IlyasMoutawwakil HF staff commited on
Commit
6327b6a
·
verified ·
1 Parent(s): c0899ff

Upload cuda_training_transformers_text-generation_openai-community/gpt2/benchmark.json with huggingface_hub

Browse files
cuda_training_transformers_text-generation_openai-community/gpt2/benchmark.json CHANGED
@@ -51,7 +51,8 @@
51
  "per_device_train_batch_size": 2,
52
  "gradient_accumulation_steps": 1,
53
  "output_dir": "./trainer_output",
54
- "evaluation_strategy": "no",
 
55
  "do_train": true,
56
  "use_cpu": false,
57
  "max_steps": 5,
@@ -108,7 +109,7 @@
108
  "overall": {
109
  "memory": {
110
  "unit": "MB",
111
- "max_ram": 1715.286016,
112
  "max_global_vram": 3559.391232,
113
  "max_process_vram": 0.0,
114
  "max_reserved": 2910.846976,
@@ -117,24 +118,24 @@
117
  "latency": {
118
  "unit": "s",
119
  "count": 5,
120
- "total": 0.9063710289001465,
121
- "mean": 0.1812742057800293,
122
- "stdev": 0.2743232494671937,
123
- "p50": 0.04429510498046875,
124
- "p90": 0.45571216583251956,
125
- "p95": 0.5928163490295408,
126
- "p99": 0.7024996955871582,
127
  "values": [
128
- 0.7299205322265625,
129
- 0.04439961624145508,
130
- 0.04395110321044922,
131
- 0.04429510498046875,
132
- 0.043804672241210936
133
  ]
134
  },
135
  "throughput": {
136
  "unit": "samples/s",
137
- "value": 55.16504654906442
138
  },
139
  "energy": null,
140
  "efficiency": null
@@ -142,7 +143,7 @@
142
  "warmup": {
143
  "memory": {
144
  "unit": "MB",
145
- "max_ram": 1715.286016,
146
  "max_global_vram": 3559.391232,
147
  "max_process_vram": 0.0,
148
  "max_reserved": 2910.846976,
@@ -151,21 +152,21 @@
151
  "latency": {
152
  "unit": "s",
153
  "count": 2,
154
- "total": 0.7743201484680176,
155
- "mean": 0.3871600742340088,
156
- "stdev": 0.3427604579925537,
157
- "p50": 0.3871600742340088,
158
- "p90": 0.6613684406280518,
159
- "p95": 0.6956444864273071,
160
- "p99": 0.7230653230667115,
161
  "values": [
162
- 0.7299205322265625,
163
- 0.04439961624145508
164
  ]
165
  },
166
  "throughput": {
167
  "unit": "samples/s",
168
- "value": 10.33164385019284
169
  },
170
  "energy": null,
171
  "efficiency": null
@@ -173,7 +174,7 @@
173
  "train": {
174
  "memory": {
175
  "unit": "MB",
176
- "max_ram": 1715.286016,
177
  "max_global_vram": 3559.391232,
178
  "max_process_vram": 0.0,
179
  "max_reserved": 2910.846976,
@@ -182,22 +183,22 @@
182
  "latency": {
183
  "unit": "s",
184
  "count": 3,
185
- "total": 0.1320508804321289,
186
- "mean": 0.044016960144042966,
187
- "stdev": 0.00020556251232615235,
188
- "p50": 0.04395110321044922,
189
- "p90": 0.044226304626464846,
190
- "p95": 0.0442607048034668,
191
- "p99": 0.04428822494506836,
192
  "values": [
193
- 0.04395110321044922,
194
- 0.04429510498046875,
195
- 0.043804672241210936
196
  ]
197
  },
198
  "throughput": {
199
  "unit": "samples/s",
200
- "value": 136.31109418654412
201
  },
202
  "energy": null,
203
  "efficiency": null
 
51
  "per_device_train_batch_size": 2,
52
  "gradient_accumulation_steps": 1,
53
  "output_dir": "./trainer_output",
54
+ "eval_strategy": "no",
55
+ "save_strategy": "no",
56
  "do_train": true,
57
  "use_cpu": false,
58
  "max_steps": 5,
 
109
  "overall": {
110
  "memory": {
111
  "unit": "MB",
112
+ "max_ram": 1138.93376,
113
  "max_global_vram": 3559.391232,
114
  "max_process_vram": 0.0,
115
  "max_reserved": 2910.846976,
 
118
  "latency": {
119
  "unit": "s",
120
  "count": 5,
121
+ "total": 0.9721169776916505,
122
+ "mean": 0.1944233955383301,
123
+ "stdev": 0.3031372390558827,
124
+ "p50": 0.04314931106567383,
125
+ "p90": 0.49779608612060555,
126
+ "p95": 0.6492467063903807,
127
+ "p99": 0.7704072026062011,
128
  "values": [
129
+ 0.8006973266601562,
130
+ 0.04314931106567383,
131
+ 0.0434442253112793,
132
+ 0.04236492919921875,
133
+ 0.042461185455322265
134
  ]
135
  },
136
  "throughput": {
137
  "unit": "samples/s",
138
+ "value": 51.43413925217927
139
  },
140
  "energy": null,
141
  "efficiency": null
 
143
  "warmup": {
144
  "memory": {
145
  "unit": "MB",
146
+ "max_ram": 1138.93376,
147
  "max_global_vram": 3559.391232,
148
  "max_process_vram": 0.0,
149
  "max_reserved": 2910.846976,
 
152
  "latency": {
153
  "unit": "s",
154
  "count": 2,
155
+ "total": 0.8438466377258301,
156
+ "mean": 0.42192331886291506,
157
+ "stdev": 0.3787740077972412,
158
+ "p50": 0.42192331886291506,
159
+ "p90": 0.724942525100708,
160
+ "p95": 0.7628199258804321,
161
+ "p99": 0.7931218465042115,
162
  "values": [
163
+ 0.8006973266601562,
164
+ 0.04314931106567383
165
  ]
166
  },
167
  "throughput": {
168
  "unit": "samples/s",
169
+ "value": 9.480395657628062
170
  },
171
  "energy": null,
172
  "efficiency": null
 
174
  "train": {
175
  "memory": {
176
  "unit": "MB",
177
+ "max_ram": 1138.93376,
178
  "max_global_vram": 3559.391232,
179
  "max_process_vram": 0.0,
180
  "max_reserved": 2910.846976,
 
183
  "latency": {
184
  "unit": "s",
185
  "count": 3,
186
+ "total": 0.12827033996582032,
187
+ "mean": 0.042756779988606775,
188
+ "stdev": 0.00048768303945494326,
189
+ "p50": 0.042461185455322265,
190
+ "p90": 0.04324761734008789,
191
+ "p95": 0.043345921325683594,
192
+ "p99": 0.04342456451416016,
193
  "values": [
194
+ 0.0434442253112793,
195
+ 0.04236492919921875,
196
+ 0.042461185455322265
197
  ]
198
  },
199
  "throughput": {
200
  "unit": "samples/s",
201
+ "value": 140.328621603376
202
  },
203
  "energy": null,
204
  "efficiency": null