Commit
•
f3acbe7
1
Parent(s):
50e5573
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 |
-
"
|
|
|
55 |
"do_train": true,
|
56 |
"use_cpu": false,
|
57 |
"max_steps": 5,
|
@@ -89,7 +90,7 @@
|
|
89 |
"gpu_count": 1,
|
90 |
"gpu_vram_mb": 68702699520,
|
91 |
"optimum_benchmark_version": "0.2.1",
|
92 |
-
"optimum_benchmark_commit": "
|
93 |
"transformers_version": "4.42.3",
|
94 |
"transformers_commit": null,
|
95 |
"accelerate_version": "0.31.0",
|
@@ -108,33 +109,33 @@
|
|
108 |
"overall": {
|
109 |
"memory": {
|
110 |
"unit": "MB",
|
111 |
-
"max_ram":
|
112 |
-
"max_global_vram": 3337.
|
113 |
-
"max_process_vram":
|
114 |
"max_reserved": 2894.06976,
|
115 |
"max_allocated": 2506.73664
|
116 |
},
|
117 |
"latency": {
|
118 |
"unit": "s",
|
119 |
"count": 5,
|
120 |
-
"total": 0.
|
121 |
-
"mean": 0.
|
122 |
-
"stdev": 0.
|
123 |
-
"p50": 0.
|
124 |
-
"p90": 0.
|
125 |
-
"p95": 0.
|
126 |
-
"p99": 0.
|
127 |
"values": [
|
128 |
-
0.
|
129 |
-
0.
|
130 |
-
0.
|
131 |
-
0.
|
132 |
-
0.
|
133 |
]
|
134 |
},
|
135 |
"throughput": {
|
136 |
"unit": "samples/s",
|
137 |
-
"value":
|
138 |
},
|
139 |
"energy": null,
|
140 |
"efficiency": null
|
@@ -142,30 +143,30 @@
|
|
142 |
"warmup": {
|
143 |
"memory": {
|
144 |
"unit": "MB",
|
145 |
-
"max_ram":
|
146 |
-
"max_global_vram": 3337.
|
147 |
-
"max_process_vram":
|
148 |
"max_reserved": 2894.06976,
|
149 |
"max_allocated": 2506.73664
|
150 |
},
|
151 |
"latency": {
|
152 |
"unit": "s",
|
153 |
"count": 2,
|
154 |
-
"total": 0.
|
155 |
-
"mean": 0.
|
156 |
-
"stdev": 0.
|
157 |
-
"p50": 0.
|
158 |
-
"p90": 0.
|
159 |
-
"p95": 0.
|
160 |
-
"p99": 0.
|
161 |
"values": [
|
162 |
-
0.
|
163 |
-
0.
|
164 |
]
|
165 |
},
|
166 |
"throughput": {
|
167 |
"unit": "samples/s",
|
168 |
-
"value":
|
169 |
},
|
170 |
"energy": null,
|
171 |
"efficiency": null
|
@@ -173,31 +174,31 @@
|
|
173 |
"train": {
|
174 |
"memory": {
|
175 |
"unit": "MB",
|
176 |
-
"max_ram":
|
177 |
-
"max_global_vram": 3337.
|
178 |
-
"max_process_vram":
|
179 |
"max_reserved": 2894.06976,
|
180 |
"max_allocated": 2506.73664
|
181 |
},
|
182 |
"latency": {
|
183 |
"unit": "s",
|
184 |
"count": 3,
|
185 |
-
"total": 0.
|
186 |
-
"mean": 0.
|
187 |
-
"stdev": 0.
|
188 |
-
"p50": 0.
|
189 |
-
"p90": 0.
|
190 |
-
"p95": 0.
|
191 |
-
"p99": 0.
|
192 |
"values": [
|
193 |
-
0.
|
194 |
-
0.
|
195 |
-
0.
|
196 |
]
|
197 |
},
|
198 |
"throughput": {
|
199 |
"unit": "samples/s",
|
200 |
-
"value":
|
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,
|
|
|
90 |
"gpu_count": 1,
|
91 |
"gpu_vram_mb": 68702699520,
|
92 |
"optimum_benchmark_version": "0.2.1",
|
93 |
+
"optimum_benchmark_commit": "9a8b29987613b8b04c221447a49b37ee314548ff",
|
94 |
"transformers_version": "4.42.3",
|
95 |
"transformers_commit": null,
|
96 |
"accelerate_version": "0.31.0",
|
|
|
109 |
"overall": {
|
110 |
"memory": {
|
111 |
"unit": "MB",
|
112 |
+
"max_ram": 1137.635328,
|
113 |
+
"max_global_vram": 3337.637888,
|
114 |
+
"max_process_vram": 306795.831296,
|
115 |
"max_reserved": 2894.06976,
|
116 |
"max_allocated": 2506.73664
|
117 |
},
|
118 |
"latency": {
|
119 |
"unit": "s",
|
120 |
"count": 5,
|
121 |
+
"total": 0.72186759185791,
|
122 |
+
"mean": 0.14437351837158202,
|
123 |
+
"stdev": 0.20630751004363443,
|
124 |
+
"p50": 0.04080127716064453,
|
125 |
+
"p90": 0.35144075622558596,
|
126 |
+
"p95": 0.4542122604370116,
|
127 |
+
"p99": 0.5364294638061523,
|
128 |
"values": [
|
129 |
+
0.5569837646484375,
|
130 |
+
0.043126243591308594,
|
131 |
+
0.04080127716064453,
|
132 |
+
0.040563514709472656,
|
133 |
+
0.04039279174804687
|
134 |
]
|
135 |
},
|
136 |
"throughput": {
|
137 |
"unit": "samples/s",
|
138 |
+
"value": 69.2647800842704
|
139 |
},
|
140 |
"energy": null,
|
141 |
"efficiency": null
|
|
|
143 |
"warmup": {
|
144 |
"memory": {
|
145 |
"unit": "MB",
|
146 |
+
"max_ram": 1137.635328,
|
147 |
+
"max_global_vram": 3337.637888,
|
148 |
+
"max_process_vram": 306795.831296,
|
149 |
"max_reserved": 2894.06976,
|
150 |
"max_allocated": 2506.73664
|
151 |
},
|
152 |
"latency": {
|
153 |
"unit": "s",
|
154 |
"count": 2,
|
155 |
+
"total": 0.600110008239746,
|
156 |
+
"mean": 0.300055004119873,
|
157 |
+
"stdev": 0.2569287605285644,
|
158 |
+
"p50": 0.300055004119873,
|
159 |
+
"p90": 0.5055980125427246,
|
160 |
+
"p95": 0.531290888595581,
|
161 |
+
"p99": 0.5518451894378662,
|
162 |
"values": [
|
163 |
+
0.5569837646484375,
|
164 |
+
0.043126243591308594
|
165 |
]
|
166 |
},
|
167 |
"throughput": {
|
168 |
"unit": "samples/s",
|
169 |
+
"value": 13.330889153916546
|
170 |
},
|
171 |
"energy": null,
|
172 |
"efficiency": null
|
|
|
174 |
"train": {
|
175 |
"memory": {
|
176 |
"unit": "MB",
|
177 |
+
"max_ram": 1137.635328,
|
178 |
+
"max_global_vram": 3337.637888,
|
179 |
+
"max_process_vram": 306795.831296,
|
180 |
"max_reserved": 2894.06976,
|
181 |
"max_allocated": 2506.73664
|
182 |
},
|
183 |
"latency": {
|
184 |
"unit": "s",
|
185 |
"count": 3,
|
186 |
+
"total": 0.12175758361816406,
|
187 |
+
"mean": 0.04058586120605469,
|
188 |
+
"stdev": 0.00016751041262836493,
|
189 |
+
"p50": 0.040563514709472656,
|
190 |
+
"p90": 0.040753724670410156,
|
191 |
+
"p95": 0.04077750091552734,
|
192 |
+
"p99": 0.04079652191162109,
|
193 |
"values": [
|
194 |
+
0.04080127716064453,
|
195 |
+
0.040563514709472656,
|
196 |
+
0.04039279174804687
|
197 |
]
|
198 |
},
|
199 |
"throughput": {
|
200 |
"unit": "samples/s",
|
201 |
+
"value": 147.83473410944663
|
202 |
},
|
203 |
"energy": null,
|
204 |
"efficiency": null
|