File size: 10,704 Bytes
252711e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 |
import argparse
import torch
import os
import json
from tqdm import tqdm
import shortuuid
from llava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN, DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN
from llava.conversation import conv_templates, SeparatorStyle
from llava.model.builder import load_pretrained_model
from llava.utils import disable_torch_init
from llava.mm_utils import tokenizer_image_token, get_model_name_from_path, KeywordsStoppingCriteria
from llava.constants import IGNORE_INDEX, DEFAULT_IMAGE_TOKEN, DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN, IMAGE_TOKEN_INDEX
from typing import Dict, Optional, Sequence, List
import transformers
import re
from PIL import Image
import math
def split_list(lst, n):
"""Split a list into n (roughly) equal-sized chunks"""
chunk_size = math.ceil(len(lst) / n) # integer division
return [lst[i:i+chunk_size] for i in range(0, len(lst), chunk_size)]
def get_chunk(lst, n, k):
chunks = split_list(lst, n)
return chunks[k]
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 = [], []
source = 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 sentence["value"] is not None and "<image>" in sentence["value"]:
num_image = len(re.findall(DEFAULT_IMAGE_TOKEN, sentence["value"]))
texts = sentence["value"].split('<image>')
_input_id = tokenizer(role).input_ids + nl_tokens
for i,text in enumerate(texts):
_input_id += tokenizer(text).input_ids
if i<len(texts)-1:
_input_id += [IMAGE_TOKEN_INDEX] + nl_tokens
_input_id += [im_end] + nl_tokens
assert sum([i==IMAGE_TOKEN_INDEX for i in _input_id])==num_image
else:
if sentence["value"] is None:
_input_id = tokenizer(role).input_ids + 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
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 input_ids
def eval_model(args):
# Model
disable_torch_init()
model_path = os.path.expanduser(args.model_path)
model_name = get_model_name_from_path(model_path)
tokenizer, model, image_processor, context_len = load_pretrained_model(model_path, args.model_base, model_name)
# Data
with open(os.path.expanduser(args.question_file)) as f:
questions = json.load(f)
questions = get_chunk(questions, args.num_chunks, args.chunk_idx)
answers_file = os.path.expanduser(args.answers_file)
os.makedirs(os.path.dirname(answers_file), exist_ok=True)
ans_file = open(answers_file, "w")
for line in tqdm(questions):
idx = line["sample_id"]
question_type = line["metadata"]["question_type"]
dataset_name = line["metadata"]["dataset"]
gt = line["conversations"][1]["value"]
image_files = line["image"]
qs = line["conversations"][0]["value"]
cur_prompt = args.extra_prompt + qs
args.conv_mode = "qwen_1_5"
conv = conv_templates[args.conv_mode].copy()
conv.append_message(conv.roles[0], qs)
conv.append_message(conv.roles[1], None)
prompt = conv.get_prompt()
input_ids = preprocess_qwen([line["conversations"][0],{'from': 'gpt','value': None}], tokenizer, has_image=True).cuda()
img_num = list(input_ids.squeeze()).count(IMAGE_TOKEN_INDEX)
image_tensors = []
for image_file in image_files:
image = Image.open(os.path.join(args.image_folder, image_file))
image_tensor = image_processor.preprocess(image, return_tensors='pt')['pixel_values']
image_tensors.append(image_tensor.half().cuda())
# image_tensors = torch.cat(image_tensors, dim=0)
stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
keywords = [stop_str]
stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)
with torch.inference_mode():
output_ids = model.generate(
input_ids,
images=image_tensors,
do_sample=True if args.temperature > 0 else False,
temperature=args.temperature,
top_p=args.top_p,
num_beams=args.num_beams,
# no_repeat_ngram_size=3,
max_new_tokens=1024,
use_cache=True)
outputs = tokenizer.batch_decode(output_ids, skip_special_tokens=True)[0]
outputs = outputs.strip()
if outputs.endswith(stop_str):
outputs = outputs[:-len(stop_str)]
outputs = outputs.strip()
ans_id = shortuuid.uuid()
ans_file.write(json.dumps({
"dataset": dataset_name,
"sample_id": idx,
"prompt": cur_prompt,
"pred_response": outputs,
"gt_response": gt,
"shortuuid": ans_id,
"model_id": model_name,
"question_type": question_type,
}) + "\n")
ans_file.flush()
if len(line["conversations"]) > 2:
for i in range(2, len(line["conversations"]), 2):
input_ids = torch.cat((input_ids, output_ids), dim=1)
gt = line["conversations"][i + 1]["value"]
qs = line["conversations"][i]["value"]
cur_prompt = args.extra_prompt + qs
args.conv_mode = "qwen_1_5"
conv = conv_templates[args.conv_mode].copy()
conv.append_message(conv.roles[0], qs)
conv.append_message(conv.roles[1], None)
prompt = conv.get_prompt()
input_ids_new = preprocess_qwen([line["conversations"][i],{'from': 'gpt','value': None}], tokenizer, has_image=True).cuda()
input_ids = torch.cat((input_ids, input_ids_new), dim=1)
img_num = list(input_ids_new.squeeze()).count(IMAGE_TOKEN_INDEX)
stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
keywords = [stop_str]
stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)
with torch.inference_mode():
output_ids = model.generate(
input_ids,
images=image_tensors,
do_sample=True if args.temperature > 0 else False,
temperature=args.temperature,
top_p=args.top_p,
num_beams=args.num_beams,
# no_repeat_ngram_size=3,
max_new_tokens=1024,
use_cache=True)
outputs = tokenizer.batch_decode(output_ids, skip_special_tokens=True)[0]
outputs = outputs.strip()
if outputs.endswith(stop_str):
outputs = outputs[:-len(stop_str)]
outputs = outputs.strip()
ans_id = shortuuid.uuid()
ans_file.write(json.dumps({
"dataset": dataset_name,
"sample_id": idx,
"prompt": cur_prompt,
"pred_response": outputs,
"gt_response": gt,
"shortuuid": ans_id,
"model_id": model_name,
"question_type": question_type,
}) + "\n")
ans_file.flush()
ans_file.close()
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--model-path", type=str, default="facebook/opt-350m")
parser.add_argument("--model-base", type=str, default=None)
parser.add_argument("--image-folder", type=str, default="")
parser.add_argument("--extra-prompt", type=str, default="")
parser.add_argument("--question-file", type=str, default="tables/question.jsonl")
parser.add_argument("--answers-file", type=str, default="answer.jsonl")
parser.add_argument("--conv-mode", type=str, default="llava_v1")
parser.add_argument("--num-chunks", type=int, default=1)
parser.add_argument("--chunk-idx", type=int, default=0)
parser.add_argument("--temperature", type=float, default=0.2)
parser.add_argument("--top_p", type=float, default=None)
parser.add_argument("--num_beams", type=int, default=1)
parser.add_argument("--test_size", type=int, default=10000000)
args = parser.parse_args()
eval_model(args) |