gilalnauman commited on
Commit
27bb1c0
·
verified ·
1 Parent(s): 29fcacd

Update leaderboard.csv

Browse files
Files changed (1) hide show
  1. leaderboard.csv +11 -6
leaderboard.csv CHANGED
@@ -1,9 +1,10 @@
1
- Method Params (B) Language Model Vision Model Evaluation Date Avg score Pubmed Tiny Pubmed All Socialpath Tiny Socialpath All Education Content Tiny Education Content All
2
  LLaVA-v1.5-7B 7.2 Vicuna-v1.5-7B CLIP ViT-L/14 12/20/2024 37.38833333 43.77 39.6 36.24 38.01 31.37 35.34
3
  LLaVA-v1.5-13B 13.4 Vicuna-v1.5-13B CLIP ViT-L/14 12/20/2024 39.26666667 45.9 37.32 39.73 37.45 39.6 35.6
4
  LLaVA-Next-Vicuna-7B 7.1 Vicuna-v1.5-7B CLIP ViT-L/14 12/20/2024 37.19 43.06 36.34 39.3 34.63 35.29 34.52
5
  LLaVA-Next-Vicuna-13B 13.4 Vicuna-v1.5-13B CLIP ViT-L/14 12/20/2024 37.71666667 40.93 37.71 42.36 36.45 34.9 33.95
6
- LLaVA-Next-Mistral-7B 7.6 Mistral-7B CLIP ViT-L/14 12/20/2024 38.91333333 47.33 36.83 39.3 34.85 40.39 34.78
 
7
  ShareGPT4V-7B 7.2 Vicuna-v1.5-7B CLIP ViT-L/14 12/20/2024 34.92166667 41.99 36.38 33.62 33.57 30.59 33.38
8
  ShareGPT4V-13B 13.4 Vicuna-v1.5-13B CLIP ViT-L/14 12/20/2024 34.10833333 40.57 35.04 34.06 31.47 32.55 30.96
9
  LLaVA-Next-Llama3 8 Llama-3-8B-Instruct CLIP ViT-L/14 12/20/2024 41.47666667 47.69 40.45 43.67 38.01 42.35 36.69
@@ -30,7 +31,8 @@ MolmoE-1B 7.2 OLMoE-1B-7B-0924 CLIP ViT-L/14 12/23/2024 37.4 39.5 36.92 40.61 37
30
  Molmo-7B-D 8 Qwen2-7B CLIP ViT-L/14 12/23/2024 42.31833333 47.33 41.32 43.23 41.55 41.17 39.31
31
  Molmo-7B-O 8 Qwen2-7B CLIP ViT-L/14 12/23/2024 38.88833333 46.26 38.13 39.3 36.5 37.64 35.5
32
  Phi-3-Vision 4.2 Phi-3 CLIP ViT-L/14 12/23/2024 39.04 44.83 38.55 38.86 38.83 36.07 37.1
33
- Phi-3.5-Vision 4 Phi-3.5 CLIP ViT-L/14 12/23/2024 43.77333333 49.11 44.16 44.97 40.55 43.92 39.93
 
34
  hareCaptioner 8 InternLM-7B EVA-G 12/25/2024 34.63666667 38.79 33.76 37.11 32.24 34.5 31.42
35
  InternLM-XComposer2 7 InternLM2-7B CLIP ViT-L/14 12/25/2024 37.74666667 40.56 36.6 36.68 35.34 41.96 35.34
36
  InternLM-XComposer2-1.8B 2 InternLM2-1.8B CLIP ViT-L/14 12/25/2024 31.73166667 35.94 32.65 30.13 31.63 30.58 29.46
@@ -44,8 +46,10 @@ InternVL2-1B 1 Qwen2-0.5B InternViT-300M 12/26/2024 37.80166667 42.35 36.15 37.5
44
  InternVL2-2B 2 InternLM2-1.8B InternViT-300M 12/26/2024 39.57 41.99 37.45 43.23 36.18 42.35 36.22
45
  InternVL2-4B 4 Phi-3 InternViT-300M 12/26/2024 49.23666667 57.65 45.47 55.46 42.94 50.2 43.7
46
  InternVL2-8B 8 InternLM2.5-7B InternViT-300M 12/26/2024 49.9 58.01 45.33 57.64 43.99 51.76 42.67
47
- InternVL2-26B 26 InternLM2-20B InternViT-6B 12/27/2024 47.61333333 53.02 44.69 53.28 41.72 50.2 42.77
48
- InternVL2-8B-MPO 8 InternLM2.5-7B InternViT-300M 12/27/2024 50.43333333 55.52 48.04 57.64 45.59 52.16 43.65
 
 
49
  InternVL2-8B-MPO-CoT 8 InternLM2.5-7B InternViT-300M 12/27/2024 3.188333333 3.2 2.9 3.93 2.55 3.14 3.41
50
  InternVL2.5-1B 1 Qwen-2.5-0.5B InternViT-300M-v2.5 12/27/2024 42.19333333 49.46 38.95 46.72 39.44 40 38.59
51
  InternVL2.5-2B 2 InternLM2.5-1.8B InternViT-300M-v2.5 12/27/2024 45.77 52.31 42.47 50.21 41.82 47.05 40.76
@@ -53,7 +57,8 @@ InternVL2.5-4B 4 Qwen-2.5-3B InternViT-300M-v2.5 12/27/2024 53.23833333 62.98 49
53
  InternVL2.5-8B 8 InternLM2.5-7B InternViT-300M-v2.5 12/27/2024 54.39166667 60.85 50.03 60.26 48.25 60.78 46.18
54
  InternVL2.5-26B 26 InternLM2.5-20B InternViT-6B-v2.5 12/27/2024 54.40666667 60.14 52.44 61.57 46.98 57.64 47.67
55
  InternVL2.5-38B 38 Qwen-2.5-32B InternViT-6B-v2.5 12/27/2024 62.72333333 65.83 61.4 62.88 57.28 69.41 59.54
56
- MMAlaya 7.8 Alaya-7B-Chat EVA-G 12/27/2024 29.92 29.18 28.74 34.06 30.24 29.8 27.5
 
57
  Llama-3.2-11B-Vision-Instruct 11 Llama-3.1-8B 12/27/2024 24.07666667 23.48 23.85 24.45 22.32 27.45 22.91
58
  LLaVA-CoT 11 12/27/2024 9.155 10.67 7.36 11.35 7.75 10.58 7.22
59
  Ovis1.5-Llama3-8B 8 Llama-3-8B-Instruct SigLIP-400M 12/31/2024 52.98833333 61.56 48.72 59.38 49.36 52.94 45.97
 
1
+ Method Params (B) Language Model Vision Model Evaluation Date Avg score Pubmed Tiny Pubmed All Socialpath Tiny Socialpath All Education Content Tiny Education Content All
2
  LLaVA-v1.5-7B 7.2 Vicuna-v1.5-7B CLIP ViT-L/14 12/20/2024 37.38833333 43.77 39.6 36.24 38.01 31.37 35.34
3
  LLaVA-v1.5-13B 13.4 Vicuna-v1.5-13B CLIP ViT-L/14 12/20/2024 39.26666667 45.9 37.32 39.73 37.45 39.6 35.6
4
  LLaVA-Next-Vicuna-7B 7.1 Vicuna-v1.5-7B CLIP ViT-L/14 12/20/2024 37.19 43.06 36.34 39.3 34.63 35.29 34.52
5
  LLaVA-Next-Vicuna-13B 13.4 Vicuna-v1.5-13B CLIP ViT-L/14 12/20/2024 37.71666667 40.93 37.71 42.36 36.45 34.9 33.95
6
+ "
7
+ LLaVA-Next-Mistral-7B" 7.6 Mistral-7B CLIP ViT-L/14 12/20/2024 38.91333333 47.33 36.83 39.3 34.85 40.39 34.78
8
  ShareGPT4V-7B 7.2 Vicuna-v1.5-7B CLIP ViT-L/14 12/20/2024 34.92166667 41.99 36.38 33.62 33.57 30.59 33.38
9
  ShareGPT4V-13B 13.4 Vicuna-v1.5-13B CLIP ViT-L/14 12/20/2024 34.10833333 40.57 35.04 34.06 31.47 32.55 30.96
10
  LLaVA-Next-Llama3 8 Llama-3-8B-Instruct CLIP ViT-L/14 12/20/2024 41.47666667 47.69 40.45 43.67 38.01 42.35 36.69
 
31
  Molmo-7B-D 8 Qwen2-7B CLIP ViT-L/14 12/23/2024 42.31833333 47.33 41.32 43.23 41.55 41.17 39.31
32
  Molmo-7B-O 8 Qwen2-7B CLIP ViT-L/14 12/23/2024 38.88833333 46.26 38.13 39.3 36.5 37.64 35.5
33
  Phi-3-Vision 4.2 Phi-3 CLIP ViT-L/14 12/23/2024 39.04 44.83 38.55 38.86 38.83 36.07 37.1
34
+ "
35
+ Phi-3.5-Vision" 4 Phi-3.5 CLIP ViT-L/14 12/23/2024 43.77333333 49.11 44.16 44.97 40.55 43.92 39.93
36
  hareCaptioner 8 InternLM-7B EVA-G 12/25/2024 34.63666667 38.79 33.76 37.11 32.24 34.5 31.42
37
  InternLM-XComposer2 7 InternLM2-7B CLIP ViT-L/14 12/25/2024 37.74666667 40.56 36.6 36.68 35.34 41.96 35.34
38
  InternLM-XComposer2-1.8B 2 InternLM2-1.8B CLIP ViT-L/14 12/25/2024 31.73166667 35.94 32.65 30.13 31.63 30.58 29.46
 
46
  InternVL2-2B 2 InternLM2-1.8B InternViT-300M 12/26/2024 39.57 41.99 37.45 43.23 36.18 42.35 36.22
47
  InternVL2-4B 4 Phi-3 InternViT-300M 12/26/2024 49.23666667 57.65 45.47 55.46 42.94 50.2 43.7
48
  InternVL2-8B 8 InternLM2.5-7B InternViT-300M 12/26/2024 49.9 58.01 45.33 57.64 43.99 51.76 42.67
49
+ "
50
+ InternVL2-26B" 26 InternLM2-20B InternViT-6B 12/27/2024 47.61333333 53.02 44.69 53.28 41.72 50.2 42.77
51
+ "
52
+ InternVL2-8B-MPO" 8 InternLM2.5-7B InternViT-300M 12/27/2024 50.43333333 55.52 48.04 57.64 45.59 52.16 43.65
53
  InternVL2-8B-MPO-CoT 8 InternLM2.5-7B InternViT-300M 12/27/2024 3.188333333 3.2 2.9 3.93 2.55 3.14 3.41
54
  InternVL2.5-1B 1 Qwen-2.5-0.5B InternViT-300M-v2.5 12/27/2024 42.19333333 49.46 38.95 46.72 39.44 40 38.59
55
  InternVL2.5-2B 2 InternLM2.5-1.8B InternViT-300M-v2.5 12/27/2024 45.77 52.31 42.47 50.21 41.82 47.05 40.76
 
57
  InternVL2.5-8B 8 InternLM2.5-7B InternViT-300M-v2.5 12/27/2024 54.39166667 60.85 50.03 60.26 48.25 60.78 46.18
58
  InternVL2.5-26B 26 InternLM2.5-20B InternViT-6B-v2.5 12/27/2024 54.40666667 60.14 52.44 61.57 46.98 57.64 47.67
59
  InternVL2.5-38B 38 Qwen-2.5-32B InternViT-6B-v2.5 12/27/2024 62.72333333 65.83 61.4 62.88 57.28 69.41 59.54
60
+ "
61
+ MMAlaya" 7.8 Alaya-7B-Chat EVA-G 12/27/2024 29.92 29.18 28.74 34.06 30.24 29.8 27.5
62
  Llama-3.2-11B-Vision-Instruct 11 Llama-3.1-8B 12/27/2024 24.07666667 23.48 23.85 24.45 22.32 27.45 22.91
63
  LLaVA-CoT 11 12/27/2024 9.155 10.67 7.36 11.35 7.75 10.58 7.22
64
  Ovis1.5-Llama3-8B 8 Llama-3-8B-Instruct SigLIP-400M 12/31/2024 52.98833333 61.56 48.72 59.38 49.36 52.94 45.97