File size: 3,799 Bytes
0fd5fe3 |
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 |
import os
import torch
from transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessor
from qwen_vl_utils import process_vision_info
import json
from tqdm import tqdm
import os
def read_json(file_path):
with open(file_path, 'r', encoding='utf-8') as file:
data = json.load(file)
return data
def write_json(file_path, data):
with open(file_path, 'w', encoding='utf-8') as file:
json.dump(data, file, ensure_ascii=False, indent=4)
# default: Load the model on the available device(s)
print(torch.cuda.device_count())
model_path = "/home/zbz5349/WorkSpace/aigeeks/Qwen2.5-VL/LLaMA-Factory/output/Qwen2.5-VL-3B_all"
# model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
# model_path, torch_dtype="auto", device_map="auto"
# )
# We recommend enabling flash_attention_2 for better acceleration and memory saving, especially in multi-image and video scenarios.
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
model_path,
torch_dtype=torch.bfloat16,
attn_implementation="flash_attention_2",
device_map="auto",
)
# default processor
processor = AutoProcessor.from_pretrained(model_path)
print(model.device)
data = read_json('/home/zbz5349/WorkSpace/aigeeks/Qwen2.5-VL/LLaMA-Factory/data/Percption_.json')
save_data = []
correct_num = 0
begin = 0
end = len(data)
batch_size = 1
for batch_idx in tqdm(range(begin, end, batch_size)):
batch = data[batch_idx:batch_idx + batch_size]
image_list = []
input_text_list = []
data_list = []
save_list = []
sd_ans = []
# while True:
for idx, i in enumerate(batch):
save_ = {
"role": "user",
"content": [
{
"type": "video",
"video": "file:///path/to/video1.mp4",
"max_pixels": 360 * 420,
"fps": 1.0,
},
{"type": "text", "text": "Describe this video."},
],
"answer":"None",
"result":"None",
}
messages = {
"role": "user",
"content": [
{
"type": "video",
"video": "file:///path/to/video1.mp4",
"max_pixels": 360 * 420,
"fps": 1.0,
},
{"type": "text", "text": "Describe this video."},
],
}
video_path = i['videos']
question = i['messages'][0]['content']
answer = i['messages'][1]['content']
messages['content'][0]['video'] = video_path
messages['content'][1]['text'] = question
save_['content'][0]['video'] = video_path
save_['content'][1]['text'] = question
save_['answer'] = answer
sd_ans.append(answer)
data_list.append(messages)
save_list.append(save_)
text = processor.apply_chat_template(data_list, tokenize=False, add_generation_prompt=True)
image_inputs, video_inputs, video_kwargs = process_vision_info(data_list, return_video_kwargs=True)
fps = 1
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
**video_kwargs,
)
inputs = inputs.to(model.device)
# Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
for idx,x in enumerate(output_text):
save_list[idx]['result'] = x
save_data.append(save_list[idx])
print("correct_num", correct_num)
write_json("infer_finetune_percption.json",save_data)
|