File size: 1,570 Bytes
03561be
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
visual_datasets = [
    dict(
        type="llava",
        vis_root="data/coco/train2017",
        ann_paths=[
            "data/llava/detail_23k.json",
            "data/llava/complex_reasoning_77k.json",
        ],
    ),
    dict(
        type="llava_dial",
        vis_root="data/coco/train2017",
        ann_paths=[
            "data/llava/conversation_58k.json",
        ],
    ),
    dict(
        type="aokvqa",
        vis_root="data/coco/images",
        ann_paths=[
            "data/aokvqa/annotations/aokvqa_v1p0_train.json",
        ],
        sample=5000,
    ),
    dict(
        type="minigpt4",
        vis_root="data/cc_sbu_align/image",
        ann_paths=[
            "data/cc_sbu_align/filter_cap.json",
        ],
    ),
    dict(
        type="coco_caption",
        vis_root="data/coco",
        ann_paths=[
            "data/coco/annotations/coco_karpathy_train_converted.json",
            "data/coco/annotations/coco_karpathy_val.json",
        ],
        sample=512,
    ),
    dict(
        type="ocr_vqa",
        vis_root="data/OCR_VQA/image",
        ann_paths=[
            "data/OCR_VQA/downloaded_dataset.json",
        ],
        sample=512,
    ),
]

language_datasets = [
    dict(
        type="dolly",
        ann_path="data/dolly/databricks-dolly-15k.jsonl",
    ),
    dict(
        type="alpaca_gpt4",
        ann_path="data/alpaca_gpt4/alpaca_gpt4_data.json",
    ),
    dict(
        type="baize",
        ann_path="data/baize/quora_chat_data.json",
    ),
]