Hugging Face's logo

Datasets:
boavizta
/
open_data_boamps

License:
Dataset card Files
xet
Community
13

Upload 3 files

#39
by leokeba - opened
base: refs/heads/main
←
from: refs/pr/39
Discussion Files changed
+333
-0
Files changed (3) hide show
  1. data/report_terra_cognita_vision_training_onPrem_1761070539.json +111 -0
  2. data/report_terra_cognita_vision_training_onPrem_1761128618.json +110 -0
  3. data/report_terra_cognita_vision_training_onPrem_1761174706.json +112 -0
data/report_terra_cognita_vision_training_onPrem_1761070539.json ADDED
@@ -0,0 +1,111 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "header": {
3
+ "licensing": "Creative Commons 4.0",
4
+ "formatVersion": "1.1.0",
5
+ "formatVersionSpecificationUri": "https://raw.githubusercontent.com/Boavizta/BoAmps/main/model/report_schema.json",
6
+ "reportId": "vision-training-1761070539",
7
+ "reportDatetime": "2025-10-21 19:31:42",
8
+ "reportStatus": "final",
9
+ "publisher": {
10
+ "name": "Terra Cognita",
11
+ "division": "Research",
12
+ "projectName": "SEEDS Lab",
13
+ "confidentialityLevel": "public"
14
+ }
15
+ },
16
+ "task": {
17
+ "taskStage": "finetuning",
18
+ "taskFamily": "multiModalTextGeneration",
19
+ "algorithms": [
20
+ {
21
+ "trainingType": "transferLearning",
22
+ "algorithmType": "vlm",
23
+ "algorithmName": "transformer",
24
+ "foundationModelName": "Qwen/Qwen2.5-VL-3B-Instruct",
25
+ "foundationModelUri": "https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct",
26
+ "parametersNumber": 3,
27
+ "framework": "PyTorch",
28
+ "frameworkVersion": "2.9.0",
29
+ "epochsNumber": 3,
30
+ "optimizer": "adamw_torch",
31
+ "quantization": "int4"
32
+ }
33
+ ],
34
+ "dataset": [
35
+ {
36
+ "dataUsage": "input",
37
+ "dataType": "image",
38
+ "source": "public",
39
+ "sourceUri": "https://huggingface.co/datasets/Barth371/cmr-all",
40
+ "fileType": "json",
41
+ "dataFormat": "json",
42
+ "dataSize": 0.5236,
43
+ "volume": 562240572,
44
+ "volumeUnit": "byte",
45
+ "dataQuantity": 663,
46
+ "owner": "Barth371",
47
+ "subset": ["train", "validation"]
48
+ }
49
+ ],
50
+ "taskDescription": "Vision-language model fine-tuning using Qwen/Qwen2.5-VL-3B-Instruct with LoRA (r=16, alpha=16) on dataset: Barth371/cmr-all"
51
+ },
52
+ "measures": [
53
+ {
54
+ "measurementMethod": "codecarbon",
55
+ "version": "2.5.0",
56
+ "cpuTrackingMode": "process",
57
+ "gpuTrackingMode": "nvml",
58
+ "powerConsumption": 0.428352,
59
+ "measurementDuration": 4559.75,
60
+ "measurementDateTime": "2025-10-21 19:31:42",
61
+ "averageUtilizationCpu": 0.0501,
62
+ "averageUtilizationGpu": 0.9142
63
+ }
64
+ ],
65
+ "infrastructure": {
66
+ "infraType": "onPremise",
67
+ "components": [
68
+ {
69
+ "componentType": "gpu",
70
+ "nbComponent": 1,
71
+ "share": 1,
72
+ "componentName": "1 x RTX 3090 Ti",
73
+ "manufacturer": "NVIDIA",
74
+ "series": "RTX 3090 Ti",
75
+ "family": "RTX",
76
+ "memorySize": 23
77
+ },
78
+ {
79
+ "componentType": "cpu",
80
+ "nbComponent": 1,
81
+ "share": 1,
82
+ "componentName": "Core i9-10850K CPU @ 3.60GHz",
83
+ "manufacturer": "Intel",
84
+ "series": "Core i9-10850K CPU @ 3.60GHz",
85
+ "family": "Core i9-10850K"
86
+ },
87
+ {
88
+ "componentType": "ram",
89
+ "nbComponent": 1,
90
+ "share": 1,
91
+ "memorySize": 31
92
+ }
93
+ ]
94
+ },
95
+ "system": {
96
+ "os": "Linux",
97
+ "distributionVersion": "6.1.0-37-amd64",
98
+ "distribution": "Debian GNU/Linux 12 (bookworm)"
99
+ },
100
+ "software": {
101
+ "language": "Python",
102
+ "version": "3.13.3"
103
+ },
104
+ "environment": {
105
+ "country": "FR",
106
+ "location": "",
107
+ "powerSupplierType": "public",
108
+ "powerSourceCarbonIntensity": 56.04
109
+ },
110
+ "quality": "high"
111
+ }
data/report_terra_cognita_vision_training_onPrem_1761128618.json ADDED
@@ -0,0 +1,110 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "header": {
3
+ "licensing": "Creative Commons 4.0",
4
+ "formatVersion": "1.1.0",
5
+ "formatVersionSpecificationUri": "https://raw.githubusercontent.com/Boavizta/BoAmps/main/model/report_schema.json",
6
+ "reportId": "vision-training-1761128618",
7
+ "reportDatetime": "2025-10-22 12:47:11",
8
+ "reportStatus": "final",
9
+ "publisher": {
10
+ "name": "Terra Cognita",
11
+ "division": "Research",
12
+ "projectName": "SEEDS Lab",
13
+ "confidentialityLevel": "public"
14
+ }
15
+ },
16
+ "task": {
17
+ "taskStage": "finetuning",
18
+ "taskFamily": "multiModalTextGeneration",
19
+ "algorithms": [
20
+ {
21
+ "trainingType": "transferLearning",
22
+ "algorithmType": "vlm",
23
+ "algorithmName": "transformer",
24
+ "foundationModelName": "Qwen/Qwen2.5-VL-7B-Instruct",
25
+ "foundationModelUri": "https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct",
26
+ "parametersNumber": 7,
27
+ "framework": "PyTorch",
28
+ "frameworkVersion": "2.9.0",
29
+ "epochsNumber": 3,
30
+ "optimizer": "adamw_8bit",
31
+ "quantization": "int4"
32
+ }
33
+ ],
34
+ "dataset": [
35
+ {
36
+ "dataUsage": "input",
37
+ "dataType": "image",
38
+ "source": "public",
39
+ "subset": ["train", "validation"],
40
+ "sourceUri": "https://huggingface.co/datasets/Barth371/cmr-all",
41
+ "fileType": "json",
42
+ "dataFormat": "json",
43
+ "dataSize": 0.5236,
44
+ "volume": 562240572,
45
+ "volumeUnit": "byte",
46
+ "owner": "Barth371"
47
+ }
48
+ ],
49
+ "taskDescription": "Vision-language model fine-tuning using Qwen/Qwen2.5-VL-7B-Instruct with LoRA (r=16, alpha=16) on dataset: Barth371/cmr-all"
50
+ },
51
+ "measures": [
52
+ {
53
+ "measurementMethod": "codecarbon",
54
+ "version": "2.5.0",
55
+ "cpuTrackingMode": "process",
56
+ "gpuTrackingMode": "nvml",
57
+ "powerConsumption": 0.85665,
58
+ "measurementDuration": 8609.35,
59
+ "measurementDateTime": "2025-10-22 12:47:11",
60
+ "averageUtilizationCpu": 0.05,
61
+ "averageUtilizationGpu": 0.9449
62
+ }
63
+ ],
64
+ "infrastructure": {
65
+ "infraType": "onPremise",
66
+ "components": [
67
+ {
68
+ "componentType": "gpu",
69
+ "nbComponent": 1,
70
+ "share": 1,
71
+ "componentName": "1 x RTX 3090 Ti",
72
+ "manufacturer": "NVIDIA",
73
+ "series": "RTX 3090 Ti",
74
+ "family": "RTX",
75
+ "memorySize": 23
76
+ },
77
+ {
78
+ "componentType": "cpu",
79
+ "nbComponent": 1,
80
+ "share": 1,
81
+ "componentName": "Core i9-10850K CPU @ 3.60GHz",
82
+ "manufacturer": "Intel",
83
+ "series": "Core i9-10850K CPU @ 3.60GHz",
84
+ "family": "Core i9-10850K"
85
+ },
86
+ {
87
+ "componentType": "ram",
88
+ "nbComponent": 1,
89
+ "share": 1,
90
+ "memorySize": 31
91
+ }
92
+ ]
93
+ },
94
+ "system": {
95
+ "os": "Linux",
96
+ "distributionVersion": "6.1.0-37-amd64",
97
+ "distribution": "Debian GNU/Linux 12 (bookworm)"
98
+ },
99
+ "software": {
100
+ "language": "Python",
101
+ "version": "3.13.3"
102
+ },
103
+ "environment": {
104
+ "country": "FR",
105
+ "location": "",
106
+ "powerSupplierType": "public",
107
+ "powerSourceCarbonIntensity": 56.04
108
+ },
109
+ "quality": "high"
110
+ }
data/report_terra_cognita_vision_training_onPrem_1761174706.json ADDED
@@ -0,0 +1,112 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "header": {
3
+ "licensing": "Creative Commons 4.0",
4
+ "formatVersion": "1.1.0",
5
+ "formatVersionSpecificationUri": "https://raw.githubusercontent.com/Boavizta/BoAmps/main/model/report_schema.json",
6
+ "reportId": "vision-training-1761174706",
7
+ "reportDatetime": "2025-10-23 00:19:41",
8
+ "reportStatus": "final",
9
+ "publisher": {
10
+ "name": "Terra Cognita",
11
+ "division": "Research",
12
+ "projectName": "SEEDS Lab",
13
+ "confidentialityLevel": "public"
14
+ }
15
+ },
16
+ "task": {
17
+ "taskStage": "finetuning",
18
+ "taskFamily": "multiModalTextGeneration",
19
+ "algorithms": [
20
+ {
21
+ "trainingType": "transferLearning",
22
+ "algorithmType": "vlm",
23
+ "algorithmName": "transformer",
24
+ "foundationModelName": "Qwen/Qwen2.5-VL-7B-Instruct",
25
+ "foundationModelUri": "https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct",
26
+ "parametersNumber": 7,
27
+ "framework": "PyTorch",
28
+ "frameworkVersion": "2.9.0",
29
+ "epochsNumber": 4,
30
+ "optimizer": "adamw_torch",
31
+ "quantization": "int4"
32
+ }
33
+ ],
34
+ "dataset": [
35
+ {
36
+ "dataUsage": "input",
37
+ "dataType": "image",
38
+ "source": "public",
39
+ "subset": ["train", "validation"],
40
+ "sourceUri": "https://huggingface.co/datasets/Barth371/cmr-blocks",
41
+ "fileType": "json",
42
+ "dataFormat": "json",
43
+ "dataSize": 0.2486,
44
+ "volume": 266976513,
45
+ "volumeUnit": "byte",
46
+ "dataQuantity": 448,
47
+ "items": 448,
48
+ "owner": "Barth371"
49
+ }
50
+ ],
51
+ "taskDescription": "Vision-language model fine-tuning using Qwen/Qwen2.5-VL-7B-Instruct with LoRA (r=16, alpha=16) on dataset: Barth371/cmr-blocks"
52
+ },
53
+ "measures": [
54
+ {
55
+ "measurementMethod": "codecarbon",
56
+ "version": "2.5.0",
57
+ "cpuTrackingMode": "process",
58
+ "gpuTrackingMode": "nvml",
59
+ "powerConsumption": 0.491807,
60
+ "measurementDuration": 4071.58,
61
+ "measurementDateTime": "2025-10-23 00:19:41",
62
+ "averageUtilizationCpu": 0.05,
63
+ "averageUtilizationGpu": 0.9514
64
+ }
65
+ ],
66
+ "infrastructure": {
67
+ "infraType": "onPremise",
68
+ "components": [
69
+ {
70
+ "componentType": "gpu",
71
+ "nbComponent": 1,
72
+ "share": 1,
73
+ "componentName": "1 x RTX 3090 Ti",
74
+ "manufacturer": "NVIDIA",
75
+ "series": "RTX 3090 Ti",
76
+ "family": "RTX",
77
+ "memorySize": 23
78
+ },
79
+ {
80
+ "componentType": "cpu",
81
+ "nbComponent": 1,
82
+ "share": 1,
83
+ "componentName": "Core i9-10850K CPU @ 3.60GHz",
84
+ "manufacturer": "Intel",
85
+ "series": "Core i9-10850K CPU @ 3.60GHz",
86
+ "family": "Core i9-10850K"
87
+ },
88
+ {
89
+ "componentType": "ram",
90
+ "nbComponent": 1,
91
+ "share": 1,
92
+ "memorySize": 31
93
+ }
94
+ ]
95
+ },
96
+ "system": {
97
+ "os": "Linux",
98
+ "distributionVersion": "6.1.0-37-amd64",
99
+ "distribution": "Debian GNU/Linux 12 (bookworm)"
100
+ },
101
+ "software": {
102
+ "language": "Python",
103
+ "version": "3.13.3"
104
+ },
105
+ "environment": {
106
+ "country": "FR",
107
+ "location": "",
108
+ "powerSupplierType": "public",
109
+ "powerSourceCarbonIntensity": 56.04
110
+ },
111
+ "quality": "high"
112
+ }