Commit
•
bda859a
1
Parent(s):
65518ca
Upload cuda_training_transformers_text-classification_FacebookAI/roberta-base/benchmark.json with huggingface_hub
Browse files
cuda_training_transformers_text-classification_FacebookAI/roberta-base/benchmark.json
CHANGED
@@ -11,7 +11,7 @@
|
|
11 |
"model": "FacebookAI/roberta-base",
|
12 |
"processor": "FacebookAI/roberta-base",
|
13 |
"device": "cuda",
|
14 |
-
"device_ids": "
|
15 |
"seed": 42,
|
16 |
"inter_op_num_threads": null,
|
17 |
"intra_op_num_threads": null,
|
@@ -119,34 +119,34 @@
|
|
119 |
"overall": {
|
120 |
"memory": {
|
121 |
"unit": "MB",
|
122 |
-
"max_ram":
|
123 |
"max_global_vram": 68702.69952,
|
124 |
-
"max_process_vram":
|
125 |
"max_reserved": 2707.423232,
|
126 |
"max_allocated": 2497.900032
|
127 |
},
|
128 |
"latency": {
|
129 |
"unit": "s",
|
130 |
"values": [
|
131 |
-
0.
|
132 |
-
0.
|
133 |
-
0.
|
134 |
-
0.
|
135 |
-
0.
|
136 |
],
|
137 |
"count": 5,
|
138 |
-
"total": 0.
|
139 |
-
"mean": 0.
|
140 |
-
"p50": 0.
|
141 |
-
"p90": 0.
|
142 |
-
"p95": 0.
|
143 |
-
"p99": 0.
|
144 |
-
"stdev": 0.
|
145 |
-
"stdev_":
|
146 |
},
|
147 |
"throughput": {
|
148 |
"unit": "samples/s",
|
149 |
-
"value":
|
150 |
},
|
151 |
"energy": null,
|
152 |
"efficiency": null
|
@@ -154,31 +154,31 @@
|
|
154 |
"warmup": {
|
155 |
"memory": {
|
156 |
"unit": "MB",
|
157 |
-
"max_ram":
|
158 |
"max_global_vram": 68702.69952,
|
159 |
-
"max_process_vram":
|
160 |
"max_reserved": 2707.423232,
|
161 |
"max_allocated": 2497.900032
|
162 |
},
|
163 |
"latency": {
|
164 |
"unit": "s",
|
165 |
"values": [
|
166 |
-
0.
|
167 |
-
0.
|
168 |
],
|
169 |
"count": 2,
|
170 |
-
"total": 0.
|
171 |
-
"mean": 0.
|
172 |
-
"p50": 0.
|
173 |
-
"p90": 0.
|
174 |
-
"p95": 0.
|
175 |
-
"p99": 0.
|
176 |
-
"stdev": 0.
|
177 |
-
"stdev_":
|
178 |
},
|
179 |
"throughput": {
|
180 |
"unit": "samples/s",
|
181 |
-
"value": 11.
|
182 |
},
|
183 |
"energy": null,
|
184 |
"efficiency": null
|
@@ -186,32 +186,32 @@
|
|
186 |
"train": {
|
187 |
"memory": {
|
188 |
"unit": "MB",
|
189 |
-
"max_ram":
|
190 |
"max_global_vram": 68702.69952,
|
191 |
-
"max_process_vram":
|
192 |
"max_reserved": 2707.423232,
|
193 |
"max_allocated": 2497.900032
|
194 |
},
|
195 |
"latency": {
|
196 |
"unit": "s",
|
197 |
"values": [
|
198 |
-
0.
|
199 |
-
0.
|
200 |
-
0.
|
201 |
],
|
202 |
"count": 3,
|
203 |
-
"total": 0.
|
204 |
-
"mean": 0.
|
205 |
-
"p50": 0.
|
206 |
-
"p90": 0.
|
207 |
-
"p95": 0.
|
208 |
-
"p99": 0.
|
209 |
-
"stdev": 0.
|
210 |
-
"stdev_":
|
211 |
},
|
212 |
"throughput": {
|
213 |
"unit": "samples/s",
|
214 |
-
"value":
|
215 |
},
|
216 |
"energy": null,
|
217 |
"efficiency": null
|
|
|
11 |
"model": "FacebookAI/roberta-base",
|
12 |
"processor": "FacebookAI/roberta-base",
|
13 |
"device": "cuda",
|
14 |
+
"device_ids": "6",
|
15 |
"seed": 42,
|
16 |
"inter_op_num_threads": null,
|
17 |
"intra_op_num_threads": null,
|
|
|
119 |
"overall": {
|
120 |
"memory": {
|
121 |
"unit": "MB",
|
122 |
+
"max_ram": 1326.923776,
|
123 |
"max_global_vram": 68702.69952,
|
124 |
+
"max_process_vram": 330356.645888,
|
125 |
"max_reserved": 2707.423232,
|
126 |
"max_allocated": 2497.900032
|
127 |
},
|
128 |
"latency": {
|
129 |
"unit": "s",
|
130 |
"values": [
|
131 |
+
0.6141828002929688,
|
132 |
+
0.05372709655761719,
|
133 |
+
0.0780350341796875,
|
134 |
+
0.04707426834106445,
|
135 |
+
0.04599682235717773
|
136 |
],
|
137 |
"count": 5,
|
138 |
+
"total": 0.8390160217285156,
|
139 |
+
"mean": 0.16780320434570312,
|
140 |
+
"p50": 0.05372709655761719,
|
141 |
+
"p90": 0.3997236938476563,
|
142 |
+
"p95": 0.5069532470703124,
|
143 |
+
"p99": 0.5927368896484375,
|
144 |
+
"stdev": 0.22348990899055132,
|
145 |
+
"stdev_": 133.18572184719676
|
146 |
},
|
147 |
"throughput": {
|
148 |
"unit": "samples/s",
|
149 |
+
"value": 59.593617648673145
|
150 |
},
|
151 |
"energy": null,
|
152 |
"efficiency": null
|
|
|
154 |
"warmup": {
|
155 |
"memory": {
|
156 |
"unit": "MB",
|
157 |
+
"max_ram": 1326.923776,
|
158 |
"max_global_vram": 68702.69952,
|
159 |
+
"max_process_vram": 330356.645888,
|
160 |
"max_reserved": 2707.423232,
|
161 |
"max_allocated": 2497.900032
|
162 |
},
|
163 |
"latency": {
|
164 |
"unit": "s",
|
165 |
"values": [
|
166 |
+
0.6141828002929688,
|
167 |
+
0.05372709655761719
|
168 |
],
|
169 |
"count": 2,
|
170 |
+
"total": 0.667909896850586,
|
171 |
+
"mean": 0.333954948425293,
|
172 |
+
"p50": 0.333954948425293,
|
173 |
+
"p90": 0.5581372299194336,
|
174 |
+
"p95": 0.5861600151062012,
|
175 |
+
"p99": 0.6085782432556153,
|
176 |
+
"stdev": 0.2802278518676758,
|
177 |
+
"stdev_": 83.91187290053402
|
178 |
},
|
179 |
"throughput": {
|
180 |
"unit": "samples/s",
|
181 |
+
"value": 11.977663510785844
|
182 |
},
|
183 |
"energy": null,
|
184 |
"efficiency": null
|
|
|
186 |
"train": {
|
187 |
"memory": {
|
188 |
"unit": "MB",
|
189 |
+
"max_ram": 1326.923776,
|
190 |
"max_global_vram": 68702.69952,
|
191 |
+
"max_process_vram": 330356.645888,
|
192 |
"max_reserved": 2707.423232,
|
193 |
"max_allocated": 2497.900032
|
194 |
},
|
195 |
"latency": {
|
196 |
"unit": "s",
|
197 |
"values": [
|
198 |
+
0.0780350341796875,
|
199 |
+
0.04707426834106445,
|
200 |
+
0.04599682235717773
|
201 |
],
|
202 |
"count": 3,
|
203 |
+
"total": 0.17110612487792967,
|
204 |
+
"mean": 0.05703537495930989,
|
205 |
+
"p50": 0.04707426834106445,
|
206 |
+
"p90": 0.0718428810119629,
|
207 |
+
"p95": 0.0749389575958252,
|
208 |
+
"p99": 0.07741581886291504,
|
209 |
+
"stdev": 0.01485551498021394,
|
210 |
+
"stdev_": 26.046142399891547
|
211 |
},
|
212 |
"throughput": {
|
213 |
"unit": "samples/s",
|
214 |
+
"value": 105.19787069481902
|
215 |
},
|
216 |
"energy": null,
|
217 |
"efficiency": null
|