Upload cuda_inference_diffusers_stable-diffusion_CompVis/stable-diffusion-v1-4/benchmark.json with huggingface_hub
Browse files
cuda_inference_diffusers_stable-diffusion_CompVis/stable-diffusion-v1-4/benchmark.json
CHANGED
|
@@ -6,19 +6,17 @@
|
|
| 6 |
"version": "2.2.2",
|
| 7 |
"_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
|
| 8 |
"task": "stable-diffusion",
|
| 9 |
-
"model": "CompVis/stable-diffusion-v1-4",
|
| 10 |
"library": "diffusers",
|
|
|
|
|
|
|
| 11 |
"device": "cuda",
|
| 12 |
"device_ids": "0",
|
| 13 |
"seed": 42,
|
| 14 |
"inter_op_num_threads": null,
|
| 15 |
"intra_op_num_threads": null,
|
| 16 |
-
"
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
"local_files_only": false,
|
| 20 |
-
"trust_remote_code": false
|
| 21 |
-
},
|
| 22 |
"no_weights": false,
|
| 23 |
"device_map": null,
|
| 24 |
"torch_dtype": null,
|
|
@@ -104,41 +102,41 @@
|
|
| 104 |
"call": {
|
| 105 |
"memory": {
|
| 106 |
"unit": "MB",
|
| 107 |
-
"max_ram":
|
| 108 |
-
"max_global_vram":
|
| 109 |
"max_process_vram": 0.0,
|
| 110 |
-
"max_reserved":
|
| 111 |
-
"max_allocated":
|
| 112 |
},
|
| 113 |
"latency": {
|
| 114 |
"unit": "s",
|
| 115 |
"count": 2,
|
| 116 |
-
"total": 1.
|
| 117 |
-
"mean": 0.
|
| 118 |
-
"stdev": 0.
|
| 119 |
-
"p50": 0.
|
| 120 |
-
"p90": 0.
|
| 121 |
-
"p95": 0.
|
| 122 |
-
"p99": 0.
|
| 123 |
"values": [
|
| 124 |
-
0.
|
| 125 |
-
0.
|
| 126 |
]
|
| 127 |
},
|
| 128 |
"throughput": {
|
| 129 |
"unit": "images/s",
|
| 130 |
-
"value": 1.
|
| 131 |
},
|
| 132 |
"energy": {
|
| 133 |
"unit": "kWh",
|
| 134 |
-
"cpu": 7.
|
| 135 |
-
"ram": 4.
|
| 136 |
-
"gpu":
|
| 137 |
-
"total": 5.
|
| 138 |
},
|
| 139 |
"efficiency": {
|
| 140 |
"unit": "images/kWh",
|
| 141 |
-
"value":
|
| 142 |
}
|
| 143 |
}
|
| 144 |
}
|
|
|
|
| 6 |
"version": "2.2.2",
|
| 7 |
"_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
|
| 8 |
"task": "stable-diffusion",
|
|
|
|
| 9 |
"library": "diffusers",
|
| 10 |
+
"model": "CompVis/stable-diffusion-v1-4",
|
| 11 |
+
"processor": "CompVis/stable-diffusion-v1-4",
|
| 12 |
"device": "cuda",
|
| 13 |
"device_ids": "0",
|
| 14 |
"seed": 42,
|
| 15 |
"inter_op_num_threads": null,
|
| 16 |
"intra_op_num_threads": null,
|
| 17 |
+
"model_kwargs": {},
|
| 18 |
+
"processor_kwargs": {},
|
| 19 |
+
"hub_kwargs": {},
|
|
|
|
|
|
|
|
|
|
| 20 |
"no_weights": false,
|
| 21 |
"device_map": null,
|
| 22 |
"torch_dtype": null,
|
|
|
|
| 102 |
"call": {
|
| 103 |
"memory": {
|
| 104 |
"unit": "MB",
|
| 105 |
+
"max_ram": 1315.57376,
|
| 106 |
+
"max_global_vram": 8143.765504,
|
| 107 |
"max_process_vram": 0.0,
|
| 108 |
+
"max_reserved": 7488.929792,
|
| 109 |
+
"max_allocated": 6529.611776
|
| 110 |
},
|
| 111 |
"latency": {
|
| 112 |
"unit": "s",
|
| 113 |
"count": 2,
|
| 114 |
+
"total": 1.2925891723632814,
|
| 115 |
+
"mean": 0.6462945861816407,
|
| 116 |
+
"stdev": 0.0018083190917968506,
|
| 117 |
+
"p50": 0.6462945861816407,
|
| 118 |
+
"p90": 0.6477412414550782,
|
| 119 |
+
"p95": 0.6479220733642578,
|
| 120 |
+
"p99": 0.6480667388916016,
|
| 121 |
"values": [
|
| 122 |
+
0.6481029052734375,
|
| 123 |
+
0.6444862670898438
|
| 124 |
]
|
| 125 |
},
|
| 126 |
"throughput": {
|
| 127 |
"unit": "images/s",
|
| 128 |
+
"value": 1.5472820311060915
|
| 129 |
},
|
| 130 |
"energy": {
|
| 131 |
"unit": "kWh",
|
| 132 |
+
"cpu": 7.633690453237957e-06,
|
| 133 |
+
"ram": 4.175484979896282e-06,
|
| 134 |
+
"gpu": 4.125433855899953e-05,
|
| 135 |
+
"total": 5.306351399213377e-05
|
| 136 |
},
|
| 137 |
"efficiency": {
|
| 138 |
"unit": "images/kWh",
|
| 139 |
+
"value": 18845.340701489196
|
| 140 |
}
|
| 141 |
}
|
| 142 |
}
|