Hugging Face's logo

Datasets:
ben-hit-save
/
data-transfer

Dataset card Files
xet
 Community
ben-hit-save commited on
Commit
d258f65
·
verified ·
1 Parent(s): bcc6027

Upload folder using huggingface_hub

Browse files
Files changed (6) hide show
  1. data_backup/lora_qwen2_5_vl_on_ECD_repro/gen-qwen2.5-vl-7b-instruct-descriptive_val.json +0 -0
  2. data_backup/lora_qwen2_5_vl_on_ECD_repro/gen-qwen2.5-vl-7b-instruct-reasoning_val.json +0 -0
  3. data_backup/lora_qwen2_5_vl_on_ECD_repro/scores-qwen2.5-vl-7b-instruct-descriptive_val.json +0 -0
  4. data_backup/lora_qwen2_5_vl_on_ECD_repro/scores-qwen2.5-vl-7b-instruct-reasoning_val.json +0 -0
  5. data_backup/lora_qwen2_5_vl_on_ECD_repro/stats-qwen2.5-vl-7b-instruct-descriptive_val.json +55 -0
  6. data_backup/lora_qwen2_5_vl_on_ECD_repro/stats-qwen2.5-vl-7b-instruct-reasoning_val.json +38 -0
data_backup/lora_qwen2_5_vl_on_ECD_repro/gen-qwen2.5-vl-7b-instruct-descriptive_val.json ADDED
The diff for this file is too large to render. See raw diff
 
data_backup/lora_qwen2_5_vl_on_ECD_repro/gen-qwen2.5-vl-7b-instruct-reasoning_val.json ADDED
The diff for this file is too large to render. See raw diff
 
data_backup/lora_qwen2_5_vl_on_ECD_repro/scores-qwen2.5-vl-7b-instruct-descriptive_val.json ADDED
The diff for this file is too large to render. See raw diff
 
data_backup/lora_qwen2_5_vl_on_ECD_repro/scores-qwen2.5-vl-7b-instruct-reasoning_val.json ADDED
The diff for this file is too large to render. See raw diff
 
data_backup/lora_qwen2_5_vl_on_ECD_repro/stats-qwen2.5-vl-7b-instruct-descriptive_val.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "Overall Score": 74.1,
3
+ "By Question": {
4
+ "Q1": 86.89,
5
+ "Q2": 77.39,
6
+ "Q3": 66.52,
7
+ "Q4": 82.49,
8
+ "Q5": 71.13,
9
+ "Q6": 67.07,
10
+ "Q7": 61.97,
11
+ "Q8": 67.41,
12
+ "Q9": 65.67,
13
+ "Q10": 84.25,
14
+ "Q11": 60.57,
15
+ "Q12": 74.18,
16
+ "Q13": 75.34,
17
+ "Q14": 92.55,
18
+ "Q15": 96.49,
19
+ "Q16": 66.67,
20
+ "Q17": 16.96,
21
+ "Q18": 92.71,
22
+ "Q19": 90.77
23
+ },
24
+ "By Category": {
25
+ "Information Extraction": 73.49,
26
+ "Enumeration": 81.6,
27
+ "Pattern Recognition": 78.38,
28
+ "Counting": 80.66,
29
+ "Compositionality": 16.96
30
+ },
31
+ "By Subplot": {
32
+ "1 Subplot": 78.95,
33
+ "2-4 Subplots": 74.21,
34
+ "5+ Subplots": 66.0
35
+ },
36
+ "By Subject": {
37
+ "Computer Science": 76.98,
38
+ "Economics": 75.72,
39
+ "Electrical Engineering and Systems Science": 77.31,
40
+ "Mathematics": 74.63,
41
+ "Physics": 65.35,
42
+ "Quantitative Biology": 72.22,
43
+ "Quantitative Finance": 73.49,
44
+ "Statistics": 77.43
45
+ },
46
+ "By Year": {
47
+ "2020": 73.99,
48
+ "2021": 71.93,
49
+ "2022": 74.59,
50
+ "2023": 76.01
51
+ },
52
+ "N_valid": 4000,
53
+ "N_invalid": 0,
54
+ "Question Type": "Descriptive"
55
+ }
data_backup/lora_qwen2_5_vl_on_ECD_repro/stats-qwen2.5-vl-7b-instruct-reasoning_val.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "Overall Score": 40.4,
3
+ "By Answer Type": {
4
+ "Text-in-Chart": 40.0,
5
+ "Text-in-General": 50.51,
6
+ "Number-in-Chart": 47.41,
7
+ "Number-in-General": 29.69
8
+ },
9
+ "By Source": {
10
+ "GPT-Sourced": 44.57,
11
+ "GPT-Inspired": 37.04,
12
+ "Completely Human": 40.33
13
+ },
14
+ "By Subject": {
15
+ "Computer Science": 40.48,
16
+ "Economics": 39.86,
17
+ "Electrical Engineering and Systems Science": 40.34,
18
+ "Mathematics": 45.19,
19
+ "Physics": 48.82,
20
+ "Quantitative Biology": 43.65,
21
+ "Quantitative Finance": 32.76,
22
+ "Statistics": 30.09
23
+ },
24
+ "By Year": {
25
+ "2020": 34.01,
26
+ "2021": 44.06,
27
+ "2022": 40.57,
28
+ "2023": 42.74
29
+ },
30
+ "By Subplot": {
31
+ "1 Subplot": 41.19,
32
+ "2-4 Subplots": 42.86,
33
+ "5+ Subplots": 35.17
34
+ },
35
+ "N_valid": 1000,
36
+ "N_invalid": 0,
37
+ "Question Type": "Reasoning"
38
+ }