Commit
•
7d4dcdf
1
Parent(s):
f7ad884
Upload cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark.json with huggingface_hub
Browse files
cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark.json
CHANGED
@@ -3,7 +3,7 @@
|
|
3 |
"name": "cpu_inference_transformers_text-generation_openai-community/gpt2",
|
4 |
"backend": {
|
5 |
"name": "pytorch",
|
6 |
-
"version": "2.
|
7 |
"_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
|
8 |
"task": "text-generation",
|
9 |
"library": "transformers",
|
@@ -45,7 +45,6 @@
|
|
45 |
"warmup_runs": 1,
|
46 |
"input_shapes": {
|
47 |
"batch_size": 1,
|
48 |
-
"num_choices": 2,
|
49 |
"sequence_length": 2
|
50 |
},
|
51 |
"new_tokens": null,
|
@@ -73,23 +72,23 @@
|
|
73 |
"environment": {
|
74 |
"cpu": " AMD EPYC 7763 64-Core Processor",
|
75 |
"cpu_count": 4,
|
76 |
-
"cpu_ram_mb":
|
77 |
"system": "Linux",
|
78 |
"machine": "x86_64",
|
79 |
-
"platform": "Linux-6.
|
80 |
"processor": "x86_64",
|
81 |
"python_version": "3.10.15",
|
82 |
"optimum_benchmark_version": "0.5.0.dev0",
|
83 |
-
"optimum_benchmark_commit": "
|
84 |
-
"transformers_version": "4.
|
85 |
"transformers_commit": null,
|
86 |
-
"accelerate_version": "1.
|
87 |
"accelerate_commit": null,
|
88 |
-
"diffusers_version": "0.
|
89 |
"diffusers_commit": null,
|
90 |
"optimum_version": null,
|
91 |
"optimum_commit": null,
|
92 |
-
"timm_version": "1.0.
|
93 |
"timm_commit": null,
|
94 |
"peft_version": null,
|
95 |
"peft_commit": null
|
@@ -101,7 +100,7 @@
|
|
101 |
"load": {
|
102 |
"memory": {
|
103 |
"unit": "MB",
|
104 |
-
"max_ram":
|
105 |
"max_global_vram": null,
|
106 |
"max_process_vram": null,
|
107 |
"max_reserved": null,
|
@@ -110,15 +109,15 @@
|
|
110 |
"latency": {
|
111 |
"unit": "s",
|
112 |
"values": [
|
113 |
-
1.
|
114 |
],
|
115 |
"count": 1,
|
116 |
-
"total": 1.
|
117 |
-
"mean": 1.
|
118 |
-
"p50": 1.
|
119 |
-
"p90": 1.
|
120 |
-
"p95": 1.
|
121 |
-
"p99": 1.
|
122 |
"stdev": 0,
|
123 |
"stdev_": 0
|
124 |
},
|
@@ -129,7 +128,7 @@
|
|
129 |
"prefill": {
|
130 |
"memory": {
|
131 |
"unit": "MB",
|
132 |
-
"max_ram":
|
133 |
"max_global_vram": null,
|
134 |
"max_process_vram": null,
|
135 |
"max_reserved": null,
|
@@ -138,51 +137,51 @@
|
|
138 |
"latency": {
|
139 |
"unit": "s",
|
140 |
"values": [
|
141 |
-
0.
|
142 |
-
0.
|
143 |
-
0.
|
144 |
-
0.
|
145 |
-
0.
|
146 |
-
0.
|
147 |
-
0.
|
148 |
-
0.
|
149 |
-
0.
|
150 |
-
0.
|
151 |
-
0.
|
152 |
-
0.
|
153 |
-
0.
|
154 |
-
0.
|
155 |
],
|
156 |
"count": 14,
|
157 |
-
"total": 0.
|
158 |
-
"mean": 0.
|
159 |
-
"p50": 0.
|
160 |
-
"p90": 0.
|
161 |
-
"p95": 0.
|
162 |
-
"p99": 0.
|
163 |
-
"stdev": 0.
|
164 |
-
"stdev_":
|
165 |
},
|
166 |
"throughput": {
|
167 |
"unit": "tokens/s",
|
168 |
-
"value": 44.
|
169 |
},
|
170 |
"energy": {
|
171 |
"unit": "kWh",
|
172 |
-
"cpu": 1.
|
173 |
-
"ram": 7.
|
174 |
"gpu": 0.0,
|
175 |
-
"total": 1.
|
176 |
},
|
177 |
"efficiency": {
|
178 |
"unit": "tokens/kWh",
|
179 |
-
"value":
|
180 |
}
|
181 |
},
|
182 |
"decode": {
|
183 |
"memory": {
|
184 |
"unit": "MB",
|
185 |
-
"max_ram":
|
186 |
"max_global_vram": null,
|
187 |
"max_process_vram": null,
|
188 |
"max_reserved": null,
|
@@ -191,45 +190,45 @@
|
|
191 |
"latency": {
|
192 |
"unit": "s",
|
193 |
"values": [
|
194 |
-
0.
|
195 |
-
0.
|
196 |
-
0.
|
197 |
-
0.
|
198 |
-
0.
|
199 |
-
0.
|
200 |
-
0.
|
201 |
-
0.
|
202 |
-
0.
|
203 |
-
0.
|
204 |
-
0.
|
205 |
-
0.
|
206 |
-
0.
|
207 |
-
0.
|
208 |
],
|
209 |
"count": 14,
|
210 |
-
"total": 0.
|
211 |
-
"mean": 0.
|
212 |
-
"p50": 0.
|
213 |
-
"p90": 0.
|
214 |
-
"p95": 0.
|
215 |
-
"p99": 0.
|
216 |
-
"stdev": 0.
|
217 |
-
"stdev_":
|
218 |
},
|
219 |
"throughput": {
|
220 |
"unit": "tokens/s",
|
221 |
-
"value": 36.
|
222 |
},
|
223 |
"energy": {
|
224 |
"unit": "kWh",
|
225 |
-
"cpu":
|
226 |
-
"ram": 4.
|
227 |
"gpu": 0.0,
|
228 |
-
"total": 1.
|
229 |
},
|
230 |
"efficiency": {
|
231 |
"unit": "tokens/kWh",
|
232 |
-
"value":
|
233 |
}
|
234 |
},
|
235 |
"per_token": {
|
@@ -237,34 +236,34 @@
|
|
237 |
"latency": {
|
238 |
"unit": "s",
|
239 |
"values": [
|
240 |
-
0.
|
241 |
-
0.
|
242 |
-
0.
|
243 |
-
0.
|
244 |
-
0.
|
245 |
-
0.
|
246 |
-
0.
|
247 |
-
0.
|
248 |
-
0.
|
249 |
-
0.
|
250 |
-
0.
|
251 |
-
0.
|
252 |
-
0.
|
253 |
-
0.
|
254 |
],
|
255 |
"count": 14,
|
256 |
-
"total": 0.
|
257 |
-
"mean": 0.
|
258 |
-
"p50": 0.
|
259 |
-
"p90": 0.
|
260 |
-
"p95": 0.
|
261 |
-
"p99": 0.
|
262 |
-
"stdev": 0.
|
263 |
-
"stdev_":
|
264 |
},
|
265 |
"throughput": {
|
266 |
"unit": "tokens/s",
|
267 |
-
"value": 37.
|
268 |
},
|
269 |
"energy": null,
|
270 |
"efficiency": null
|
|
|
3 |
"name": "cpu_inference_transformers_text-generation_openai-community/gpt2",
|
4 |
"backend": {
|
5 |
"name": "pytorch",
|
6 |
+
"version": "2.5.1+cpu",
|
7 |
"_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
|
8 |
"task": "text-generation",
|
9 |
"library": "transformers",
|
|
|
45 |
"warmup_runs": 1,
|
46 |
"input_shapes": {
|
47 |
"batch_size": 1,
|
|
|
48 |
"sequence_length": 2
|
49 |
},
|
50 |
"new_tokens": null,
|
|
|
72 |
"environment": {
|
73 |
"cpu": " AMD EPYC 7763 64-Core Processor",
|
74 |
"cpu_count": 4,
|
75 |
+
"cpu_ram_mb": 16757.342208,
|
76 |
"system": "Linux",
|
77 |
"machine": "x86_64",
|
78 |
+
"platform": "Linux-6.5.0-1025-azure-x86_64-with-glibc2.35",
|
79 |
"processor": "x86_64",
|
80 |
"python_version": "3.10.15",
|
81 |
"optimum_benchmark_version": "0.5.0.dev0",
|
82 |
+
"optimum_benchmark_commit": "6807ba28334c8c98abf72a03d78f86133328d180",
|
83 |
+
"transformers_version": "4.46.3",
|
84 |
"transformers_commit": null,
|
85 |
+
"accelerate_version": "1.1.1",
|
86 |
"accelerate_commit": null,
|
87 |
+
"diffusers_version": "0.31.0",
|
88 |
"diffusers_commit": null,
|
89 |
"optimum_version": null,
|
90 |
"optimum_commit": null,
|
91 |
+
"timm_version": "1.0.11",
|
92 |
"timm_commit": null,
|
93 |
"peft_version": null,
|
94 |
"peft_commit": null
|
|
|
100 |
"load": {
|
101 |
"memory": {
|
102 |
"unit": "MB",
|
103 |
+
"max_ram": 1130.385408,
|
104 |
"max_global_vram": null,
|
105 |
"max_process_vram": null,
|
106 |
"max_reserved": null,
|
|
|
109 |
"latency": {
|
110 |
"unit": "s",
|
111 |
"values": [
|
112 |
+
1.7419912789999898
|
113 |
],
|
114 |
"count": 1,
|
115 |
+
"total": 1.7419912789999898,
|
116 |
+
"mean": 1.7419912789999898,
|
117 |
+
"p50": 1.7419912789999898,
|
118 |
+
"p90": 1.7419912789999898,
|
119 |
+
"p95": 1.7419912789999898,
|
120 |
+
"p99": 1.7419912789999898,
|
121 |
"stdev": 0,
|
122 |
"stdev_": 0
|
123 |
},
|
|
|
128 |
"prefill": {
|
129 |
"memory": {
|
130 |
"unit": "MB",
|
131 |
+
"max_ram": 985.68192,
|
132 |
"max_global_vram": null,
|
133 |
"max_process_vram": null,
|
134 |
"max_reserved": null,
|
|
|
137 |
"latency": {
|
138 |
"unit": "s",
|
139 |
"values": [
|
140 |
+
0.047665391999998974,
|
141 |
+
0.049731436999991274,
|
142 |
+
0.04895271399999501,
|
143 |
+
0.04833657500000754,
|
144 |
+
0.04391997499999434,
|
145 |
+
0.043644501000017044,
|
146 |
+
0.04382332400001587,
|
147 |
+
0.04319213599995919,
|
148 |
+
0.04416296799996644,
|
149 |
+
0.044141969000008885,
|
150 |
+
0.04314894600003072,
|
151 |
+
0.04354990399997405,
|
152 |
+
0.043227575999992496,
|
153 |
+
0.04241360399998939
|
154 |
],
|
155 |
"count": 14,
|
156 |
+
"total": 0.6299110209999412,
|
157 |
+
"mean": 0.04499364435713866,
|
158 |
+
"p50": 0.043871649500005105,
|
159 |
+
"p90": 0.04876787229999877,
|
160 |
+
"p95": 0.04922526704999371,
|
161 |
+
"p99": 0.04963020300999176,
|
162 |
+
"stdev": 0.0024009023029775032,
|
163 |
+
"stdev_": 5.336092102076141
|
164 |
},
|
165 |
"throughput": {
|
166 |
"unit": "tokens/s",
|
167 |
+
"value": 44.45072250927105
|
168 |
},
|
169 |
"energy": {
|
170 |
"unit": "kWh",
|
171 |
+
"cpu": 1.6854151983796455e-06,
|
172 |
+
"ram": 7.043250781543132e-08,
|
173 |
"gpu": 0.0,
|
174 |
+
"total": 1.7558477061950768e-06
|
175 |
},
|
176 |
"efficiency": {
|
177 |
"unit": "tokens/kWh",
|
178 |
+
"value": 1139050.9512547653
|
179 |
}
|
180 |
},
|
181 |
"decode": {
|
182 |
"memory": {
|
183 |
"unit": "MB",
|
184 |
+
"max_ram": 985.68192,
|
185 |
"max_global_vram": null,
|
186 |
"max_process_vram": null,
|
187 |
"max_reserved": null,
|
|
|
190 |
"latency": {
|
191 |
"unit": "s",
|
192 |
"values": [
|
193 |
+
0.030765811999970083,
|
194 |
+
0.030026732999999695,
|
195 |
+
0.03023100500001874,
|
196 |
+
0.02618368299999929,
|
197 |
+
0.026192708999985825,
|
198 |
+
0.026696138000033898,
|
199 |
+
0.026371824000023025,
|
200 |
+
0.026651877000006152,
|
201 |
+
0.026894829000013942,
|
202 |
+
0.026896673000010196,
|
203 |
+
0.027024220999976478,
|
204 |
+
0.02588880300004348,
|
205 |
+
0.026431090000016866,
|
206 |
+
0.025466253999979926
|
207 |
],
|
208 |
"count": 14,
|
209 |
+
"total": 0.3817216510000776,
|
210 |
+
"mean": 0.02726583221429126,
|
211 |
+
"p50": 0.026674007500020025,
|
212 |
+
"p90": 0.030169723400013025,
|
213 |
+
"p95": 0.03041818745000171,
|
214 |
+
"p99": 0.030696287089976407,
|
215 |
+
"stdev": 0.0016615794659438356,
|
216 |
+
"stdev_": 6.093998719294277
|
217 |
},
|
218 |
"throughput": {
|
219 |
"unit": "tokens/s",
|
220 |
+
"value": 36.67593903390393
|
221 |
},
|
222 |
"energy": {
|
223 |
"unit": "kWh",
|
224 |
+
"cpu": 1.0257383568054851e-06,
|
225 |
+
"ram": 4.287120110748343e-08,
|
226 |
"gpu": 0.0,
|
227 |
+
"total": 1.068609557912968e-06
|
228 |
},
|
229 |
"efficiency": {
|
230 |
"unit": "tokens/kWh",
|
231 |
+
"value": 935795.4854465602
|
232 |
}
|
233 |
},
|
234 |
"per_token": {
|
|
|
236 |
"latency": {
|
237 |
"unit": "s",
|
238 |
"values": [
|
239 |
+
0.030299631999980647,
|
240 |
+
0.029594156000030125,
|
241 |
+
0.029819386999974995,
|
242 |
+
0.02586457799998243,
|
243 |
+
0.025855570000032913,
|
244 |
+
0.02636103300000059,
|
245 |
+
0.026048771000034776,
|
246 |
+
0.02631907600004979,
|
247 |
+
0.02656132700002445,
|
248 |
+
0.02656453299999839,
|
249 |
+
0.02669703099996923,
|
250 |
+
0.02555920800000422,
|
251 |
+
0.026109781000002386,
|
252 |
+
0.02515429199996788
|
253 |
],
|
254 |
"count": 14,
|
255 |
+
"total": 0.3768083750000528,
|
256 |
+
"mean": 0.0269148839285752,
|
257 |
+
"p50": 0.02634005450002519,
|
258 |
+
"p90": 0.029751817699991533,
|
259 |
+
"p95": 0.029987472749976974,
|
260 |
+
"p99": 0.030237200149979913,
|
261 |
+
"stdev": 0.001616638367978842,
|
262 |
+
"stdev_": 6.006484636043616
|
263 |
},
|
264 |
"throughput": {
|
265 |
"unit": "tokens/s",
|
266 |
+
"value": 37.154163571863386
|
267 |
},
|
268 |
"energy": null,
|
269 |
"efficiency": null
|