Upload cuda_training_transformers_text-generation_hf-internal-testing/tiny-random-LlamaForCausalLM/benchmark.json with huggingface_hub
Browse files
cuda_training_transformers_text-generation_hf-internal-testing/tiny-random-LlamaForCausalLM/benchmark.json
CHANGED
@@ -11,7 +11,7 @@
|
|
11 |
"library": "transformers",
|
12 |
"model_type": "llama",
|
13 |
"device": "cuda",
|
14 |
-
"device_ids": "
|
15 |
"seed": 42,
|
16 |
"inter_op_num_threads": null,
|
17 |
"intra_op_num_threads": null,
|
@@ -120,7 +120,7 @@
|
|
120 |
"overall": {
|
121 |
"memory": {
|
122 |
"unit": "MB",
|
123 |
-
"max_ram":
|
124 |
"max_global_vram": 68702.69952,
|
125 |
"max_process_vram": 0.0,
|
126 |
"max_reserved": 182.452224,
|
@@ -129,25 +129,25 @@
|
|
129 |
"latency": {
|
130 |
"unit": "s",
|
131 |
"values": [
|
132 |
-
1.
|
133 |
-
0.
|
134 |
-
0.
|
135 |
-
0.
|
136 |
-
0.
|
137 |
],
|
138 |
"count": 5,
|
139 |
-
"total": 1.
|
140 |
-
"mean": 0.
|
141 |
-
"p50": 0.
|
142 |
-
"p90": 0.
|
143 |
-
"p95": 0.
|
144 |
-
"p99":
|
145 |
-
"stdev": 0.
|
146 |
-
"stdev_": 190.
|
147 |
},
|
148 |
"throughput": {
|
149 |
"unit": "samples/s",
|
150 |
-
"value":
|
151 |
},
|
152 |
"energy": null,
|
153 |
"efficiency": null
|
@@ -155,7 +155,7 @@
|
|
155 |
"warmup": {
|
156 |
"memory": {
|
157 |
"unit": "MB",
|
158 |
-
"max_ram":
|
159 |
"max_global_vram": 68702.69952,
|
160 |
"max_process_vram": 0.0,
|
161 |
"max_reserved": 182.452224,
|
@@ -164,22 +164,22 @@
|
|
164 |
"latency": {
|
165 |
"unit": "s",
|
166 |
"values": [
|
167 |
-
1.
|
168 |
-
0.
|
169 |
],
|
170 |
"count": 2,
|
171 |
-
"total": 1.
|
172 |
-
"mean": 0.
|
173 |
-
"p50": 0.
|
174 |
-
"p90": 0.
|
175 |
-
"p95":
|
176 |
-
"p99": 1.
|
177 |
-
"stdev": 0.
|
178 |
-
"stdev_": 97.
|
179 |
},
|
180 |
"throughput": {
|
181 |
"unit": "samples/s",
|
182 |
-
"value": 7.
|
183 |
},
|
184 |
"energy": null,
|
185 |
"efficiency": null
|
@@ -187,7 +187,7 @@
|
|
187 |
"train": {
|
188 |
"memory": {
|
189 |
"unit": "MB",
|
190 |
-
"max_ram":
|
191 |
"max_global_vram": 68702.69952,
|
192 |
"max_process_vram": 0.0,
|
193 |
"max_reserved": 182.452224,
|
@@ -196,23 +196,23 @@
|
|
196 |
"latency": {
|
197 |
"unit": "s",
|
198 |
"values": [
|
199 |
-
0.
|
200 |
-
0.
|
201 |
-
0.
|
202 |
],
|
203 |
"count": 3,
|
204 |
-
"total": 0.
|
205 |
-
"mean": 0.
|
206 |
-
"p50": 0.
|
207 |
-
"p90": 0.
|
208 |
-
"p95": 0.
|
209 |
-
"p99": 0.
|
210 |
-
"stdev":
|
211 |
-
"stdev_":
|
212 |
},
|
213 |
"throughput": {
|
214 |
"unit": "samples/s",
|
215 |
-
"value":
|
216 |
},
|
217 |
"energy": null,
|
218 |
"efficiency": null
|
|
|
11 |
"library": "transformers",
|
12 |
"model_type": "llama",
|
13 |
"device": "cuda",
|
14 |
+
"device_ids": "5",
|
15 |
"seed": 42,
|
16 |
"inter_op_num_threads": null,
|
17 |
"intra_op_num_threads": null,
|
|
|
120 |
"overall": {
|
121 |
"memory": {
|
122 |
"unit": "MB",
|
123 |
+
"max_ram": 2344.697856,
|
124 |
"max_global_vram": 68702.69952,
|
125 |
"max_process_vram": 0.0,
|
126 |
"max_reserved": 182.452224,
|
|
|
129 |
"latency": {
|
130 |
"unit": "s",
|
131 |
"values": [
|
132 |
+
1.0875611572265624,
|
133 |
+
0.012012459754943848,
|
134 |
+
0.010878222465515137,
|
135 |
+
0.011063180923461913,
|
136 |
+
0.011084941864013672
|
137 |
],
|
138 |
"count": 5,
|
139 |
+
"total": 1.1325999622344969,
|
140 |
+
"mean": 0.22651999244689938,
|
141 |
+
"p50": 0.011084941864013672,
|
142 |
+
"p90": 0.6573416782379151,
|
143 |
+
"p95": 0.8724514177322386,
|
144 |
+
"p99": 1.0445392093276977,
|
145 |
+
"stdev": 0.43052076387531185,
|
146 |
+
"stdev_": 190.05861655952253
|
147 |
},
|
148 |
"throughput": {
|
149 |
"unit": "samples/s",
|
150 |
+
"value": 44.14621372700333
|
151 |
},
|
152 |
"energy": null,
|
153 |
"efficiency": null
|
|
|
155 |
"warmup": {
|
156 |
"memory": {
|
157 |
"unit": "MB",
|
158 |
+
"max_ram": 2344.697856,
|
159 |
"max_global_vram": 68702.69952,
|
160 |
"max_process_vram": 0.0,
|
161 |
"max_reserved": 182.452224,
|
|
|
164 |
"latency": {
|
165 |
"unit": "s",
|
166 |
"values": [
|
167 |
+
1.0875611572265624,
|
168 |
+
0.012012459754943848
|
169 |
],
|
170 |
"count": 2,
|
171 |
+
"total": 1.0995736169815062,
|
172 |
+
"mean": 0.5497868084907531,
|
173 |
+
"p50": 0.5497868084907531,
|
174 |
+
"p90": 0.9800062874794006,
|
175 |
+
"p95": 1.0337837223529815,
|
176 |
+
"p99": 1.0768056702518463,
|
177 |
+
"stdev": 0.5377743487358093,
|
178 |
+
"stdev_": 97.81506948340216
|
179 |
},
|
180 |
"throughput": {
|
181 |
"unit": "samples/s",
|
182 |
+
"value": 7.275547427157441
|
183 |
},
|
184 |
"energy": null,
|
185 |
"efficiency": null
|
|
|
187 |
"train": {
|
188 |
"memory": {
|
189 |
"unit": "MB",
|
190 |
+
"max_ram": 2344.697856,
|
191 |
"max_global_vram": 68702.69952,
|
192 |
"max_process_vram": 0.0,
|
193 |
"max_reserved": 182.452224,
|
|
|
196 |
"latency": {
|
197 |
"unit": "s",
|
198 |
"values": [
|
199 |
+
0.010878222465515137,
|
200 |
+
0.011063180923461913,
|
201 |
+
0.011084941864013672
|
202 |
],
|
203 |
"count": 3,
|
204 |
+
"total": 0.033026345252990724,
|
205 |
+
"mean": 0.011008781750996907,
|
206 |
+
"p50": 0.011063180923461913,
|
207 |
+
"p90": 0.011080589675903321,
|
208 |
+
"p95": 0.011082765769958496,
|
209 |
+
"p99": 0.011084506645202637,
|
210 |
+
"stdev": 9.274581716554546e-05,
|
211 |
+
"stdev_": 0.8424712131035461
|
212 |
},
|
213 |
"throughput": {
|
214 |
"unit": "samples/s",
|
215 |
+
"value": 545.0194340946641
|
216 |
},
|
217 |
"energy": null,
|
218 |
"efficiency": null
|