Spaces:
Running
Running
upload results logs for reference
Browse files- Archive/BGE-VL-v1.5-mmeb.json +13 -0
- Archive/BGE-VL-v1.5-zs.json +13 -0
- Archive/LLaVE.json +35 -0
- Archive/UniME_results.json +35 -0
- Archive/meta-score-BGE-VL-v1.5-mmeb.json +51 -0
- Archive/meta-score-BGE-VL-v1.5-zs.json +51 -0
- Archive/mmE5_mmeb.json +13 -0
- Archive/mmeb-qqmm.json +1 -0
Archive/BGE-VL-v1.5-mmeb.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"Model": "BGE-VL-v1.5 (FT; LlaVA-1.6-Mistral)",
|
4 |
+
"URL": "https://huggingface.co/BAAI/BGE-VL-v1.5-mmeb",
|
5 |
+
"Model Size(B)": 7.57,
|
6 |
+
"Data Source": "Self-Reported",
|
7 |
+
"V1-Overall": 69.4,
|
8 |
+
"I-CLS": 63.7,
|
9 |
+
"I-QA": 64.9,
|
10 |
+
"I-RET": 72.2,
|
11 |
+
"I-VG": 86.6
|
12 |
+
}
|
13 |
+
]
|
Archive/BGE-VL-v1.5-zs.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"Model": "BGE-VL-v1.5 (zeroshot; LlaVA-1.6-Mistral)",
|
4 |
+
"URL": "https://huggingface.co/BAAI/BGE-VL-v1.5-zs",
|
5 |
+
"Model Size(B)": 7.57,
|
6 |
+
"Data Source": "Self-Reported",
|
7 |
+
"V1-Overall": 60.1,
|
8 |
+
"I-CLS": 56.1,
|
9 |
+
"I-QA": 55.3,
|
10 |
+
"I-RET": 63.9,
|
11 |
+
"I-VG": 70.8
|
12 |
+
}
|
13 |
+
]
|
Archive/LLaVE.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"Model": "LLaVE-7B",
|
4 |
+
"URL": "https://huggingface.co/zhibinlan/LLaVE-7B",
|
5 |
+
"Model Size(B)": 8.03B,
|
6 |
+
"Data Source": "Self-Reported",
|
7 |
+
"Overall": 70.3,
|
8 |
+
"Classification": 65.7,
|
9 |
+
"VQA": 65.4,
|
10 |
+
"Retrieval": 70.9,
|
11 |
+
"Grounding": 91.9
|
12 |
+
},
|
13 |
+
{
|
14 |
+
"Model": "LLaVE-2B",
|
15 |
+
"URL": "https://huggingface.co/zhibinlan/LLaVE-2B",
|
16 |
+
"Model Size(B)": 1.95B,
|
17 |
+
"Data Source": "Self-Reported",
|
18 |
+
"Overall": 65.2,
|
19 |
+
"Classification": 62.1,
|
20 |
+
"VQA": 60.2,
|
21 |
+
"Retrieval": 65.2,
|
22 |
+
"Grounding": 84.9
|
23 |
+
},
|
24 |
+
{
|
25 |
+
"Model": "LLaVE-0.5B",
|
26 |
+
"URL": "https://huggingface.co/zhibinlan/LLaVE-0.5B",
|
27 |
+
"Model Size(B)": 894M,
|
28 |
+
"Data Source": "Self-Reported",
|
29 |
+
"Overall": 59.1,
|
30 |
+
"Classification": 57.4,
|
31 |
+
"VQA": 50.3,
|
32 |
+
"Retrieval": 59.8,
|
33 |
+
"Grounding": 82.9
|
34 |
+
},
|
35 |
+
]
|
Archive/UniME_results.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"Model": "UniME(LLaVA-OneVision-7B-LoRA-Res336)",
|
4 |
+
"URL": "https://huggingface.co/DeepGlint-AI/UniME-LLaVA-OneVision-7B",
|
5 |
+
"Model Size(B)": 8.03,
|
6 |
+
"Data Source": "Self-Reported",
|
7 |
+
"Overall": 70.7,
|
8 |
+
"Classification": 66.8,
|
9 |
+
"VQA": 66.6,
|
10 |
+
"Retrieval": 70.5,
|
11 |
+
"Grounding": 90.9
|
12 |
+
},
|
13 |
+
{
|
14 |
+
"Model": "UniME(LLaVA-1.6-7B-LoRA-LowRes)",
|
15 |
+
"URL": "https://huggingface.co/DeepGlint-AI/UniME-LLaVA-1.6-7B",
|
16 |
+
"Model Size(B)": 7.57,
|
17 |
+
"Data Source": "Self-Reported",
|
18 |
+
"Overall": 66.6,
|
19 |
+
"Classification": 60.6,
|
20 |
+
"VQA": 52.9,
|
21 |
+
"Retrieval": 67.9,
|
22 |
+
"Grounding": 85.1
|
23 |
+
},
|
24 |
+
{
|
25 |
+
"Model": "UniME(Phi-3.5-V-LoRA)",
|
26 |
+
"URL": "https://huggingface.co/DeepGlint-AI/UniME-Phi3.5-V-4.2B",
|
27 |
+
"Model Size(B)": 4.2,
|
28 |
+
"Data Source": "Self-Reported",
|
29 |
+
"Overall": 64.2,
|
30 |
+
"Classification": 54.8,
|
31 |
+
"VQA": 55.9,
|
32 |
+
"Retrieval": 64.5,
|
33 |
+
"Grounding": 81.8
|
34 |
+
}
|
35 |
+
]
|
Archive/meta-score-BGE-VL-v1.5-mmeb.json
ADDED
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"Classification": {
|
3 |
+
"ImageNet-1K": 0.78,
|
4 |
+
"N24News": 0.812,
|
5 |
+
"HatefulMemes": 0.616,
|
6 |
+
"VOC2007": 0.928,
|
7 |
+
"SUN397": 0.836,
|
8 |
+
"Place365": 0.468,
|
9 |
+
"ImageNet-A": 0.449,
|
10 |
+
"ImageNet-R": 0.771,
|
11 |
+
"ObjectNet": 0.54,
|
12 |
+
"Country211": 0.172,
|
13 |
+
"avg": 0.6372
|
14 |
+
},
|
15 |
+
"VQA": {
|
16 |
+
"OK-VQA": 0.749,
|
17 |
+
"A-OKVQA": 0.599,
|
18 |
+
"DocVQA": 0.854,
|
19 |
+
"InfographicsVQA": 0.47,
|
20 |
+
"ChartQA": 0.487,
|
21 |
+
"Visual7W": 0.613,
|
22 |
+
"ScienceQA": 0.679,
|
23 |
+
"VizWiz": 0.493,
|
24 |
+
"GQA": 0.728,
|
25 |
+
"TextVQA": 0.814,
|
26 |
+
"avg": 0.6486
|
27 |
+
},
|
28 |
+
"Retrieval": {
|
29 |
+
"VisDial": 0.846,
|
30 |
+
"CIRR": 0.635,
|
31 |
+
"VisualNews_t2i": 0.787,
|
32 |
+
"VisualNews_i2t": 0.816,
|
33 |
+
"MSCOCO_t2i": 0.793,
|
34 |
+
"MSCOCO_i2t": 0.755,
|
35 |
+
"NIGHTS": 0.698,
|
36 |
+
"WebQA": 0.904,
|
37 |
+
"OVEN": 0.576,
|
38 |
+
"FashionIQ": 0.269,
|
39 |
+
"EDIS": 0.904,
|
40 |
+
"Wiki-SS-NQ": 0.676,
|
41 |
+
"avg": 0.7215833333333332
|
42 |
+
},
|
43 |
+
"Visual Grounding": {
|
44 |
+
"MSCOCO": 0.824,
|
45 |
+
"Visual7W-Pointing": 0.912,
|
46 |
+
"RefCOCO": 0.911,
|
47 |
+
"RefCOCO-Matching": 0.815,
|
48 |
+
"avg": 0.8655
|
49 |
+
},
|
50 |
+
"avg": 0.693861111111111
|
51 |
+
}
|
Archive/meta-score-BGE-VL-v1.5-zs.json
ADDED
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"Classification": {
|
3 |
+
"ImageNet-1K": 0.636,
|
4 |
+
"N24News": 0.511,
|
5 |
+
"HatefulMemes": 0.51,
|
6 |
+
"VOC2007": 0.78,
|
7 |
+
"SUN397": 0.722,
|
8 |
+
"Place365": 0.431,
|
9 |
+
"ImageNet-A": 0.488,
|
10 |
+
"ImageNet-R": 0.758,
|
11 |
+
"ObjectNet": 0.618,
|
12 |
+
"Country211": 0.152,
|
13 |
+
"avg": 0.5606000000000001
|
14 |
+
},
|
15 |
+
"VQA": {
|
16 |
+
"OK-VQA": 0.642,
|
17 |
+
"A-OKVQA": 0.53,
|
18 |
+
"DocVQA": 0.688,
|
19 |
+
"InfographicsVQA": 0.319,
|
20 |
+
"ChartQA": 0.355,
|
21 |
+
"Visual7W": 0.488,
|
22 |
+
"ScienceQA": 0.462,
|
23 |
+
"VizWiz": 0.515,
|
24 |
+
"GQA": 0.755,
|
25 |
+
"TextVQA": 0.777,
|
26 |
+
"avg": 0.5531
|
27 |
+
},
|
28 |
+
"Retrieval": {
|
29 |
+
"VisDial": 0.701,
|
30 |
+
"CIRR": 0.601,
|
31 |
+
"VisualNews_t2i": 0.63,
|
32 |
+
"VisualNews_i2t": 0.669,
|
33 |
+
"MSCOCO_t2i": 0.719,
|
34 |
+
"MSCOCO_i2t": 0.63,
|
35 |
+
"NIGHTS": 0.598,
|
36 |
+
"WebQA": 0.852,
|
37 |
+
"OVEN": 0.529,
|
38 |
+
"FashionIQ": 0.246,
|
39 |
+
"EDIS": 0.807,
|
40 |
+
"Wiki-SS-NQ": 0.687,
|
41 |
+
"avg": 0.6390833333333334
|
42 |
+
},
|
43 |
+
"Visual Grounding": {
|
44 |
+
"MSCOCO": 0.521,
|
45 |
+
"Visual7W-Pointing": 0.789,
|
46 |
+
"RefCOCO": 0.74,
|
47 |
+
"RefCOCO-Matching": 0.784,
|
48 |
+
"avg": 0.7084999999999999
|
49 |
+
},
|
50 |
+
"avg": 0.6011111111111112
|
51 |
+
}
|
Archive/mmE5_mmeb.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"Model": "mmE5-mllama-11b-instruct",
|
4 |
+
"URL": "https://huggingface.co/intfloat/mmE5-mllama-11b-instruct",
|
5 |
+
"Model Size(B)": 10.6,
|
6 |
+
"Data Source": Self-Reported,
|
7 |
+
"Overall": 69.8,
|
8 |
+
"Classification": 67.6,
|
9 |
+
"VQA": 62.6,
|
10 |
+
"Retrieval": 71.0,
|
11 |
+
"Grounding": 89.6
|
12 |
+
},
|
13 |
+
]
|
Archive/mmeb-qqmm.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
[{"Model": "QQMM-embed", "Model Size(B)": 8.297, "Data Source": "Self-Reported", "Overall": 72.175, "Classification": 70.07, "VQA": 69.52, "Retrieval": 71.175, "Grounding": 87.075}]
|