IlyasMoutawwakil HF Staff commited on
Commit
d2de28d
·
verified ·
1 Parent(s): c01cb04

Upload cpu_training_transformers_text-generation_hf-internal-testing/tiny-random-LlamaForCausalLM/benchmark.json with huggingface_hub

Browse files
cpu_training_transformers_text-generation_hf-internal-testing/tiny-random-LlamaForCausalLM/benchmark.json CHANGED
@@ -3,7 +3,7 @@
3
  "name": "cpu_training_transformers_text-generation_hf-internal-testing/tiny-random-LlamaForCausalLM",
4
  "backend": {
5
  "name": "pytorch",
6
- "version": "2.6.0+cpu",
7
  "_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
8
  "model": "hf-internal-testing/tiny-random-LlamaForCausalLM",
9
  "processor": "hf-internal-testing/tiny-random-LlamaForCausalLM",
@@ -25,6 +25,7 @@
25
  "low_cpu_mem_usage": null,
26
  "attn_implementation": null,
27
  "cache_implementation": null,
 
28
  "autocast_enabled": false,
29
  "autocast_dtype": null,
30
  "torch_compile": false,
@@ -51,7 +52,6 @@
51
  "per_device_train_batch_size": 2,
52
  "gradient_accumulation_steps": 1,
53
  "output_dir": "./trainer_output",
54
- "evaluation_strategy": "no",
55
  "eval_strategy": "no",
56
  "save_strategy": "no",
57
  "do_train": true,
@@ -79,23 +79,23 @@
79
  "environment": {
80
  "cpu": " AMD EPYC 7763 64-Core Processor",
81
  "cpu_count": 4,
82
- "cpu_ram_mb": 16766.763008,
83
  "system": "Linux",
84
  "machine": "x86_64",
85
- "platform": "Linux-6.8.0-1020-azure-x86_64-with-glibc2.39",
86
  "processor": "x86_64",
87
- "python_version": "3.10.16",
88
  "optimum_benchmark_version": "0.6.0.dev0",
89
- "optimum_benchmark_commit": "e0b65f8587af7a3174a7414ab6972a55c4d97a28",
90
- "transformers_version": "4.48.2",
91
  "transformers_commit": null,
92
- "accelerate_version": "1.3.0",
93
  "accelerate_commit": null,
94
- "diffusers_version": "0.32.2",
95
  "diffusers_commit": null,
96
  "optimum_version": null,
97
  "optimum_commit": null,
98
- "timm_version": "1.0.14",
99
  "timm_commit": null,
100
  "peft_version": null,
101
  "peft_commit": null
@@ -107,7 +107,7 @@
107
  "overall": {
108
  "memory": {
109
  "unit": "MB",
110
- "max_ram": 532.107264,
111
  "max_global_vram": null,
112
  "max_process_vram": null,
113
  "max_reserved": null,
@@ -116,42 +116,42 @@
116
  "latency": {
117
  "unit": "s",
118
  "values": [
119
- 0.02078934000002164,
120
- 0.013656629000024623,
121
- 0.013431731000025593,
122
- 0.013472086000035688,
123
- 0.01307712100003755
124
  ],
125
  "count": 5,
126
- "total": 0.07442690700014509,
127
- "mean": 0.014885381400029018,
128
- "p50": 0.013472086000035688,
129
- "p90": 0.01793625560002283,
130
- "p95": 0.019362797800022234,
131
- "p99": 0.020504031560021757,
132
- "stdev": 0.002957934030981527,
133
- "stdev_": 19.871402361082673
134
  },
135
  "throughput": {
136
  "unit": "samples/s",
137
- "value": 671.8000520954408
138
  },
139
  "energy": {
140
  "unit": "kWh",
141
- "cpu": 1.0123467316666139e-05,
142
- "ram": 4.228948615278479e-07,
143
  "gpu": 0,
144
- "total": 1.0546362178193988e-05
145
  },
146
  "efficiency": {
147
  "unit": "samples/kWh",
148
- "value": 948194.2522964303
149
  }
150
  },
151
  "warmup": {
152
  "memory": {
153
  "unit": "MB",
154
- "max_ram": 532.107264,
155
  "max_global_vram": null,
156
  "max_process_vram": null,
157
  "max_reserved": null,
@@ -160,22 +160,22 @@
160
  "latency": {
161
  "unit": "s",
162
  "values": [
163
- 0.02078934000002164,
164
- 0.013656629000024623
165
  ],
166
  "count": 2,
167
- "total": 0.03444596900004626,
168
- "mean": 0.01722298450002313,
169
- "p50": 0.01722298450002313,
170
- "p90": 0.020076068900021938,
171
- "p95": 0.020432704450021787,
172
- "p99": 0.02071801289002167,
173
- "stdev": 0.003566355499998508,
174
- "stdev_": 20.706954128616434
175
  },
176
  "throughput": {
177
  "unit": "samples/s",
178
- "value": 232.2477849291816
179
  },
180
  "energy": null,
181
  "efficiency": null
@@ -183,7 +183,7 @@
183
  "train": {
184
  "memory": {
185
  "unit": "MB",
186
- "max_ram": 532.107264,
187
  "max_global_vram": null,
188
  "max_process_vram": null,
189
  "max_reserved": null,
@@ -192,23 +192,23 @@
192
  "latency": {
193
  "unit": "s",
194
  "values": [
195
- 0.013431731000025593,
196
- 0.013472086000035688,
197
- 0.01307712100003755
198
  ],
199
  "count": 3,
200
- "total": 0.03998093800009883,
201
- "mean": 0.013326979333366277,
202
- "p50": 0.013431731000025593,
203
- "p90": 0.013464015000033669,
204
- "p95": 0.013468050500034679,
205
- "p99": 0.013471278900035487,
206
- "stdev": 0.00017744298907313021,
207
- "stdev_": 1.331456923842244
208
  },
209
  "throughput": {
210
  "unit": "samples/s",
211
- "value": 450.21454974256744
212
  },
213
  "energy": null,
214
  "efficiency": null
 
3
  "name": "cpu_training_transformers_text-generation_hf-internal-testing/tiny-random-LlamaForCausalLM",
4
  "backend": {
5
  "name": "pytorch",
6
+ "version": "2.7.0+cpu",
7
  "_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
8
  "model": "hf-internal-testing/tiny-random-LlamaForCausalLM",
9
  "processor": "hf-internal-testing/tiny-random-LlamaForCausalLM",
 
25
  "low_cpu_mem_usage": null,
26
  "attn_implementation": null,
27
  "cache_implementation": null,
28
+ "allow_tf32": false,
29
  "autocast_enabled": false,
30
  "autocast_dtype": null,
31
  "torch_compile": false,
 
52
  "per_device_train_batch_size": 2,
53
  "gradient_accumulation_steps": 1,
54
  "output_dir": "./trainer_output",
 
55
  "eval_strategy": "no",
56
  "save_strategy": "no",
57
  "do_train": true,
 
79
  "environment": {
80
  "cpu": " AMD EPYC 7763 64-Core Processor",
81
  "cpu_count": 4,
82
+ "cpu_ram_mb": 16766.410752,
83
  "system": "Linux",
84
  "machine": "x86_64",
85
+ "platform": "Linux-6.11.0-1012-azure-x86_64-with-glibc2.39",
86
  "processor": "x86_64",
87
+ "python_version": "3.10.17",
88
  "optimum_benchmark_version": "0.6.0.dev0",
89
+ "optimum_benchmark_commit": "78d6fd7c05e4903fc03580f46bd45ef32863e57b",
90
+ "transformers_version": "4.51.3",
91
  "transformers_commit": null,
92
+ "accelerate_version": "1.6.0",
93
  "accelerate_commit": null,
94
+ "diffusers_version": "0.33.1",
95
  "diffusers_commit": null,
96
  "optimum_version": null,
97
  "optimum_commit": null,
98
+ "timm_version": "1.0.15",
99
  "timm_commit": null,
100
  "peft_version": null,
101
  "peft_commit": null
 
107
  "overall": {
108
  "memory": {
109
  "unit": "MB",
110
+ "max_ram": 543.506432,
111
  "max_global_vram": null,
112
  "max_process_vram": null,
113
  "max_reserved": null,
 
116
  "latency": {
117
  "unit": "s",
118
  "values": [
119
+ 0.020551179999984015,
120
+ 0.014195597000025373,
121
+ 0.014158407000024908,
122
+ 0.014290523999989091,
123
+ 0.014062617999968552
124
  ],
125
  "count": 5,
126
+ "total": 0.07725832599999194,
127
+ "mean": 0.015451665199998387,
128
+ "p50": 0.014195597000025373,
129
+ "p90": 0.018046917599986046,
130
+ "p95": 0.01929904879998503,
131
+ "p99": 0.02030075375998422,
132
+ "stdev": 0.002550802865778706,
133
+ "stdev_": 16.50827165074074
134
  },
135
  "throughput": {
136
  "unit": "samples/s",
137
+ "value": 647.179437980642
138
  },
139
  "energy": {
140
  "unit": "kWh",
141
+ "cpu": 5.994502313333593e-06,
142
+ "ram": 2.140121688888996e-06,
143
  "gpu": 0,
144
+ "total": 8.13462400222259e-06
145
  },
146
  "efficiency": {
147
  "unit": "samples/kWh",
148
+ "value": 1229313.118500344
149
  }
150
  },
151
  "warmup": {
152
  "memory": {
153
  "unit": "MB",
154
+ "max_ram": 543.506432,
155
  "max_global_vram": null,
156
  "max_process_vram": null,
157
  "max_reserved": null,
 
160
  "latency": {
161
  "unit": "s",
162
  "values": [
163
+ 0.020551179999984015,
164
+ 0.014195597000025373
165
  ],
166
  "count": 2,
167
+ "total": 0.03474677700000939,
168
+ "mean": 0.017373388500004694,
169
+ "p50": 0.017373388500004694,
170
+ "p90": 0.019915621699988152,
171
+ "p95": 0.020233400849986084,
172
+ "p99": 0.02048762416998443,
173
+ "stdev": 0.003177791499979321,
174
+ "stdev_": 18.291143952594293
175
  },
176
  "throughput": {
177
  "unit": "samples/s",
178
+ "value": 230.2371814225486
179
  },
180
  "energy": null,
181
  "efficiency": null
 
183
  "train": {
184
  "memory": {
185
  "unit": "MB",
186
+ "max_ram": 543.506432,
187
  "max_global_vram": null,
188
  "max_process_vram": null,
189
  "max_reserved": null,
 
192
  "latency": {
193
  "unit": "s",
194
  "values": [
195
+ 0.014158407000024908,
196
+ 0.014290523999989091,
197
+ 0.014062617999968552
198
  ],
199
  "count": 3,
200
+ "total": 0.04251154899998255,
201
+ "mean": 0.014170516333327518,
202
+ "p50": 0.014158407000024908,
203
+ "p90": 0.014264100599996254,
204
+ "p95": 0.014277312299992673,
205
+ "p99": 0.014287881659989808,
206
+ "stdev": 9.343540790698045e-05,
207
+ "stdev_": 0.659364879226246
208
  },
209
  "throughput": {
210
  "unit": "samples/s",
211
+ "value": 423.4143526505559
212
  },
213
  "energy": null,
214
  "efficiency": null