Commit
•
c04ada2
1
Parent(s):
e5fa517
Upload cuda_training_transformers_fill-mask_google-bert/bert-base-uncased/benchmark.json with huggingface_hub
Browse files
cuda_training_transformers_fill-mask_google-bert/bert-base-uncased/benchmark.json
CHANGED
@@ -3,7 +3,7 @@
|
|
3 |
"name": "cuda_training_transformers_fill-mask_google-bert/bert-base-uncased",
|
4 |
"backend": {
|
5 |
"name": "pytorch",
|
6 |
-
"version": "2.
|
7 |
"_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
|
8 |
"task": "fill-mask",
|
9 |
"library": "transformers",
|
@@ -71,7 +71,7 @@
|
|
71 |
"name": "process",
|
72 |
"_target_": "optimum_benchmark.launchers.process.launcher.ProcessLauncher",
|
73 |
"device_isolation": true,
|
74 |
-
"device_isolation_action": "
|
75 |
"numactl": false,
|
76 |
"numactl_kwargs": {},
|
77 |
"start_method": "spawn"
|
@@ -101,7 +101,7 @@
|
|
101 |
"optimum_benchmark_commit": null,
|
102 |
"transformers_version": "4.44.2",
|
103 |
"transformers_commit": null,
|
104 |
-
"accelerate_version": "0.
|
105 |
"accelerate_commit": null,
|
106 |
"diffusers_version": "0.30.2",
|
107 |
"diffusers_commit": null,
|
@@ -117,33 +117,33 @@
|
|
117 |
"overall": {
|
118 |
"memory": {
|
119 |
"unit": "MB",
|
120 |
-
"max_ram":
|
121 |
-
"max_global_vram":
|
122 |
-
"max_process_vram":
|
123 |
"max_reserved": 2497.708032,
|
124 |
"max_allocated": 2195.345408
|
125 |
},
|
126 |
"latency": {
|
127 |
"unit": "s",
|
128 |
"count": 5,
|
129 |
-
"total": 0.
|
130 |
-
"mean": 0.
|
131 |
-
"stdev": 0.
|
132 |
-
"p50": 0.
|
133 |
-
"p90": 0.
|
134 |
-
"p95": 0.
|
135 |
-
"p99": 0.
|
136 |
"values": [
|
137 |
-
0.
|
138 |
-
0.
|
139 |
-
0.
|
140 |
-
0.
|
141 |
-
0.
|
142 |
]
|
143 |
},
|
144 |
"throughput": {
|
145 |
"unit": "samples/s",
|
146 |
-
"value":
|
147 |
},
|
148 |
"energy": null,
|
149 |
"efficiency": null
|
@@ -151,30 +151,30 @@
|
|
151 |
"warmup": {
|
152 |
"memory": {
|
153 |
"unit": "MB",
|
154 |
-
"max_ram":
|
155 |
-
"max_global_vram":
|
156 |
-
"max_process_vram":
|
157 |
"max_reserved": 2497.708032,
|
158 |
"max_allocated": 2195.345408
|
159 |
},
|
160 |
"latency": {
|
161 |
"unit": "s",
|
162 |
"count": 2,
|
163 |
-
"total": 0.
|
164 |
-
"mean": 0.
|
165 |
-
"stdev": 0.
|
166 |
-
"p50": 0.
|
167 |
-
"p90": 0.
|
168 |
-
"p95": 0.
|
169 |
-
"p99": 0.
|
170 |
"values": [
|
171 |
-
0.
|
172 |
-
0.
|
173 |
]
|
174 |
},
|
175 |
"throughput": {
|
176 |
"unit": "samples/s",
|
177 |
-
"value": 13.
|
178 |
},
|
179 |
"energy": null,
|
180 |
"efficiency": null
|
@@ -182,31 +182,31 @@
|
|
182 |
"train": {
|
183 |
"memory": {
|
184 |
"unit": "MB",
|
185 |
-
"max_ram":
|
186 |
-
"max_global_vram":
|
187 |
-
"max_process_vram":
|
188 |
"max_reserved": 2497.708032,
|
189 |
"max_allocated": 2195.345408
|
190 |
},
|
191 |
"latency": {
|
192 |
"unit": "s",
|
193 |
"count": 3,
|
194 |
-
"total": 0.
|
195 |
-
"mean": 0.
|
196 |
-
"stdev":
|
197 |
-
"p50": 0.
|
198 |
-
"p90": 0.
|
199 |
-
"p95": 0.
|
200 |
-
"p99": 0.
|
201 |
"values": [
|
202 |
-
0.
|
203 |
-
0.
|
204 |
-
0.
|
205 |
]
|
206 |
},
|
207 |
"throughput": {
|
208 |
"unit": "samples/s",
|
209 |
-
"value":
|
210 |
},
|
211 |
"energy": null,
|
212 |
"efficiency": null
|
|
|
3 |
"name": "cuda_training_transformers_fill-mask_google-bert/bert-base-uncased",
|
4 |
"backend": {
|
5 |
"name": "pytorch",
|
6 |
+
"version": "2.3.1+rocm5.7",
|
7 |
"_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
|
8 |
"task": "fill-mask",
|
9 |
"library": "transformers",
|
|
|
71 |
"name": "process",
|
72 |
"_target_": "optimum_benchmark.launchers.process.launcher.ProcessLauncher",
|
73 |
"device_isolation": true,
|
74 |
+
"device_isolation_action": "warn",
|
75 |
"numactl": false,
|
76 |
"numactl_kwargs": {},
|
77 |
"start_method": "spawn"
|
|
|
101 |
"optimum_benchmark_commit": null,
|
102 |
"transformers_version": "4.44.2",
|
103 |
"transformers_commit": null,
|
104 |
+
"accelerate_version": "0.34.0",
|
105 |
"accelerate_commit": null,
|
106 |
"diffusers_version": "0.30.2",
|
107 |
"diffusers_commit": null,
|
|
|
117 |
"overall": {
|
118 |
"memory": {
|
119 |
"unit": "MB",
|
120 |
+
"max_ram": 1271.894016,
|
121 |
+
"max_global_vram": 68702.69952,
|
122 |
+
"max_process_vram": 283400.167424,
|
123 |
"max_reserved": 2497.708032,
|
124 |
"max_allocated": 2195.345408
|
125 |
},
|
126 |
"latency": {
|
127 |
"unit": "s",
|
128 |
"count": 5,
|
129 |
+
"total": 0.7092256927490235,
|
130 |
+
"mean": 0.14184513854980468,
|
131 |
+
"stdev": 0.2020748678933304,
|
132 |
+
"p50": 0.04065860366821289,
|
133 |
+
"p90": 0.3441763854980469,
|
134 |
+
"p95": 0.4450853546142577,
|
135 |
+
"p99": 0.5258125299072265,
|
136 |
"values": [
|
137 |
+
0.5459943237304687,
|
138 |
+
0.04144947814941406,
|
139 |
+
0.0405379638671875,
|
140 |
+
0.040585323333740236,
|
141 |
+
0.04065860366821289
|
142 |
]
|
143 |
},
|
144 |
"throughput": {
|
145 |
"unit": "samples/s",
|
146 |
+
"value": 70.49942001705472
|
147 |
},
|
148 |
"energy": null,
|
149 |
"efficiency": null
|
|
|
151 |
"warmup": {
|
152 |
"memory": {
|
153 |
"unit": "MB",
|
154 |
+
"max_ram": 1271.894016,
|
155 |
+
"max_global_vram": 68702.69952,
|
156 |
+
"max_process_vram": 283400.167424,
|
157 |
"max_reserved": 2497.708032,
|
158 |
"max_allocated": 2195.345408
|
159 |
},
|
160 |
"latency": {
|
161 |
"unit": "s",
|
162 |
"count": 2,
|
163 |
+
"total": 0.5874438018798828,
|
164 |
+
"mean": 0.2937219009399414,
|
165 |
+
"stdev": 0.2522724227905273,
|
166 |
+
"p50": 0.2937219009399414,
|
167 |
+
"p90": 0.4955398391723633,
|
168 |
+
"p95": 0.520767081451416,
|
169 |
+
"p99": 0.5409488752746582,
|
170 |
"values": [
|
171 |
+
0.5459943237304687,
|
172 |
+
0.04144947814941406
|
173 |
]
|
174 |
},
|
175 |
"throughput": {
|
176 |
"unit": "samples/s",
|
177 |
+
"value": 13.618323956094434
|
178 |
},
|
179 |
"energy": null,
|
180 |
"efficiency": null
|
|
|
182 |
"train": {
|
183 |
"memory": {
|
184 |
"unit": "MB",
|
185 |
+
"max_ram": 1271.894016,
|
186 |
+
"max_global_vram": 68702.69952,
|
187 |
+
"max_process_vram": 283400.167424,
|
188 |
"max_reserved": 2497.708032,
|
189 |
"max_allocated": 2195.345408
|
190 |
},
|
191 |
"latency": {
|
192 |
"unit": "s",
|
193 |
"count": 3,
|
194 |
+
"total": 0.12178189086914062,
|
195 |
+
"mean": 0.040593963623046875,
|
196 |
+
"stdev": 4.9628495492934084e-05,
|
197 |
+
"p50": 0.040585323333740236,
|
198 |
+
"p90": 0.040643947601318356,
|
199 |
+
"p95": 0.04065127563476562,
|
200 |
+
"p99": 0.040657138061523435,
|
201 |
"values": [
|
202 |
+
0.0405379638671875,
|
203 |
+
0.040585323333740236,
|
204 |
+
0.04065860366821289
|
205 |
]
|
206 |
},
|
207 |
"throughput": {
|
208 |
"unit": "samples/s",
|
209 |
+
"value": 147.8052267996208
|
210 |
},
|
211 |
"energy": null,
|
212 |
"efficiency": null
|