite other previous settings
version: Optional[str] = field(default="v0")
freeze_backbone: bool = field(default=False)
tune_mm_mlp_adapter: bool = field(default=False)
tune_mm_vision_resampler: bool = field(default=False)
vision_tower: Optional[str] = field(default=None)
vision_tower_pretrained: Optional[str] = field(default=None) # default to the last layer
unfreeze_mm_vision_tower: bool = field(default=False)
unfreeze_language_model: bool = field(default=False)
mm_vision_select_layer: Optional[int] = field(default=-1) # default to the last layer
pretrain_mm_mlp_adapter: Optional[str] = field(default=None)
mm_projector_type: Optional[str] = field(default="linear")
mm_use_im_start_end: bool = field(default=False)
mm_use_im_patch_token: bool = field(default=True)
mm_patch_merge_type: Optional[str] = field(default="flat")
mm_vision_select_feature: Optional[str] = field(default="patch")
mm_resampler_type: Optional[str] = field(default=None)
mm_mask_drop_mode: str = field(default="fixed")
mm_mask_drop_skip_percentage: float = field(default=0.0)
mm_mask_drop_ratio: float = field(default=0.25)
mm_mask_drop_ratio_upper: Optional[float] = field(default=None)
mm_mask_drop_ratio_lower: Optional[float] = field(default=None)
mm_spatial_pool_stride: Optional[int] = field(default=None)
mm_spatial_pool_mode: str = field(default="average")
mm_spatial_pool_out_channels: Optional[int] = field(default=None)
mm_perceiver_depth: Optional[int] = field(default=3)
mm_perceiver_latents: Optional[int] = field(default=32)
mm_perceiver_ff_mult: Optional[float] = field(default=4)
mm_perceiver_pretrained: Optional[str] = field(default=None)
mm_qformer_depth: Optional[int] = field(default=3)
mm_qformer_latents: Optional[int] = field(default=32)
mm_qformer_pretrained: Optional[str] = field(default=None)
rope_scaling_factor: Optional[float] = field(default=None)
rope_scaling_type: Optional[str] = field(default=None)
s2: Optional[bool] = field(default=False)
s2_scales: Optional[str] = field(default="336,672,1008")
@dataclass
class DataArguments:
data_path: str = field(default=None, metadata={"help": "Path to the training data, in llava's instruction.json format. Supporting multiple json files via /path/to/{a,b,c}.json"})
lazy_preprocess: bool = False
is_multimodal: bool = False
image_folder: Optional[str] = field(default=None)
video_folder: Optional[str] = field(default=None)
video_fps: Optional[int] = field(default=1)
image_aspect_ratio: str = "square"
image_grid_pinpoints: Optional[str] = field(default=None)
image_crop_resolution: int = 384
image_split_resolution: int = 384
input_prompt: Optional[str] = field(default=None)
refine_prompt: Optional[bool] = field(default=False)
frames_upbound: Optional[int] = field(default=0)
num_sample: Optional[int] = field(default=None)
@dataclass
class TrainingArguments(transformers.TrainingArguments):
cache_dir: Optional[str] = field(default=None)
optim: str = field(default="adamw_torch")
remove_unused_columns: bool = field(default=False)
freeze_mm_mlp_adapter: bool = field(default=False)
freeze_mm_vision_resampler: bool = field(default=False)
mpt_attn_impl: Optional[str] = field(default="triton")
model_max_length: int = field(
default=4096,
metadata={"help": "Maximum sequence length. Sequences will be right padded (and possibly truncated)."},
)
double_quant: bool = field(default=True, metadata={"help": "Compress the quantization statistics through double quantization."})
quant_type: str = field(default="nf4", metadata={"help": "Quantization data type to use. Should be one of `fp4` or `nf4`."})
bits: int = field(default=16, metadata={"help": "How many bits to use."})
lora_enable: bool = False
lora_r: int = 64
lora_alpha: int = 16
lora_dropout: float = 0.05
lora_weight_path: str = ""
lora_bias: str = "none"
mm_projector_lr: Optional[float] = None
mm_vision_tower_lr: Optional[float] = None
group_by_varlen: bool = field(default=False)
group_by_modality_length: bool = field(default=False)
group_by_modality_length_auto: bool = field(default=False)
auto_find_batch_size: bool = field(default=False)
gradient_checkpointing: bool = field(default=True)
verbose_logging: bool = field(default=False)
attn_implementation: str = field(default="flash_attention_2", metadata={"help": "Use transformers attention implementation."})
dpo_alpha: float = field(default=1.0)
beta: float = field(default=0.1)
gamma: float = field(default=1.0)
generate_during_eval: bool = field(default=False)
precompute_ref_log_probs: bool = field(default=False)
def maybe_zero_3(param, ignore_status=False, name=None):
from deepspeed import zero
from deepspeed.runtime.zero.partition_parameters import ZeroParamStatus
if hasattr(param, "ds_id"):
if param.ds_status == ZeroParamStatus.NOT_AVAILABLE:
if not ignore_status:
logging.warning(f"{name}: param.ds_status != ZeroParamStatus.NOT_AVAILABLE: {param.ds_status}")
with zero.GatheredParameters([param]):
param = param.data.detach().cpu().clone()
else:
param = param.detach().cpu().clone()
return param
# Borrowed from peft.utils.get_peft_model_state_dict
def get_peft_state_maybe_zero_3(named_params, bias):
if bias == "none":
to_return = {k: t for k, t in named_params if "lora_" in k}
elif bias == "all":
to_return = {k: t for k, t in named_params if "lora_" in k or "bias" in k}
elif bias == "lora_only":
to_return = {}
maybe_lora_bias = {}
lora_bias_names = set()
for k, t in named_params:
if "lora_" in k:
to_return[k] = t
bias_name = k.split("lora_")[0] + "bias"
lora_bias_names.add(bias_name)
elif "bias" in k:
maybe_lora_bias[k] = t
for k, t in maybe_lora_bias:
if bias_name in lora_bias_names:
to_return[bias_name] = t
else:
raise NotImplementedError
to_return = {k: maybe_zero_3(v, ignore_status=True) for k, v in to_return.items()}
return to_return
def get_peft_state_non_lora_maybe_zero_3(named_params, require_grad_only=True):
to_return = {k: t for k, t in named_params if "lora_" not in k}
if require_grad_only:
to_return = {k: t for k, t in to_return.items() if t.requires_grad}
to_return = {k: maybe_zero_3(v, ignore_status=True).cpu() for k, v in to_return.items()}
return to_return
def get_mm_adapter_state_maybe_zero_3(named_params, keys_to_match):
to_return = {k: t for k, t in named_params if any(key_match in k for key_match in keys_to_match)}
to_return = {k: maybe_zero_3(v, ignore_status=True).cpu() for k, v in to_return.items()}
return to_return
def find_all_linear_names(model):
cls = torch.nn.Linear
lora_module_names = set()
multimodal_keywords = ["mm_projector", "vision_tower", "vision_resampler"]
for name, module in model.named_modules():
if any(mm_keyword in name for mm_keyword in multimodal_keywords):
continue
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 safe_save_model_for_hf_trainer(trainer: transformers.Trainer, output_dir: str):
"""Collects the state dict and dump to disk."""
if hasattr(trainer.args, "tune_mm_mlp_adapter") and trainer.args.tune_mm_mlp_adapter:
check_only_save_mm_adapter_tunnable = True
# only has mm_mlp_adapter and mm_vision_resampler in the tuneable parts
elif hasattr(trainer.args, "mm_tunable_parts") and (len(trainer.args.mm_tunable_parts.split(",")) == 1 and ("mm_mlp_adapter" in trainer.args.mm_tunable_parts or "mm_vision_resampler" in trainer.args.mm_tunable_parts)):
check_only_save_mm_adapter_tunnable = True
else:
check_only_save_mm_adapter_tunnable = False
trainer.accelerator.wait_for_everyone()
torch.cuda.synchronize()
rank0_print(f"Only save projectors: {check_only_save_mm_adapter_tunnable}")
if check_only_save_mm_adapter_tunnable:
# Only save Adapter
keys_to_match = ["mm_projector", "vision_resampler"]
if getattr(trainer.args, "use_im_start_end", False):
keys_to_match.extend(["embed_tokens", "embed_in"])
weight_to_save = get_mm_adapter_state_maybe_zero_3(trainer.model.named_parameters(), keys_to_match)
trainer.model.config.save_pretrained(output_dir)
current_folder = output_dir.split("/")[-1]
parent_folder = os.path.dirname(output_dir)
if trainer.args.local_rank == 0 or trainer.args.local_rank == -1:
if current_folder.startswith("checkpoint-"):
mm_projector_folder = os.path.join(parent_folder, "mm_projector")
os.makedirs(mm_projector_folder, exist_ok=True)
torch.save(weight_to_save, os.path.join(mm_projector_folder, f"{current_folder}.bin"))
else:
torch.save(weight_to_save, os.path.join(output_dir, f"mm_projector.bin"))
return
if trainer.deepspeed:
trainer.save_model(output_dir)
return
state_dict = trainer.model.state_dict()
if trainer.args.should_save:
cpu_state_dict = {key: value.cpu() for key, value in state_dict.items()}
del state_dict
trainer._save(output_dir, state_dict=cpu_state_dict) # noqa
def smart_tokenizer_and_embedding_resize(
special_tokens_dict: Dict,
tokenizer: transformers.PreTrainedTokenizer,
model: transformers.PreTrainedModel,
):
"""Resize tokenizer and embedding.
Note: This is the unoptimized version that may make your embedding size not be divisible by 64.
"""
num_new_tokens = tokenizer.add_special_tokens(special_tokens_dict)
model.resize_token_embeddings(len(tokenizer))
if num_new_tokens > 0:
input_embeddings = model.get_input_embeddings().weight.data
output_embeddings = model.get_output_embeddings().weight.data
input_embeddings_avg = input_embeddings[:-num_new_tokens].mean(dim=0, keepdim=True)
output_embeddings_avg = output_embeddings[:-num_new_tokens].mean(dim=0, keepdim=True)
input_embeddings[-num_new_tokens:] = input_embeddings_avg
output_embeddings[-num_new_tokens:] = output_embeddings_avg
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 _mask_targets(target, tokenized_lens, speakers):
# cur_idx = 0
cur_idx = tokenized_lens[0]
tokenized_lens = tokenized_lens[1:]
target[:cur_idx] = IGNORE_INDEX
for tokenized_len, speaker in zip(tokenized_lens, speakers):
if speaker == "human":
target[cur_idx + 2 : cur_idx + tokenized_len] = IGNORE_INDEX
cur_idx += tokenized_len
def _add_speaker_and_signal(header, source, get_conversation=True):
"""Add speaker and start/end signal on each round."""
BEGIN_SIGNAL = "### "
END_SIGNAL = "\n"
conversation = header
for sentence in source:
from_str = sentence["from"]
if from_str.lower() == "human":
from_str = conversation_lib.default_conversation.roles[0]
elif from_str.lower() == "gpt":
from_str = conversation_lib.default_conversation.roles[1]
else:
from_str = "unknown"
sentence["value"] = BEGIN_SIGNAL + from_str + ": " + sentence["value"] + END_SIGNAL
if get_conversation:
conversation += sentence["value"]
conversation += BEGIN_SIGNAL
return conversation
def preprocess_multimodal(sources: Sequence[str], data_args: DataArguments) -> Dict:
is_multimodal = data_args.is_multimodal
if not is_multimodal:
return sources
for source in sources:
for sentence in source:
if DEFAULT_IMAGE_TOKEN in sentence["value"] and not sentence["value"].startswith(DEFAULT_IMAGE_TOKEN):
sentence["value"] = sentence["value"].replace(DEFAULT_IMAGE_TOKEN, "").strip()
sentence["value"] = DEFAULT_IMAGE_TOKEN + "\n" + sentence["value"]
sentence["value"] = sentence["value"].strip()
if "mmtag" in conversation_lib.default_conversation.version:
sentence["value"] = sentence["value"].replace(DEFAULT_IMAGE_TOKEN, "" + DEFAULT_IMAGE_TOKEN + "")
replace_token = DEFAULT_IMAGE_TOKEN
if data_args.mm_use_im_start_end:
replace_token = DEFAULT_IM_START_TOKEN + replace_token + DEFAULT_IM_END_TOKEN
sentence["value"] = sentence["value"].replace(DEFAULT_IMAGE_TOKEN, replace_token)
return sources
def preprocess_multimodal_movie(sources: Sequence[str], data_args: DataArguments, video_inputs: str) -> Dict:
is_multimodal = data_args.is_multimodal
if not is_multimodal:
return sources
for source in sources:
for sentence in source:
if DEFAULT_IMAGE_TOKEN in sentence["value"]:
prompt = sentence["value"].replace(DEFAULT_IMAGE_TOKEN, "").strip()
replace_token = video_inputs
if data_args.mm_use_im_start_end:
replace_token = DEFAULT_IM_START_TOKEN + replace_token + DEFAULT_IM_END_TOKEN
sentence["value"] = sentence["value"].replace(DEFAULT_IMAGE_TOKEN, replace_token)
return sources, prompt
def preprocess_llama_2(sources, tokenizer: transformers.PreTrainedTokenizer, has_image: bool = False) -> Dict:
conv = conversation_lib.default_conversation.copy()
roles = {"human": conv.roles[0], "gpt": conv.roles[1]}
# Apply prompt templates
conversations = []
for i, source in enumerate(sources):
if roles[source[0]["from"]] != conv.roles[0]:
# Skip the first one if it is not from human
source = source[1:]
conv.messages = []
for j, sentence in enumerate(source):
role = roles[sentence["from"]]
assert role == conv.roles[j % 2], f"{i}"
conv.append_message(role, sentence["value"])
conversations.append(conv.get_prompt())
# Tokenize conversations
if has_image:
input_ids = torch.stack([tokenizer_image_token(prompt, tokenizer, return_tensors="pt") for prompt in conversations], dim=0)
else:
input_ids = tokenizer(
conversations,
return_tensors="pt",
padding="longest",
max_length=tokenizer.model_max_length,
truncation=True,
).input_ids
targets = input_ids.clone()
assert conv.sep_style == conversation_lib.SeparatorStyle.LLAMA_2
# Mask targets
sep = "[/INST] "
for conversation, target in zip(conversations, targets):
total_len = int(target.ne(tokenizer.pad_token_id).sum())
rounds = conversation.split(conv.sep2)
cur_len = 1
target[:cur_len] = IGNORE_INDEX
for i, rou in enumerate(rounds):
if rou == "":
break
parts = rou.split(sep)
if len(parts) != 2:
break
parts[0] += sep
if has_image:
round_len = len(tokenizer_image_token(rou, tokenizer))
instruction_len = len(tokenizer_image_token(parts[0], tokenizer)) - 2
else:
round_len = len(tokenizer(rou).input_ids)
instruction_len = len(tokenizer(parts[0]).input_ids) - 2
target[cur_len : cur_len + instruction_len] = IGNORE_INDEX
cur_len += round_len
target[cur_len:] = IGNORE_INDEX
if cur_len < tokenizer.model_max_length:
if cur_len != total_len:
target[:] = IGNORE_INDEX
rank0_print(f"WARNING: tokenization mismatch: {cur_len} vs. {total_len}." f" (ignored)")
return dict(
input_ids=input_ids,
labels=targets,
)
def make_conv(prompt, answer):
return [
{
"from": "human",
"value": prompt,
},
{
"from": "gpt",
"value": answer,
},
]
def preprocess_gemma(sources: List[List[Dict[str, str]]], tokenizer: transformers.PreTrainedTokenizer, has_image: bool = False) -> Dict:
conv: conversation_lib.Conversation = conversation_lib.default_conversation.copy()
roles: Dict[str, str] = {"human": conv.roles[0], "gpt": conv.roles[1]}
# Apply prompt templates
conversations: List[str] = []
for i, source in enumerate(sources):
if roles[source[0]["from"]] != conv.roles[0]:
# Skip the first one if it is not from human
source: List[Dict[str, str]] = source[1:]
conv.messages = []
for j, sentence in enumerate(source):
role: str = roles[sentence["from"]]
assert role == conv.roles[j % 2], f"{i}"
conv.append_message(role, sentence["value"])
conversations.append(conv.get_prompt())
# Tokenize conversations
if has_image:
input_ids: torch.Tensor = torch.stack([tokenizer_image_token(prompt, tokenizer, return_tensors="pt") for prompt in conversations], dim=0)
else:
input_ids: torch.Tensor = tokenizer(
conversations,
return_tensors="pt",
padding="longest",
max_length=tokenizer.model_max_length,
truncation=True,
).input_ids
targets: torch.Tensor = input_ids.clone()
assert conv.sep_style == conversation_lib.SeparatorStyle.GEMMA
# Mask target
sep: str = conv.sep + conv.roles[1]
for conversation, target in zip(conversations, targets):
total_len: int = int(target.ne(tokenizer.pad_token_id).sum())
rounds: List[str] = conversation.split(conv.sep)
re_rounds = []
for conv_idx in range(0, len(rounds), 2):
re_rounds.append(conv.sep.join(rounds[conv_idx : conv_idx + 2]))
cur_len = 1 # Ignore
target[:cur_len] = IGNORE_INDEX
for i, rou in enumerate(re_rounds):
if rou == "":
break
parts = rou.split(sep)
if len(parts) != 2:
break
parts[0] += sep # Re-append sep because split on this
# Now "".join(parts)==rou
if has_image:
round_len = len(tokenizer_image_token(rou, tokenizer)) - 1 # Ignore
instruction_len = len(tokenizer_image_token(parts[0], tokenizer)) - 1 # Ignore
else:
round_len = len(tokenizer(rou).input_ids) - 1 # Ignore
instruction_len = len(tokenizer(parts[0]).input_ids) - 1 # Ignore
round_len += 2 # sep: \n takes 2 tokens
target[cur_len : cur_len + instruction_len] = IGNORE_INDEX
cur_len += round_len
target[cur_len:] = IGNORE_INDEX
if cur_len < tokenizer.model_max_length:
if cur_len != total_len:
target[:] = IGNORE_INDEX
rank0_print(f"WARNING: tokenization mismatch: {cur_len} vs. {total_len}." f" (ignored)")
return dict(
input_ids=input_ids,
labels=targets,
)
def preprocess_qwen(sources, tokenizer: transformers.PreTrainedTokenizer, has_image: bool = False, max_len=2048, system_message: str = "You are a helpful assistant.") -> Dict:
roles = {"human": "<|im_start|>user", "gpt": "<|im_start|>assistant"}
im_start, im_end = tokenizer.additional_special_tokens_ids
nl_tokens = tokenizer("\n").input_ids
_system = tokenizer("system").input_ids + nl_tokens
_user = tokenizer("user").input_ids + nl_tokens
_assistant = tokenizer("assistant").input_ids + nl_tokens
# Apply prompt templates
input_ids, targets = [], []
for i, source in enumerate(sources):
if roles[source[0]["from"]] != roles["human"]:
source = source[1:]
input_id, target = [], []
system = [im_start] + _system + tokenizer(system_message).input_ids + [im_end] + nl_tokens
input_id += system
target += [im_start] + [IGNORE_INDEX] * (len(system) - 3) + [im_end] + nl_tokens
assert len(input_id) == len(target)
for j, sentence in enumerate(source):
role = roles[sentence["from"]]
if has_image and "" in sentence["value"]:
assert sentence["value"].startswith(""), print(sentence["value"])
_input_id = tokenizer(role).input_ids + nl_tokens + [IMAGE_TOKEN_INDEX] + nl_tokens + tokenizer(sentence["value"][len("") :]).input_ids + [im_end] + nl_tokens
else:
_input_id = tokenizer(role).input_ids + nl_tokens + tokenizer(sentence["value"]).input_ids + [im_end] + nl_tokens
input_id += _input_id
if role == "<|im_start|>user":
_target = [im_start] + [IGNORE_INDEX] * (len(_input_id) - 3) + [im_end] + nl_tokens
elif role == "<|im_start|>assistant":
_target = [im_start] + [IGNORE_INDEX] * len(tokenizer(role).input_ids) + _input_id[len(tokenizer(role).input_ids) + 1 : -2] + [im_end] + nl_tokens
else:
raise NotImplementedError
target += _target
assert len(input_id) == len(target)
# input_id += [tokenizer.pad_token_id] * (max_len - len(input_id))
# target += [IGNORE_INDEX] * (max_len - len(target))
input_ids.append(input_id)
targets.append(target)
input_ids = torch.tensor(input_ids, dtype=torch.long)
targets = torch.tensor(targets, dtype=torch.long)
return dict(
input_ids=input_ids, # tensor(bs x seq_len)
labels=targets, # tensor(bs x seq_len)
# attention_mask=input_ids.ne(tokenizer.pad_token_id), # tensor(bs x seq_len)
)
def preprocess_llama3(
sources,
tokenizer: transformers.PreTrainedTokenizer,
has_image: bool = False,
max_len=2048,
system_message: str = "You are a helpful language and vision assistant. You are able to understand the visual content that the user provides, and assist the user with a variety of tasks using natural language.",
) -> Dict:
roles = {"human": "<|start_header_id|>user<|end_header_id|>", "gpt": "<|start_header_id|>assistant<|end_header_id|>"}
eot_id = tokenizer.convert_tokens_to_ids("<|eot_id|>")
nl_tokens = tokenizer("\n").input_ids
# Apply prompt templates
input_ids, targets = [], []
for i, source in enumerate(sources):
if roles[source[0]["from"]] != roles["human"]:
source = source[1:]
input_id, target = [], []
system = tokenizer("<|begin_of_text|>").input_ids + tokenizer("<|start_header_id|>system<|end_header_id|>").input_ids + nl_tokens * 2 + tokenizer(system_message).input_ids + [eot_id]
input_id += system
target += [IGNORE_INDEX] * len(system)
for j, sentence in enumerate(source):
role = roles[sentence["from"]]
if has_image and "" in sentence["value"]:
assert sentence["value"].startswith(""), print(sentence["value"])
_input_id = tokenizer(role).input_ids + nl_tokens * 2 + [IMAGE_TOKEN_INDEX] + tokenizer(sentence["value"][len("") :]).input_ids + [eot_id]
else:
_input_id = tokenizer(role).input_ids + nl_tokens * 2 + tokenizer(sentence["value"]).input_ids + [eot_id]
input_id += _input_id
if role == "<|start_header_id|>user<|end_header_id|>":
_target = [IGNORE_INDEX] * len(_input_id)
elif role == "<|start_header_id|>assistant<|end_header_id|>":
_target = [IGNORE_INDEX] * (len(tokenizer(role).input_ids) + 2) + _input_id[len(tokenizer(role).input_ids) + 2 : -1] + [eot_id]
else:
raise NotImplementedError
target += _target
assert len(input_id) == len(target), f"{len(input_id)} != {len(target)}"
input_ids.append(input_id)
targets.append(target)
input_ids = torch.tensor(input_ids, dtype=torch.long)
targets = torch.tensor(targets, dtype=torch.long)
return dict(
input_ids=input_ids, # tensor(bs x seq_len)
labels=targets, # tensor(bs x seq_len)
)
def preprocess_v1(sources, tokenizer: transformers.PreTrainedTokenizer, has_image: bool = False) -> Dict:
conv = conversation_lib.default_conversation.copy()
roles = {"human": conv.roles[0], "gpt": conv.roles[1]}
# Apply prompt templates
conversations = []
for i, source in enumerate(sources):
if roles[source[0]["from"]] != conv.roles[0]:
# Skip the first one if it is not from human
source = source[1:]
conv.messages = []
for j, sentence in enumerate(source):
role = roles[sentence["from"]]
assert role == conv.roles[j % 2], f"{i}"
conv.append_message(role, sentence["value"])
conversations.append(conv.get_prompt())
# Tokenize conversations
if has_image:
input_ids = torch.stack([tokenizer_image_token(prompt, tokenizer, return_tensors="pt") for prompt in conversations], dim=0)
else:
input_ids = tokenizer(
conversations,
return_tensors="pt",
padding="longest",
max_length=tokenizer.model_max_length,
truncation=True,
).input_ids
targets = input_ids.clone()
assert conv.sep_style == conversation_lib.SeparatorStyle.TWO
# Mask targets
sep = conv.sep + conv.roles[1] + ": "
for conversation, target in zip(conversations, targets):
total_len = int(target.ne(tokenizer.pad_token_id).sum())
rounds = conversation.split(conv.sep2)
cur_len = 1
target[:cur_len] = IGNORE_INDEX
for i, rou in enumerate(rounds):
if rou == "":
break
parts = rou.split(sep)
if len(parts) != 2:
break
parts[0] += sep
if has_image:
round_len = len(tokenizer_image_token(rou, tokenizer))
instruction_len = len(tokenizer_image_token(parts[0], tokenizer)) - 2
else:
round_len = len(tokenizer(rou).input_ids)
instruction_len = len(tokenizer(parts[0]).input_ids) - 2
if i != 0 and not tokenizer.legacy and IS_TOKENIZER_GREATER_THAN_0_14:
round_len -= 1
instruction_len -= 1
target[cur_len : cur_len + instruction_len] = IGNORE_INDEX
cur_len += round_len
target[cur_len:] = IGNORE_INDEX
if cur_len < tokenizer.model_max_length:
if cur_len != total_len:
target[:] = IGNORE_INDEX
print(f"WARNING: tokenization mismatch: {cur_len} vs. {total_len}." f" (ignored)")
return dict(
input_ids=input_ids,
labels=targets,
)
def preprocess_mpt(sources, tokenizer: transformers.PreTrainedTokenizer, has_image: bool = False) -> Dict:
conv = conversation_lib.default_conversation.copy()
roles = {"human": conv.roles[0], "gpt": conv.roles[1]}
# Apply prompt templates
conversations = []
for i, source in enumerate(sources):
if roles[source[0]["from"]] != conv.roles[0]:
# Skip the first one if it is not from human
source = source[1:]
conv.messages = []
for j, sentence in enumerate(source):
role = roles[sentence["from"]]
assert role == conv.roles[j % 2], f"{i}"
conv.append_message(role, sentence["value"])
conversations.append(conv.get_prompt())
# Tokenize conversations
if has_image:
input_ids = torch.stack([tokenizer_image_token(prompt, tokenizer, return_tensors="pt") for prompt in conversations], dim=0)
else:
input_ids = tokenizer(
conversations,
return_tensors="pt",
padding="longest",
max_length=tokenizer.model_max_length,
truncation=True,
).input_ids
targets = input_ids.clone()
assert conv.sep_style == conversation_lib.SeparatorStyle.MPT
# Mask targets
sep = conv.sep + conv.roles[1]
for conversation, target in zip(conversations, targets):
total_len = int(target.ne(tokenizer.pad_token_id).sum())
rounds = conversation.split(conv.sep)
re_rounds = [conv.sep.join(rounds[:3])] # system + user + gpt
for conv_idx in range(3, len(rounds), 2):
re_rounds.append(conv.sep.join(rounds[conv_idx : conv_idx + 2])) # user + gpt
cur_len = 1
target[:cur_len] = IGNORE_INDEX
for i, rou in enumerate(re_rounds):
if rou == "":
break
parts = rou.split(sep)
if len(parts) != 2:
break
parts[0] += sep
if has_image:
round_len = len(tokenizer_image_token(rou, tokenizer))
instruction_len = len(tokenizer_image_token(parts[0], tokenizer)) - 1
else:
round_len = len(tokenizer(rou).input_ids)
instruction_len = len(tokenizer(parts[0]).input_ids) - 1
if i != 0 and getattr(tokenizer, "legacy", False) and IS_TOKENIZER_GREATER_THAN_0_14:
round_len += 1
instruction_len += 1
target[cur_len : cur_len + instruction_len] = IGNORE_INDEX
cur_len += round_len
target[cur_len:] = IGNORE_INDEX
if cur_len < tokenizer.model_max_length:
if cur_len != total_len:
target[:] = IGNORE_INDEX
print(f"WARNING: tokenization mismatch: {cur_len} vs. {total_len}." f"(#turns={len(re_rounds)} ignored)")
return dict(
input_ids=input_ids,
labels=targets,
)
def preprocess_plain(
sources: Sequence[str],
tokenizer: transformers.PreTrainedTokenizer,
) -> Dict:
# add end signal and concatenate together
conversations = []
for source in sources:
assert len(source) == 2
assert DEFAULT_IMAGE_TOKEN in source[0]["value"]
source[0]["value"] = DEFAULT_IMAGE_TOKEN
conversation = source[0]["value"] + source[1]["value"] + conversation_lib.default_conversation.sep
conversations.append(conversation)
# tokenize conversations
input_ids = [tokenizer_image_token(prompt, tokenizer, return_tensors="pt") for prompt in conversations]
targets = copy.deepcopy(input_ids)
for target, source in zip(targets, sources):
tokenized_len = len(tokenizer_image_token(source[0]["value"], tokenizer))
target[:tokenized_len] = IGNORE_INDEX
return dict(input_ids=input_ids, labels=targets)
def preprocess(sources: Sequence[str], tokenizer: transformers.PreTrainedTokenizer, has_image: bool = False) -> Dict:
"""
Given a list of sources, each is a conversation list. This transform:
1. Add signal '### ' at the beginning each sentence, with end signal '\n';
2. Concatenate conversations together;
3. Tokenize the concatenated conversation;
4. Make a deepcopy as the target. Mask human words with IGNORE_INDEX.
"""
if conversation_lib.default_conversation.sep_style == conversation_lib.SeparatorStyle.PLAIN:
return preprocess_plain(sources, tokenizer)
if conversation_lib.default_conversation.sep_style == conversation_lib.SeparatorStyle.LLAMA_2:
return preprocess_llama_2(sources, tokenizer, has_image=has_image)
if conversation_lib.default_conversation.version.startswith("v1"):
return preprocess_v1(sources, tokenizer, has_image=has_image)
if conversation_lib.default_conversation.version == "mpt":
return preprocess_mpt(sources, tokenizer, has_image=has_image)
if conversation_lib.default_conversation.version == "qwen":
return preprocess_qwen(sources, tokenizer, has_image=has_image)
if conversation_lib.default_conversation.version == "gemma":
return preprocess_gemma(sources, tokenizer, has_image=has_image)
if conversation_lib.default_conversation.version == "llama_v3":
return preprocess_llama3(sources, tokenizer, has_image=has_image)
# add end signal and concatenate together
conversations = []
for source in sources:
header = f"{conversation_lib.default_conversation.system}\n\n"
conversation = _add_speaker_and_signal(header, source)
conversations.append(conversation)
# tokenize conversations
def get_tokenize_len(prompts):
return [len(tokenizer_image_token(prompt, tokenizer)) for prompt in prompts]
if has_image:
input_ids = [tokenizer_image_token(prompt, tokenizer, return_tensors="pt") for prompt in conversations]
else:
conversations_tokenized = _tokenize_fn(conversations, tokenizer)
input_ids = conversations_tokenized["input_ids"]
targets = copy.deepcopy(input_ids)
for target, source in zip(targets, sources):
if has_image:
tokenized_lens = get_tokenize_len([header] + [s["value"] for s in source])
else:
tokenized_lens = _tokenize_fn([header] + [s["value"] for s in source], tokenizer)["input_ids_lens"]
speakers = [sentence["from"] for sentence in source]
_mask_targets(target, tokenized_lens, speakers)
return dict(input_ids=input_ids, labels=targets)
def load_data(data_path):
if "jsonl" in data_path:
data_list = load_jsonl(data_path)
else:
data_list = load_json(data_path)
return data_list
class DPODataset(Dataset):
"""Dataset for DPODataset fine-tuning."""
def __init__(self, data_path: str, tokenizer: transformers.PreTrainedTokenizer, data_args: DataArguments):
super(DPODataset, self).__init__()
# Handle multiple JSON files specified in the data_path
self.list_data_dict = []
if "{" in data_path and "}" in data_path:
base_path, file_pattern = re.match(r"^(.*)\{(.*)\}\.json$", data_path).groups()
file_names = file_pattern.split(",")
rank0_print(f"Loading {file_names} from {base_path}")
data_args.dataset_paths = []
for file_name in file_names:
data_args.dataset_paths.append(f"{base_path}{file_name}.json")
full_path = f"{base_path}{file_name}.json"
rank0_print(f"Loading {full_path}")
cur_data_dict = load_data(full_path)
rank0_print(f"Loaded {len(cur_data_dict)} samples from {full_path}")
self.list_data_dict.extend(cur_data_dict)
elif data_path.endswith(".yaml"):
with open(data_path, "r") as file:
yaml_data = yaml.safe_load(file)
datasets = yaml_data.get("datasets")
# file should be in the format of:
# datasets:
# - json_path: xxxx1.json
# sampling_strategy: first:1000
# - json_path: xxxx2.json
# sampling_strategy: end:3000
# - json_path: xxxx3.json
# sampling_strategy: random:999
data_args.dataset_paths = [dataset.get("json_path") for dataset in datasets]
for dataset in datasets:
json_path = dataset.get("json_path")
sampling_strategy = dataset.get("sampling_strategy", "all")
sampling_number = None
rank0_print(f"Loading {json_path} with {sampling_strategy} sampling strategy")
cur_data_dict = load_data(json_path)
if ":" in sampling_strategy:
sampling_strategy, sampling_number = sampling_strategy.split(":")
if "%" in sampling_number:
sampling_number = math.ceil(int(sampling_number.split("%")[0]) * len(cur_data_dict) / 100)
else:
sampling_number = int(sampling_number)
# Apply the sampling strategy
if sampling_strategy == "first" and sampling_number is not None:
cur_data_dict = cur_data_dict[:sampling_number]
elif sampling_strategy == "end" and sampling_number is not None:
cur_data_dict = cur_data_dict[-sampling_number:]
elif sampling_strategy == "random" and sampling_number is not None:
random.shuffle(cur_data_dict)
cur_data_dict = cur_data_dict[:sampling_number]
rank0_print(f"Loaded {len(cur_data_dict)} samples from {json_path}")
self.list_data_dict.extend(cur_data_dict)
else:
data_args.dataset_paths = [data_path]
rank0_print(f"Loading {data_path}")
cur_data_dict = load_data(data_path)
rank0_print(f"Loaded {len(cur_data_dict)} samples from {data_path}")
self.list_data_dict.extend(cur_data_dict)
rank0_print("Formatting inputs...Skip in lazy mode")
self.tokenizer = tokenizer
self.data_args = data_args
def __len__(self):
return len(self.list_data_dict)
@property
def lengths(self):
length_list = []
for sample in self.list_data_dict:
# Calculate the length of the prompt, answer, chosen, and rejected text
cur_len = len(sample["prompt"].split()) + len(sample["answer"].split()) + len(sample["chosen"].split()) + len(sample["rejected"].split())
# Add additional tokens if an image is present
img_tokens = 128 if "image" in sample else 0
length_list.append(cur_len + img_tokens)
return length_list
@property
def modality_lengths(self):
length_list = []
for sample in self.list_data_dict:
# Calculate the length of the prompt, answer, chosen, and rejected text
cur_len = len(sample["prompt"].split()) + len(sample["answer"].split()) + len(sample["chosen"].split()) + len(sample["rejected"].split())
# If the sample includes a video, the length is positive; otherwise, it is negative
cur_len = cur_len if ("video" in sample or "image" in sample) else -cur_len
length_list.append(cur_len)
return length_list
def process_image(self, image_file):
image_folder = self.data_args.image_folder
processor = self.data_args.image_processor
# print(f"\n\nInspecting the image path, folder = {image_folder}, image={image_file}\n\n")
try:
image = Image.open(os.path.join(image_folder, image_file)).convert("RGB")
except Exception as exn:
print(f"Failed to open image {image_file}. Exception:", exn)
raise exn
image_size = image.size
if self.data_args.image_aspect_ratio == "highres":
image = process_highres_image(image, self.data_args.image_processor, self.data_args.image_grid_pinpoints)
elif self.data_args.image_aspect_ratio == "anyres" or "anyres" in self.data_args.image_aspect_ratio:
image = process_anyres_image(image, self.data_args.image_processor, self.data_args.image_grid_pinpoints)
elif self.data_args.image_aspect_ratio == "crop_split":
image = process_highres_image_crop_split(image, self.data_args)
elif self.data_args.image_aspect_ratio == "pad":
def expand2square(pil_img, background_color):
width, height = pil_img.size
if width == height:
return pil_img
elif width > height:
result = Image.new(pil_img.mode, (width, width), background_color)
result.paste(pil_img, (0, (width - height) // 2))
return result
else:
result = Image.new(pil_img.mode, (height, height), background_color)
result.paste(pil_img, ((height - width) // 2, 0))
return result
image = expand2square(image, tuple(int(x * 255) for x in processor.image_mean))
image = processor.preprocess(image, return_tensors="pt")["pixel_values"][0]
else:
image = processor.preprocess(image, return_tensors="pt")["pixel_values"][0]
return image, image_size, "image"
def __getitem__(self, i) -> Dict[str, torch.Tensor]:
# TODO: define number of retries somewhere else
num_base_retries = 3
num_final_retries = 300
# try the current sample first
for attempt_idx in range(num_base_retries):
try:
sample = self._get_item(i)
return sample
except Exception as e:
# sleep 1s in case it is a cloud disk issue
print(f"[Try #{attempt_idx}] Failed to fetch sample {i}. Exception:", e)
time.sleep(1)
# try other samples, in case it is file corruption issue
for attempt_idx in range(num_base_retries):
try:
next_index = min(i + 1, len(self.list_data_dict) - 1)
# sample_idx = random.choice(range(len(self)))
sample = self._get_item(next_index)
return sample
except Exception as e:
# no need to sleep
print(f"[Try other #{attempt_idx}] Failed to fetch sample {next_index}. Exception:", e)
pass
# still fail, most likely to be path issue or cloud disk issue, retry the same sample for longer
# for attempt_idx in range(num_final_retries):
# try:
# sample = self._get_item(i)
# return sample
# except Exception as e:
# # sleep 1s in case it is a cloud disk issue
# print(f"[Final try #{attempt_idx}] Failed to fetch sample {i}. Exception:", e)
# time.sleep(1)
# Finally raise exception on failing.
assert False, "Failed to fetch sample."
def _get_item(self, i) -> Dict[str, torch.Tensor]:
sources = self.list_data_dict[i]
if isinstance(i, int):
sources = [sources]
assert len(sources) == 1, "Don't know why it is wrapped to a list" # FIXME
suffix = None
if "image" in sources[0]:
image_file = self.list_data_dict[i]["image"]
if type(image_file) is list:
image = [self.process_image(f) for f in image_file]
else:
image = [self.process_image(image_file)]
# sources = preprocess_multimodal(copy.deepcopy([e["conversations"] for e in sources]), self.data_args)
elif "video" in sources[0]: # FIXME: This logic should be largely improved by Yuanhan. It's too messy now.
video_file = self.list_data_dict[i]["video"]
video_folder = self.data_args.video_folder
video_file = os.path.join(video_folder, video_file)
suffix = video_file.split(".")[-1]
if not os.path.exists(video_file):
print("File {} not exist!".format(video_file))
if suffix == "pkl":
video_info = pickle.load(open(video_file, "rb"))
image = torch.from_numpy(video_info["feats"][:, 1:])
input_prompt = video_info["inputs"].replace("...", "")
# replace the default image token with multiple tokens
input_prompt = input_prompt.replace(DEFAULT_IMAGE_TOKEN, DEFAULT_IMAGE_TOKEN * self.data_args.video_token)
sources, query_prompt = preprocess_multimodal_movie(copy.deepcopy([e["conversations"] for e in sources]), self.data_args, input_prompt)
else: # using videoreader
if "shareVideoGPTV" not in video_file and "liangke" not in video_file:
vr = VideoReader(video_file, ctx=cpu(0))
total_frame_num = len(vr)
avg_fps = round(vr.get_avg_fps() / self.data_args.video_fps)
frame_idx = [i for i in range(0, total_frame_num, avg_fps)]
if self.data_args.frames_upbound > 0:
if len(frame_idx) > self.data_args.frames_upbound:
uniform_sampled_frames = np.linspace(0, total_frame_num - 1, self.data_args.frames_upbound, dtype=int)
frame_idx = uniform_sampled_frames.tolist()
video = vr.get_batch(frame_idx).asnumpy()
video = np.array(video)
else:
if "liangke" in video_file:
video_file = self.list_data_dict[i]["video"]
frame_files = [os.path.join(video_file, f) for f in os.listdir(video_file) if os.path.isfile(os.path.join(video_file, f))]
frame_files.sort() # Ensure the frames are sorted if they are named sequentially
# TODO: Hard CODE: Determine the indices for uniformly sampling 10 frames
num_frames_to_sample = 10
total_frames = len(frame_files)
sampled_indices = np.linspace(0, total_frames - 1, num_frames_to_sample, dtype=int)
# Read and store the sampled frames
video = []
for idx in sampled_indices:
frame_path = frame_files[idx]
try:
with Image.open(frame_path) as img:
frame = img.convert("RGB")
video.append(frame)
except IOError:
print(f"Failed to read frame at path: {frame_path}")
processor = self.data_args.image_processor
image = processor.preprocess(video, return_tensors="pt")["pixel_values"]
image = [(image, video[0].size, "video")]
# sources = preprocess_multimodal(copy.deepcopy([e["conversations"] for e in sources]), self.data_args)
else:
sources = copy.deepcopy([e["conversations"] for e in sources])
has_image = ("image" in self.list_data_dict[i]) or ("video" in self.list_data_dict[i])
# data_dict = preprocess(sources, self.tokenizer, has_image=has_image)
data_dict = copy.deepcopy(self.list_data_dict[i]) # inplace modification following
if "prompt" in data_dict:
prompt = data_dict["prompt"]
prompt = prompt.replace("", "").strip()
prompt = "\n" + prompt
data_dict["prompt"] = prompt
else:
prompt = None
if suffix == "pkl":
prompt = [query_prompt]
# image exist in the data
if "image" in self.list_data_dict[i]:
data_dict["image"] = image
elif "video" in self.list_data_dict[i]:
data_dict["image"] = image
elif self.data_args.is_multimodal:
# image does not exist in the data, but the model is multimodal
crop_size = self.data_args.image_processor.crop_size
data_dict["image"] = [
(torch.zeros(1, 3, crop_size["height"], crop_size["width"]), (crop_size["width"], crop_size["height"]), "text"),
]
# prompt exist in the data
data_dict["has_image"] = has_image
return data_dict
@dataclass
class DPODataCollator(DPODataCollatorWithPadding):
"""Collate examples for DPO fine-tuning."""
# tokenizer: transformers.PreTrainedTokenizer
def collate(self, batch):
# first, pad everything to the same length
# 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)
# input_ids = input_ids[:, :self.tokenizer.model_max_length]
# labels = labels[:, :self.tokenizer.model_max_length]
# batch = dict(
# input_ids=input_ids,
# labels=labels,
# attention_mask=input_ids.ne(self.tokenizer.pad_token_id),
# )
padded_batch = {}
for k in batch[0].keys():
if k.endswith("_input_ids") or k.endswith("_attention_mask") or k.endswith("_labels"):
# if "prompt" in k:
# to_pad = [torch.LongTensor(ex[k][::-1]) for ex in batch]
# else:
to_pad = [torch.LongTensor(ex[k]) for ex in batch]
if k.endswith("_input_ids"):
padding_value = self.tokenizer.pad_token_id
elif k.endswith("_labels"):
padding_value = self.label_pad_token_id
else:
continue
# elif k.endswith("_attention_mask"):
# padding_value = self.padding_value
# else:
# raise ValueError(f"Unexpected key in batch '{k}'")
padded_batch[k] = torch.nn.utils.rnn.pad_sequence(to_pad, batch_first=True, padding_value=padding_value)
# for the prompt, flip back so padding is on left side
# if "prompt" in k:
# padded_batch[k] = padded_batch[k].flip(dims=[1])
else:
padded_batch[k] = [ex[k] for ex in batch]
for k in ["chosen_input_ids", "rejected_input_ids"]:
attn_k = k.replace("input_ids", "attention_mask")
padded_batch[attn_k] = padded_batch[k].ne(self.tokenizer.pad_token_id)
return padded_batch
def tokenize_batch_element(self, prompt: str, chosen: str, rejected: str, has_image: bool = True) -> Dict:
"""Tokenize a single batch element.
At this stage, we don't convert to PyTorch tensors yet; we just handle the truncation
in case the prompt + chosen or prompt + rejected responses is/are too long. First
we truncate the prompt; if we're still too long, we truncate the chosen/rejected.
We also create the labels for the chosen/rejected responses, which are of length equal to
the sum of the length of the prompt and the chosen/rejected response, with
label_pad_token_id for the prompt tokens.
"""
# import pdb; pdb.set_trace()
batch = {}
chosen_sources = make_conv(prompt, chosen)
rejected_sources = make_conv(prompt, rejected)
chosen_data_dict = preprocess([chosen_sources], self.tokenizer, has_image=has_image)
# chosen_data_dict['attention_mask'] = chosen_data_dict["input_ids"].ne(self.tokenizer.pad_token_id)
rejected_data_dict = preprocess([rejected_sources], self.tokenizer, has_image=has_image)
# rejected_data_dict['attention_mask'] = rejected_data_dict["input_ids"].ne(self.tokenizer.pad_token_id)
chosen_data_dict = {k: v[0] for k, v in chosen_data_dict.items()}
rejected_data_dict = {k: v[0] for k, v in rejected_data_dict.items()}
for k, toks in {
"chosen": chosen_data_dict,
"rejected": rejected_data_dict,
}.items():
for type_key, tokens in toks.items():
if type_key == "token_type_ids":
continue
batch[f"{k}_{type_key}"] = tokens
return batch
def __call__(self, features: List[Dict[str, Any]]) -> Dict[str, Any]:
tokenized_batch = []
Xs, keys = [], []
for feature in features:
prompt = feature["prompt"]
chosen = feature["chosen"]
rejected = feature["rejected"]
has_image = feature["has_image"]
# Xs.append(feature[has_X])
# keys.append(has_X)
batch_element = self.tokenize_batch_element(prompt, chosen, rejected, has_image=has_image)
tokenized_batch.append(batch_element)
# return collated batch
padded_batch = self.collate(tokenized_batch)
# import pdb;pdb.set_trace()
if "image" in features[0]:
# instances[1]['image'][0][0].shape
# torch.Size([5, 3, 224, 224])
images = [instance["image"] for instance in features]
padded_batch["image_sizes"] = [im[1] for im_list in images for im in im_list]
padded_batch["modalities"] = [im[2] for im_list in images for im in im_list]
images = [im[0] for im_list in images for im in im_list]
# import pdb;pdb.set_trace()
padded_batch["images"] = images
# padded_batch["images"] =[padded_batch["modalities"], images]
return padded_batch
def make_dpo_data_module(tokenizer: transformers.PreTrainedTokenizer, data_args) -> Dict:
"""Make dataset and collator for supervised fine-tuning."""
train_dataset = DPODataset(tokenizer=tokenizer, data_path=data_args.data_path, data_args=data_args)
return train_dataset
def get_model(model_args, training_args, bnb_model_from_pretrained_args):
assert training_args.attn_implementation
if training_args.attn_implementation == "sdpa" and torch.__version__ < "2.1.2":
raise ValueError("The 'sdpa' attention implementation requires torch version 2.1.2 or higher.")
######################### Overwrite config #########################
customized_kwargs = dict()
customized_kwargs.update(bnb_model_from_pretrained_args)
overwrite_config = {}
cfg_pretrained = None
if "qwen" in model_args.model_name_or_path.lower():
cfg_pretrained = LlavaQwenConfig.from_pretrained(model_args.model_name_or_path)
elif "mistral" in model_args.model_name_or_path.lower() or "zephyr" in model_args.model_name_or_path.lower():
cfg_pretrained = LlavaMistralConfig.from_pretrained(model_args.model_name_or_path)
elif (
"wizardlm-2" in model_args.model_name_or_path.lower()
or "vicuna" in model_args.model_name_or_path.lower()
or "llama" in model_args.model_name_or_path.lower()
or "yi" in model_args.model_name_or_path.lower()
or "nous-hermes" in model_args.model_name_or_path.lower()
and "wizard-2" in model_args.model_name_or_path.lower()
):
cfg_pretrained = LlavaConfig.from_pretrained(model_args.model_name_or_path)
else:
cfg_pretrained = AutoConfig.from_pretrained(model_args.model_name_or_path)
if model_args.rope_scaling_factor is not None and model_args.rope_scaling_type is not None and cfg_pretrained is not None:
overwrite_config["rope_scaling"] = {
"factor": model_args.rope_scaling_factor,
"type": model_args.rope_scaling_type,
}
if training_args.model_max_length is None:
training_args.model_max_length = cfg_pretrained.max_position_embeddings * model_args.rope_scaling_factor
overwrite_config["max_sequence_length"] = training_args.model_max_length
assert training_args.model_max_length == int(cfg_pretrained.max_position_embeddings * model_args.rope_scaling_factor), print(
f"model_max_length: {training_args.model_max_length}, max_position_embeddings: {cfg_pretrained.max_position_embeddings}, rope_scaling_factor: {model_args.rope_scaling_factor}"
)
# overwrite_config["max_sequence_length"] = model_args.max_sequence_length
# overwrite_config["tokenizer_model_max_length"] = model_args.tokenizer_model_max_length
if model_args.mm_spatial_pool_stride is not None and model_args.mm_spatial_pool_out_channels is not None and model_args.mm_spatial_pool_mode is not None and model_args.mm_resampler_type is not None and cfg_pretrained is not None:
overwrite_config["mm_resampler_type"] = model_args.mm_resampler_type
overwrite_config["mm_spatial_pool_stride"] = model_args.mm_spatial_pool_stride
overwrite_config["mm_spatial_pool_out_channels"] = model_args.mm_spatial_pool_out_channels
overwrite_config["mm_spatial_pool_mode"] = model_args.mm_spatial_pool_mode
if overwrite_config:
rank0_print(f"Overwriting config with {overwrite_config}")
for k, v in overwrite_config.items():
setattr(cfg_pretrained, k, v)
customized_kwargs["config"] = cfg_pretrained
######################### Finish Overwrite ###########################
ref_model = None
if model_args.model_class_name is not None:
actual_model_class_name = f"{model_args.model_class_name}ForCausalLM"
model_class = getattr(transformers, actual_model_class_name)
rank0_print(f"Using model class {model_class} from {model_args.model_class_name}")
model = model_class.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
elif model_args.vision_tower is not None:
if "mixtral" in model_args.model_name_or_path.lower():
model = LlavaMixtralForCausalLM.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
from transformers.models.mixtral.modeling_mixtral import MixtralSparseMoeBlock
deepspeed.utils.set_z3_leaf_modules(model, [MixtralSparseMoeBlock])
elif "mistral" in model_args.model_name_or_path.lower() or "zephyr" in model_args.model_name_or_path.lower():
model = LlavaMistralForCausalLM.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
elif (
"wizardlm-2" in model_args.model_name_or_path.lower()
or "vicuna" in model_args.model_name_or_path.lower()
or "llama" in model_args.model_name_or_path.lower()
or "yi" in model_args.model_name_or_path.lower()
or "nous-hermes" in model_args.model_name_or_path.lower()
and "wizard-2" in model_args.model_name_or_path.lower()
):
model = LlavaLlamaForCausalLM.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
if "zero3" in training_args.deepspeed:
rank0_print("#### Initialize reference model #####")
ref_model = LlavaLlamaForCausalLM.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
elif "qwen" in model_args.model_name_or_path.lower() or "quyen" in model_args.model_name_or_path.lower():
if "moe" in model_args.model_name_or_path.lower():
model = LlavaQwenMoeForCausalLM.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
from transformers.models.qwen2_moe.modeling_qwen2_moe import Qwen2MoeSparseMoeBlock
deepspeed.utils.set_z3_leaf_modules(model, [Qwen2MoeSparseMoeBlock])
else:
model = LlavaQwenForCausalLM.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
if "zero3" in training_args.deepspeed:
rank0_print("#### Initialize reference model #####")
ref_model = LlavaQwenForCausalLM.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
elif "gemma" in model_args.model_name_or_path.lower():
model = LlavaGemmaForCausalLM.from_pretrained(
model_args.model_name_or_path,
cache_dir=training_args.cache_dir,
attn_implementation=training_args.attn_implementation,
torch_dtype=(torch.bfloat16 if training_args.bf16 else None),
low_cpu_mem_usage=False,
**customized_kwargs,
)
else:
raise ValueError(f"Unknown model class {model_args}")
else:
model = transformers.LlamaForCausalLM.from_pretrained(
model_args.model_name_or_path, cache_dir=training_args.cache_dir, attn_implementation=training_args.attn_implementation, torch_dtype=(torch.bfloat16 if training_args.bf16 else None), **customized_kwargs
)
return model, ref_model
def train(attn_implementation=None):
global local_rank
parser = transformers.HfArgumentParser((ModelArguments, DataArguments, TrainingArguments))
model_args, data_args, training_args = parser.parse_args_into_dataclasses()
if training_args.verbose_logging:
rank0_print(f"Inspecting experiment hyperparameters:\n")
rank0_print(f"model_args = {vars(model_args)}\n\n")
rank0_print(f"data_args = {vars(data_args)}\n\n")
rank0_print(f"training_args = {vars(training_args)}\n\n")
# rank0_print(f"evaluation_args = {vars(evaluation_args)}\n\n")
local_rank = training_args.local_rank
compute_dtype = torch.float16 if training_args.fp16 else (torch.bfloat16 if training_args.bf16 else torch.float32)
bnb_model_from_pretrained_args = {}
if training_args.bits in [4, 8]:
from transformers import BitsAndBytesConfig
bnb_model_from_pretrained_args.update(
dict(
device_map={"": training_args.device},
load_in_4bit=training_args.bits == 4,
load_in_8bit=training_args.bits == 8,
quantization_config=BitsAndBytesConfig(
load_in_4bit=training_args.bits == 4,
load_in_8bit=training_args.bits == 8,
llm_int8_threshold=6.0,
llm_int8_has_fp16_weight=False,
bnb_4bit_compute_dtype=compute_dtype,
bnb_4bit_use_double_quant=training_args.double_quant,
bnb_4bit_quant_type=training_args.quant_type, # {'fp4', 'nf4'}
),
)
)
model, ref_model = get_model(model_args, training_args, bnb_model_from_pretrained_args)
model.config.use_cache = False
if model_args.freeze_backbone:
model.model.requires_grad_(False)
if training_args.bits in [4, 8]:
from peft import prepare_model_for_kbit_training
model.config.torch_dtype = torch.float32 if training_args.fp16 else (torch.bfloat16 if training_args.bf16 else torch.float32)
model = prepare_model_for_kbit_training(model, use_gradient_checkpointing=training_args.gradient_checkpointing)
if training_args.gradient_checkpointing:
if hasattr(model, "enable_input_require_grads"):
model.enable_input_require_grads()
if ref_model is not None:
ref_model.enable_input_require_grads()
else:
def make_inputs_require_grad(module, input, output):
output.requires_grad_(True)
model.get_input_embeddings().register_forward_hook(make_inputs_require_grad)
if ref_model is not None:
ref_model.get_input_embeddings().register_forward_hook(make_inputs_require_grad)
if training_args.lora_enable:
from peft import LoraConfig, get_peft_model
lora_config = LoraConfig(
r=training_args.lora_r,
lora_alpha=training_args.lora_alpha,
target_modules=find_all_linear_names(model),
lora_dropout=training_args.lora_dropout,
bias=training_args.lora_bias,
task_type="CAUSAL_LM",
)
if training_args.bits == 16:
if training_args.bf16:
model.to(torch.bfloat16)
if training_args.fp16:
model.to(torch.float16)
rank0_print("Adding LoRA adapters...")
model = get_peft_model(model, lora_config)
if "mpt" in model_args.model_name_or_path:
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")
elif "mistral" in model_args.model_name_or_path.lower() or "mixtral" in model_args.model_name_or_path.lower() or "zephyr" in model_args.model_name_or_path.lower():
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="left")
elif "qwen" in model_args.model_name_or_path.lower():
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")
else: # for all other models
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,
)
rank0_print(f"Prompt version: {model_args.version}")
if model_args.version == "v0":
if tokenizer.pad_token is None:
smart_tokenizer_and_embedding_resize(
special_tokens_dict=dict(pad_token="[PAD]"),
tokenizer=tokenizer,
model=model,
)
elif model_args.version == "v0.5":
tokenizer.pad_token = tokenizer.unk_token
else:
if tokenizer.unk_token is not None:
tokenizer.pad_token = tokenizer.unk_token
if model_args.version in conversation_lib.conv_templates:
conversation_lib.default_conversation = conversation_lib.conv_templates[model_args.version]
else:
conversation_lib.default_conversation = conversation_lib.conv_templates["vicuna_v1"]
if model_args.vision_tower is not None:
model.get_model().initialize_vision_modules(model_args=model_args, fsdp=training_args.fsdp)
vision_tower = model.get_vision_tower()
vision_tower.to(dtype=torch.bfloat16 if training_args.bf16 else torch.float16, device=training_args.device)
data_args.image_processor = vision_tower.image_processor
data_args.is_multimodal = True
model.config.image_aspect_ratio = data_args.image_aspect_ratio
if data_args.image_grid_pinpoints is not None:
# for input like "(1x1)...(3x3)", convert to [(1, 1), (2, 1), (3, 1), (1, 2), (2, 2), (3, 2), (1, 3), (2, 3), (3, 3)]
if "x" in data_args.image_grid_pinpoints and "..." in data_args.image_grid_pinpoints:
vis_encoder_size = data_args.image_processor.size[0]
matches = re.findall(r"\((\d+)x(\d+)\)", data_args.image_grid_pinpoints)
range_start = tuple(map(int, matches[0]))
range_end = tuple(map(int, matches[-1]))
grid_pinpoints = [(i, j) for i in range(range_start[0], range_end[0] + 1) for j in range(range_start[1], range_end[1] + 1)]
grid_pinpoints = [[dim * vis_encoder_size for dim in pair] for pair in grid_pinpoints]
data_args.image_grid_pinpoints = grid_pinpoints
elif "x" in data_args.image_grid_pinpoints:
vis_encoder_size = data_args.image_processor.size[0]
assert vis_encoder_size in [224, 336, 384, 448, 512], "vis_encoder_size should be in [224, 336, 384, 448, 512]"
grid_pinpoints = data_args.image_grid_pinpoints.replace(" ", "").replace("x", ",")[1:-1].split("),(")
data_args.image_grid_pinpoints = [[int(x) * vis_encoder_size for x in item.split(",")] for item in grid_pinpoints]
else:
data_args.image_grid_pinpoints = ast.literal_eval(data_args.image_grid_pinpoints) # for backward compatibility
model.config.image_grid_pinpoints = data_args.image_grid_pinpoints
model.config.image_crop_resolution = data_args.image_crop_resolution
model.config.image_split_resolution = data_args.image_split_resolution
model.config.tokenizer_padding_side = tokenizer.padding_side
model.config.tokenizer_model_max_length = tokenizer.model_max_length
### Deciding train which part of the model
if model_args.mm_tunable_parts is None: # traditional way of deciding which part to train
model.config.tune_mm_mlp_adapter = training_args.tune_mm_mlp_adapter = model_args.tune_mm_mlp_adapter
model.config.tune_mm_vision_resampler = training_args.tune_mm_vision_resampler = model_args.tune_mm_vision_resampler
if model_args.tune_mm_mlp_adapter or model_args.tune_mm_vision_resampler:
model.requires_grad_(False)
if model_args.tune_mm_mlp_adapter:
for p in model.get_model().mm_projector.parameters():
p.requires_grad = True
if model_args.tune_mm_vision_resampler:
for p in model.get_model().vision_resampler.parameters():
p.requires_grad = True
model.config.freeze_mm_mlp_adapter = training_args.freeze_mm_mlp_adapter
if training_args.freeze_mm_mlp_adapter:
for p in model.get_model().mm_projector.parameters():
p.requires_grad = False
model.config.freeze_mm_vision_resampler = training_args.freeze_mm_vision_resampler
if training_args.freeze_mm_vision_resampler:
for p in model.get_model().vision_resampler.parameters():
p.requires_grad = False
model.config.unfreeze_mm_vision_tower = model_args.unfreeze_mm_vision_tower
if model_args.unfreeze_mm_vision_tower:
vision_tower.requires_grad_(True)
else:
vision_tower.requires_grad_(False)
else:
rank0_print(f"Using mm_tunable_parts: {model_args.mm_tunable_parts}")
model.config.mm_tunable_parts = training_args.mm_tunable_parts = model_args.mm_tunable_parts
# Set the entire model to not require gradients by default
model.requires_grad_(False)
vision_tower.requires_grad_(False)
model.get_model().mm_projector.requires_grad_(False)
model.get_model().vision_resampler.requires_grad_(False)
# Parse the mm_tunable_parts to decide which parts to unfreeze
tunable_parts = model_args.mm_tunable_parts.split(",")
if "mm_mlp_adapter" in tunable_parts:
for p in model.get_model().mm_projector.parameters():
p.requires_grad = True
if "mm_vision_resampler" in tunable_parts:
for p in model.get_model().vision_resampler.parameters():
p.requires_grad = True
if "mm_vision_tower" in tunable_parts:
for name, param in model.named_parameters():
if "vision_tower" in name:
param.requires_grad_(True)
if "mm_language_model" in tunable_parts:
for name, param in model.named_parameters():
if "vision_tower" not in name and "mm_projector" not in name and "vision_resampler" not in name:
param.requires_grad_(True)
total_params = sum(p.ds_numel if hasattr(p, "ds_numel") else p.numel() for p in model.parameters())
trainable_params = sum(p.ds_numel if hasattr(p, "ds_numel") else p.numel() for p in model.parameters() if p.requires_grad)
rank0_print(f"Total parameters: ~{total_params/1e6:.2f} MB)")
rank0_print(f"Trainable parameters: ~{trainable_params/1e6:.2f} MB)")
if training_args.bits in [4, 8]:
model.get_model().mm_projector.to(dtype=compute_dtype, device=training_args.device)
model.config.mm_use_im_start_end = data_args.mm_use_im_start_end = model_args.mm_use_im_start_end
model.config.mm_projector_lr = training_args.mm_projector_lr
model.config.mm_vision_tower_lr = training_args.mm_vision_tower_lr
training_args.use_im_start_end = model_args.mm_use_im_start_end
model.config.mm_use_im_patch_token = model_args.mm_use_im_patch_token
model.initialize_vision_tokenizer(model_args, tokenizer=tokenizer)
if ref_model is not None:
ref_model.get_model().initialize_vision_modules(model_args=model_args, fsdp=training_args.fsdp)
ref_vision_tower = ref_model.get_vision_tower()
ref_vision_tower.to(dtype=torch.bfloat16 if training_args.bf16 else torch.float16, device=training_args.device)
ref_model.config.image_aspect_ratio = data_args.image_aspect_ratio
ref_model.config.image_grid_pinpoints = data_args.image_grid_pinpoints
ref_model.config.image_crop_resolution = data_args.image_crop_resolution
ref_model.config.image_split_resolution = data_args.image_split_resolution
ref_model.config.tokenizer_padding_side = tokenizer.padding_side
ref_model.config.tokenizer_model_max_length = tokenizer.model_max_length
ref_model.config.mm_use_im_start_end = data_args.mm_use_im_start_end
ref_model.config.mm_use_im_patch_token = model_args.mm_use_im_patch_token
ref_model.initialize_vision_tokenizer(model_args, tokenizer=tokenizer)
parameter_names = [n for n, _ in ref_model.named_parameters()]
for param_name in parameter_names:
param = ref_model.get_parameter(param_name)
param.requires_grad = False
ref_model.eval()
if training_args.bits in [4, 8]:
from peft.tuners.lora import LoraLayer
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)
train_dataset = make_dpo_data_module(tokenizer=tokenizer, data_args=data_args)
data_collator = DPODataCollator(
tokenizer,
label_pad_token_id=IGNORE_INDEX,
pad_token_id=tokenizer.pad_token_id,
)
trainer = LLaVADPOTrainer(
model,
ref_model,
args=training_args,
dpo_alpha=training_args.dpo_alpha,
beta=training_args.beta,
gamma=training_args.gamma,
train_dataset=train_dataset,
eval_dataset=None,
data_collator=data_collator,
tokenizer=tokenizer,
max_length=training_args.model_max_length,
generate_during_eval=False, # training_args.generate_during_eval,
precompute_ref_log_probs=training_args.precompute_ref_log_probs,
)
if list(pathlib.Path(training_args.output_dir).glob("checkpoint-*")):
trainer.train(resume_from_checkpoint=True)
else:
trainer.train()
trainer.save_state()
model.config.use_cache = True
if training_args.lora_enable:
state_dict = get_peft_state_maybe_zero_3(model.named_parameters(), training_args.lora_bias)
non_lora_state_dict = get_peft_state_non_lora_maybe_zero_3(model.named_parameters())
if training_args.local_rank == 0 or training_args.local_rank == -1:
if hasattr(model, "config"):
model.config.save_pretrained(training_args.output_dir)
if hasattr(model, "generation_config"):
model.generation_config.save_pretrained(training_args.output_dir)
model.save_pretrained(training_args.output_dir, state_dict=state_dict)
torch.save(non_lora_state_dict, os.path.join(training_args.output_dir, "non_lora_trainables.bin"))
else:
safe_save_model_for_hf_trainer(trainer=trainer, output_dir=training_args.output_dir)
rank0_print(f"Model saved to {training_args.output_dir}")
if __name__ == "__main__":
train()