Upload cuda_training_transformers_multiple-choice_FacebookAI/roberta-base/benchmark.json with huggingface_hub

cuda_training_transformers_multiple-choice_FacebookAI/roberta-base/benchmark.json

@@ -3,7 +3,7 @@
         "name": "cuda_training_transformers_multiple-choice_FacebookAI/roberta-base",
         "backend": {
             "name": "pytorch",
-            "version": "2.4.0+cu124",
+            "version": "2.4.0+cu121",
             "_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
             "task": "multiple-choice",
             "library": "transformers",
@@ -110,7 +110,7 @@
         "overall": {
             "memory": {
                 "unit": "MB",
-                "max_ram": 1299.554304,
+                "max_ram": 1167.527936,
                 "max_global_vram": 3379.03616,
                 "max_process_vram": 0.0,
                 "max_reserved": 2730.491904,
@@ -119,24 +119,24 @@
             "latency": {
                 "unit": "s",
                 "count": 5,
-                "total": 0.5483161697387695,
-                "mean": 0.1096632339477539,
-                "stdev": 0.11634699812622813,
-                "p50": 0.051469310760498044,
-                "p90": 0.2265274429321289,
-                "p95": 0.2844401741027831,
-                "p99": 0.3307703590393066,
+                "total": 0.8808540458679199,
+                "mean": 0.176170809173584,
+                "stdev": 0.2524065846847361,
+                "p50": 0.05005311965942383,
+                "p90": 0.4287940780639649,
+                "p95": 0.5548888298034667,
+                "p99": 0.6557646311950683,
                 "values": [
-                    0.3423529052734375,
-                    0.051469310760498044,
-                    0.052789249420166016,
-                    0.050936832427978515,
-                    0.050767871856689455
+                    0.6809835815429688,
+                    0.05050982284545898,
+                    0.049616897583007816,
+                    0.04969062423706055,
+                    0.05005311965942383
                 ]
             },
             "throughput": {
                 "unit": "samples/s",
-                "value": 91.18826465362339
+                "value": 56.76309285806161
             },
             "energy": null,
             "efficiency": null
@@ -144,7 +144,7 @@
         "warmup": {
             "memory": {
                 "unit": "MB",
-                "max_ram": 1299.554304,
+                "max_ram": 1167.527936,
                 "max_global_vram": 3379.03616,
                 "max_process_vram": 0.0,
                 "max_reserved": 2730.491904,
@@ -153,21 +153,21 @@
             "latency": {
                 "unit": "s",
                 "count": 2,
-                "total": 0.3938222160339355,
-                "mean": 0.19691110801696776,
-                "stdev": 0.14544179725646972,
-                "p50": 0.19691110801696776,
-                "p90": 0.3132645458221435,
-                "p95": 0.3278087255477905,
-                "p99": 0.33944406932830806,
+                "total": 0.7314934043884278,
+                "mean": 0.3657467021942139,
+                "stdev": 0.3152368793487549,
+                "p50": 0.3657467021942139,
+                "p90": 0.6179362056732178,
+                "p95": 0.6494598936080933,
+                "p99": 0.6746788439559936,
                 "values": [
-                    0.3423529052734375,
-                    0.051469310760498044
+                    0.6809835815429688,
+                    0.05050982284545898
                 ]
             },
             "throughput": {
                 "unit": "samples/s",
-                "value": 20.313734660694315
+                "value": 10.93653059891699
             },
             "energy": null,
             "efficiency": null
@@ -175,7 +175,7 @@
         "train": {
             "memory": {
                 "unit": "MB",
-                "max_ram": 1299.554304,
+                "max_ram": 1167.527936,
                 "max_global_vram": 3379.03616,
                 "max_process_vram": 0.0,
                 "max_reserved": 2730.491904,
@@ -184,22 +184,22 @@
             "latency": {
                 "unit": "s",
                 "count": 3,
-                "total": 0.154493953704834,
-                "mean": 0.051497984568278,
-                "stdev": 0.0009156639146738015,
-                "p50": 0.050936832427978515,
-                "p90": 0.05241876602172851,
-                "p95": 0.05260400772094727,
-                "p99": 0.05275220108032227,
+                "total": 0.14936064147949218,
+                "mean": 0.049786880493164064,
+                "stdev": 0.0001906504220075295,
+                "p50": 0.04969062423706055,
+                "p90": 0.04998062057495117,
+                "p95": 0.0500168701171875,
+                "p99": 0.050045869750976565,
                 "values": [
-                    0.052789249420166016,
-                    0.050936832427978515,
-                    0.050767871856689455
+                    0.049616897583007816,
+                    0.04969062423706055,
+                    0.05005311965942383
                 ]
             },
             "throughput": {
                 "unit": "samples/s",
-                "value": 116.50941391783925
+                "value": 120.5136763052231
             },
             "energy": null,
             "efficiency": null