File size: 4,556 Bytes
431dab7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7222efe
431dab7
 
 
 
 
 
7222efe
 
 
 
 
 
431dab7
2533bec
 
2fb0735
7222efe
2533bec
7222efe
 
2533bec
7222efe
2533bec
431dab7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
10ffe76
431dab7
 
 
 
10ffe76
431dab7
 
 
 
 
 
 
 
 
 
 
2533bec
431dab7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
43605d2
 
 
 
 
 
 
 
431dab7
 
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
import gradio as gr
from huggingface_hub import InferenceClient

from transformers import AutoTokenizer
from llava.model.language_model.llava_mistral import LlavaMistralForCausalLM
from llava.model.builder import load_pretrained_model
from llava.mm_utils import (
    process_images,
    tokenizer_image_token,
    get_model_name_from_path,
)
from llava.constants import (
    IMAGE_TOKEN_INDEX,
    DEFAULT_IMAGE_TOKEN,
    DEFAULT_IM_START_TOKEN,
    DEFAULT_IM_END_TOKEN,
    IMAGE_PLACEHOLDER,
)
from llava.conversation import conv_templates, SeparatorStyle

import argparse
import torch
import requests
from PIL import Image
from io import BytesIO
import re

parser = argparse.ArgumentParser()
parser.add_argument("--model-path", type=str, default="liuhaotian/llava-v1.6-mistral-7b")
parser.add_argument("--image-file", type=str, required=True)
parser.add_argument("--inference-type", type=str, default="auto")
parser.add_argument("--prompt", type=str, default="Explain this image")
cmd_args = parser.parse_args()

# Line 138 uncomment the cuda() to use GPUs

# device = "cpu"
device = cmd_args.inference_type

prompt = cmd_args.prompt
image_file = cmd_args.image_file

model_path = cmd_args.model_path



# Functions for inference
def image_parser(args):
    out = args.image_file.split(args.sep)
    return out


def load_image(image_file):
    if image_file.startswith("http") or image_file.startswith("https"):
        response = requests.get(image_file)
        image = Image.open(BytesIO(response.content)).convert("RGB")
    else:
        image = Image.open(image_file).convert("RGB")
    return image


def load_images(image_files):
    out = []
    for image_file in image_files:
        image = load_image(image_file)
        out.append(image)
    return out


model_name = get_model_name_from_path('llava-v1.6-mistral-7b')

args = type('Args', (), {
    "model_path": model_path,
    "model_base": None,
    "model_name": model_name,
    "query": prompt,
    "conv_mode": None,
    "image_file": image_file,
    "sep": ",",
    "temperature": 0,
    "top_p": None,
    "num_beams": 1,
    "max_new_tokens": 512
})()

tokenizer, model, image_processor, context_len = load_pretrained_model(
        model_path, None, model_name, device_map=device
    )

qs = args.query
image_token_se = DEFAULT_IM_START_TOKEN + DEFAULT_IMAGE_TOKEN + DEFAULT_IM_END_TOKEN
if IMAGE_PLACEHOLDER in qs:
    if model.config.mm_use_im_start_end:
        qs = re.sub(IMAGE_PLACEHOLDER, image_token_se, qs)
    else:
        qs = re.sub(IMAGE_PLACEHOLDER, DEFAULT_IMAGE_TOKEN, qs)
else:
    if model.config.mm_use_im_start_end:
        qs = image_token_se + "\n" + qs
    else:
        qs = DEFAULT_IMAGE_TOKEN + "\n" + qs

if "llama-2" in model_name.lower():
    conv_mode = "llava_llama_2"
elif "mistral" in model_name.lower():
    conv_mode = "mistral_instruct"
elif "v1.6-34b" in model_name.lower():
    conv_mode = "chatml_direct"
elif "v1" in model_name.lower():
    conv_mode = "llava_v1"
elif "mpt" in model_name.lower():
    conv_mode = "mpt"
else:
    conv_mode = "llava_v0"

if args.conv_mode is not None and conv_mode != args.conv_mode:
    print(
        "[WARNING] the auto inferred conversation mode is {}, while `--conv-mode` is {}, using {}".format(
            conv_mode, args.conv_mode, args.conv_mode
        )
    )
else:
    args.conv_mode = conv_mode

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()

image_files = image_parser(args)
images = load_images(image_files)
image_sizes = [x.size for x in images]
images_tensor = process_images(
    images,
    image_processor,
    model.config
).to(model.device, dtype=torch.float16)

input_ids = (
    tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors="pt")
    .unsqueeze(0)
    # .cuda()
)

with torch.inference_mode():
    output_ids = model.generate(
        input_ids,
        images=images_tensor,
        image_sizes=image_sizes,
        do_sample=True if args.temperature > 0 else False,
        temperature=args.temperature,
        top_p=args.top_p,
        num_beams=args.num_beams,
        max_new_tokens=args.max_new_tokens,
        use_cache=True,
    )

outputs = tokenizer.batch_decode(output_ids, skip_special_tokens=True)[0].strip()

if "dataset1" in image_file:
    print("Num of words: ", len(outputs))
elif "dataset2" in image_file:
    print()
else:
    print("Is single word?", len((outputs).split()) == 1)

print(outputs)
# End Llava inference