hf-transformers-bot commited on
Commit
a2880e0
·
verified ·
1 Parent(s): e4160f0

Upload benchmark results for run 18503598619

Browse files
2025-10-14/18503598619/benchmark_results/Llama-2-7b-hf/Llama-2-7b-hf_benchmark_20251014_164807.json ADDED
@@ -0,0 +1,1121 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "Llama-2-7b-hf",
3
+ "benchmark_scenarios": [
4
+ {
5
+ "scenario_name": "eager_eager_attn",
6
+ "metadata": {
7
+ "timestamp": "2025-10-14T16:45:25.574509",
8
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
9
+ "hardware_info": {
10
+ "gpu_name": "unknown",
11
+ "gpu_memory_total_mb": 0,
12
+ "cpu_count": 256,
13
+ "memory_total_mb": 3091588,
14
+ "python_version": "3.12.10",
15
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
16
+ "cuda_version": null
17
+ },
18
+ "config": {
19
+ "name": "eager",
20
+ "model_id": "meta-llama/Llama-2-7b-hf",
21
+ "variant": "eager",
22
+ "warmup_iterations": 3,
23
+ "measurement_iterations": 5,
24
+ "num_tokens_to_generate": 100,
25
+ "device": "cuda",
26
+ "torch_dtype": "float16",
27
+ "compile_mode": null,
28
+ "compile_options": {},
29
+ "use_cache": true,
30
+ "batch_size": 1,
31
+ "sequence_length": null,
32
+ "attn_implementation": "eager",
33
+ "sdpa_backend": null,
34
+ "custom_params": {}
35
+ }
36
+ },
37
+ "measurements": {
38
+ "latency_seconds": {
39
+ "name": "latency_seconds",
40
+ "measurements": [
41
+ 1.3390462646484376,
42
+ 1.204842529296875,
43
+ 1.1875853271484376,
44
+ 1.1835208740234375,
45
+ 1.1794454345703125
46
+ ],
47
+ "mean": 1.2188880859375,
48
+ "median": 1.1875853271484376,
49
+ "std": 0.0606987490904525,
50
+ "min": 1.1794454345703125,
51
+ "max": 1.3390462646484376,
52
+ "p25": 1.1835208740234375,
53
+ "p75": 1.204842529296875,
54
+ "p90": 1.2853647705078126,
55
+ "p95": 1.312205517578125,
56
+ "p99": 1.333678115234375,
57
+ "unit": "seconds"
58
+ },
59
+ "time_to_first_token_seconds": {
60
+ "name": "time_to_first_token_seconds",
61
+ "measurements": [
62
+ 0.014856541633605957,
63
+ 0.014307372093200684,
64
+ 0.012768298149108888,
65
+ 0.012747146606445313,
66
+ 0.012734288215637208
67
+ ],
68
+ "mean": 0.013482729339599611,
69
+ "median": 0.012768298149108888,
70
+ "std": 0.0009142268744166934,
71
+ "min": 0.012734288215637208,
72
+ "max": 0.014856541633605957,
73
+ "p25": 0.012747146606445313,
74
+ "p75": 0.014307372093200684,
75
+ "p90": 0.014636873817443847,
76
+ "p95": 0.014746707725524902,
77
+ "p99": 0.014834574851989746,
78
+ "unit": "seconds"
79
+ },
80
+ "tokens_per_second": {
81
+ "name": "tokens_per_second",
82
+ "measurements": [
83
+ 74.6800186371863,
84
+ 82.99839818765217,
85
+ 84.20447585026528,
86
+ 84.49365126957589,
87
+ 84.78560946435925
88
+ ],
89
+ "mean": 82.23243068180777,
90
+ "median": 84.20447585026528,
91
+ "std": 3.8248251332683734,
92
+ "min": 74.6800186371863,
93
+ "max": 84.78560946435925,
94
+ "p25": 82.99839818765217,
95
+ "p75": 84.49365126957589,
96
+ "p90": 84.66882618644591,
97
+ "p95": 84.72721782540258,
98
+ "p99": 84.77393113656792,
99
+ "unit": "tokens/sec"
100
+ },
101
+ "time_per_output_token_seconds": {
102
+ "name": "time_per_output_token_seconds",
103
+ "measurements": [
104
+ 0.013390462646484375,
105
+ 0.01204842529296875,
106
+ 0.011875853271484375,
107
+ 0.011835208740234375,
108
+ 0.011794454345703125
109
+ ],
110
+ "mean": 0.012188880859375001,
111
+ "median": 0.011875853271484375,
112
+ "std": 0.0006069874909045248,
113
+ "min": 0.011794454345703125,
114
+ "max": 0.013390462646484375,
115
+ "p25": 0.011835208740234375,
116
+ "p75": 0.01204842529296875,
117
+ "p90": 0.012853647705078125,
118
+ "p95": 0.01312205517578125,
119
+ "p99": 0.01333678115234375,
120
+ "unit": "seconds/token"
121
+ }
122
+ },
123
+ "gpu_metrics": {
124
+ "gpu_monitoring_status": "disabled",
125
+ "gpu_monitoring_reason": "no_gpus_available"
126
+ }
127
+ },
128
+ {
129
+ "scenario_name": "compiled_compile_max-autotune_eager_attn",
130
+ "metadata": {
131
+ "timestamp": "2025-10-14T16:45:40.204358",
132
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
133
+ "hardware_info": {
134
+ "gpu_name": "unknown",
135
+ "gpu_memory_total_mb": 0,
136
+ "cpu_count": 256,
137
+ "memory_total_mb": 3091588,
138
+ "python_version": "3.12.10",
139
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
140
+ "cuda_version": null
141
+ },
142
+ "config": {
143
+ "name": "compiled",
144
+ "model_id": "meta-llama/Llama-2-7b-hf",
145
+ "variant": "compiled",
146
+ "warmup_iterations": 3,
147
+ "measurement_iterations": 5,
148
+ "num_tokens_to_generate": 100,
149
+ "device": "cuda",
150
+ "torch_dtype": "float16",
151
+ "compile_mode": "max-autotune",
152
+ "compile_options": {},
153
+ "use_cache": true,
154
+ "batch_size": 1,
155
+ "sequence_length": null,
156
+ "attn_implementation": "eager",
157
+ "sdpa_backend": null,
158
+ "custom_params": {}
159
+ }
160
+ },
161
+ "measurements": {
162
+ "latency_seconds": {
163
+ "name": "latency_seconds",
164
+ "measurements": [
165
+ 2.82615966796875,
166
+ 2.813818115234375,
167
+ 2.85666455078125,
168
+ 2.90904052734375,
169
+ 2.913306884765625
170
+ ],
171
+ "mean": 2.86379794921875,
172
+ "median": 2.85666455078125,
173
+ "std": 0.04114252050667118,
174
+ "min": 2.813818115234375,
175
+ "max": 2.913306884765625,
176
+ "p25": 2.82615966796875,
177
+ "p75": 2.90904052734375,
178
+ "p90": 2.911600341796875,
179
+ "p95": 2.91245361328125,
180
+ "p99": 2.91313623046875,
181
+ "unit": "seconds"
182
+ },
183
+ "time_to_first_token_seconds": {
184
+ "name": "time_to_first_token_seconds",
185
+ "measurements": [
186
+ 0.015611336708068848,
187
+ 0.015397101402282716,
188
+ 0.015785394668579103,
189
+ 0.015571879386901856,
190
+ 0.015817082405090332
191
+ ],
192
+ "mean": 0.01563655891418457,
193
+ "median": 0.015611336708068848,
194
+ "std": 0.00015290540441206035,
195
+ "min": 0.015397101402282716,
196
+ "max": 0.015817082405090332,
197
+ "p25": 0.015571879386901856,
198
+ "p75": 0.015785394668579103,
199
+ "p90": 0.01580440731048584,
200
+ "p95": 0.015810744857788085,
201
+ "p99": 0.015815814895629884,
202
+ "unit": "seconds"
203
+ },
204
+ "tokens_per_second": {
205
+ "name": "tokens_per_second",
206
+ "measurements": [
207
+ 35.383705009092125,
208
+ 35.53889978125703,
209
+ 35.005860233975206,
210
+ 34.375595341502574,
211
+ 34.32525441206479
212
+ ],
213
+ "mean": 34.92586295557835,
214
+ "median": 35.005860233975206,
215
+ "std": 0.5010704800318887,
216
+ "min": 34.32525441206479,
217
+ "max": 35.53889978125703,
218
+ "p25": 34.375595341502574,
219
+ "p75": 35.383705009092125,
220
+ "p90": 35.47682187239107,
221
+ "p95": 35.50786082682405,
222
+ "p99": 35.53269199037043,
223
+ "unit": "tokens/sec"
224
+ },
225
+ "time_per_output_token_seconds": {
226
+ "name": "time_per_output_token_seconds",
227
+ "measurements": [
228
+ 0.0282615966796875,
229
+ 0.028138181152343747,
230
+ 0.0285666455078125,
231
+ 0.0290904052734375,
232
+ 0.029133068847656252
233
+ ],
234
+ "mean": 0.028637979492187503,
235
+ "median": 0.0285666455078125,
236
+ "std": 0.00041142520506671273,
237
+ "min": 0.028138181152343747,
238
+ "max": 0.029133068847656252,
239
+ "p25": 0.0282615966796875,
240
+ "p75": 0.0290904052734375,
241
+ "p90": 0.02911600341796875,
242
+ "p95": 0.029124536132812504,
243
+ "p99": 0.029131362304687502,
244
+ "unit": "seconds/token"
245
+ }
246
+ },
247
+ "gpu_metrics": {
248
+ "gpu_monitoring_status": "disabled",
249
+ "gpu_monitoring_reason": "no_gpus_available"
250
+ }
251
+ },
252
+ {
253
+ "scenario_name": "eager_sdpa_default",
254
+ "metadata": {
255
+ "timestamp": "2025-10-14T16:46:18.271211",
256
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
257
+ "hardware_info": {
258
+ "gpu_name": "unknown",
259
+ "gpu_memory_total_mb": 0,
260
+ "cpu_count": 256,
261
+ "memory_total_mb": 3091588,
262
+ "python_version": "3.12.10",
263
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
264
+ "cuda_version": null
265
+ },
266
+ "config": {
267
+ "name": "eager",
268
+ "model_id": "meta-llama/Llama-2-7b-hf",
269
+ "variant": "eager",
270
+ "warmup_iterations": 3,
271
+ "measurement_iterations": 5,
272
+ "num_tokens_to_generate": 100,
273
+ "device": "cuda",
274
+ "torch_dtype": "float16",
275
+ "compile_mode": null,
276
+ "compile_options": {},
277
+ "use_cache": true,
278
+ "batch_size": 1,
279
+ "sequence_length": null,
280
+ "attn_implementation": "sdpa",
281
+ "sdpa_backend": null,
282
+ "custom_params": {}
283
+ }
284
+ },
285
+ "measurements": {
286
+ "latency_seconds": {
287
+ "name": "latency_seconds",
288
+ "measurements": [
289
+ 1.2940545654296876,
290
+ 1.2887750244140626,
291
+ 1.290135009765625,
292
+ 1.283328369140625,
293
+ 1.282864013671875
294
+ ],
295
+ "mean": 1.287831396484375,
296
+ "median": 1.2887750244140626,
297
+ "std": 0.0042397416620737986,
298
+ "min": 1.282864013671875,
299
+ "max": 1.2940545654296876,
300
+ "p25": 1.283328369140625,
301
+ "p75": 1.290135009765625,
302
+ "p90": 1.2924867431640625,
303
+ "p95": 1.293270654296875,
304
+ "p99": 1.293897783203125,
305
+ "unit": "seconds"
306
+ },
307
+ "time_to_first_token_seconds": {
308
+ "name": "time_to_first_token_seconds",
309
+ "measurements": [
310
+ 0.013973237991333008,
311
+ 0.01388634967803955,
312
+ 0.013933819770812988,
313
+ 0.013996712684631348,
314
+ 0.0140050048828125
315
+ ],
316
+ "mean": 0.01395902500152588,
317
+ "median": 0.013973237991333008,
318
+ "std": 4.393979134993714e-05,
319
+ "min": 0.01388634967803955,
320
+ "max": 0.0140050048828125,
321
+ "p25": 0.013933819770812988,
322
+ "p75": 0.013996712684631348,
323
+ "p90": 0.01400168800354004,
324
+ "p95": 0.01400334644317627,
325
+ "p99": 0.014004673194885254,
326
+ "unit": "seconds"
327
+ },
328
+ "tokens_per_second": {
329
+ "name": "tokens_per_second",
330
+ "measurements": [
331
+ 77.27649410733716,
332
+ 77.5930617102583,
333
+ 77.51126761389624,
334
+ 77.92237934159014,
335
+ 77.95058473405548
336
+ ],
337
+ "mean": 77.65075750142748,
338
+ "median": 77.5930617102583,
339
+ "std": 0.255551996539586,
340
+ "min": 77.27649410733716,
341
+ "max": 77.95058473405548,
342
+ "p25": 77.51126761389624,
343
+ "p75": 77.92237934159014,
344
+ "p90": 77.93930257706934,
345
+ "p95": 77.94494365556241,
346
+ "p99": 77.94945651835687,
347
+ "unit": "tokens/sec"
348
+ },
349
+ "time_per_output_token_seconds": {
350
+ "name": "time_per_output_token_seconds",
351
+ "measurements": [
352
+ 0.012940545654296875,
353
+ 0.012887750244140626,
354
+ 0.01290135009765625,
355
+ 0.012833283691406249,
356
+ 0.01282864013671875
357
+ ],
358
+ "mean": 0.012878313964843749,
359
+ "median": 0.012887750244140626,
360
+ "std": 4.239741662073795e-05,
361
+ "min": 0.01282864013671875,
362
+ "max": 0.012940545654296875,
363
+ "p25": 0.012833283691406249,
364
+ "p75": 0.01290135009765625,
365
+ "p90": 0.012924867431640626,
366
+ "p95": 0.01293270654296875,
367
+ "p99": 0.01293897783203125,
368
+ "unit": "seconds/token"
369
+ }
370
+ },
371
+ "gpu_metrics": {
372
+ "gpu_monitoring_status": "disabled",
373
+ "gpu_monitoring_reason": "no_gpus_available"
374
+ }
375
+ },
376
+ {
377
+ "scenario_name": "eager_sdpa_math",
378
+ "metadata": {
379
+ "timestamp": "2025-10-14T16:46:33.477794",
380
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
381
+ "hardware_info": {
382
+ "gpu_name": "unknown",
383
+ "gpu_memory_total_mb": 0,
384
+ "cpu_count": 256,
385
+ "memory_total_mb": 3091588,
386
+ "python_version": "3.12.10",
387
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
388
+ "cuda_version": null
389
+ },
390
+ "config": {
391
+ "name": "eager",
392
+ "model_id": "meta-llama/Llama-2-7b-hf",
393
+ "variant": "eager",
394
+ "warmup_iterations": 3,
395
+ "measurement_iterations": 5,
396
+ "num_tokens_to_generate": 100,
397
+ "device": "cuda",
398
+ "torch_dtype": "float16",
399
+ "compile_mode": null,
400
+ "compile_options": {},
401
+ "use_cache": true,
402
+ "batch_size": 1,
403
+ "sequence_length": null,
404
+ "attn_implementation": "sdpa",
405
+ "sdpa_backend": "math",
406
+ "custom_params": {}
407
+ }
408
+ },
409
+ "measurements": {
410
+ "latency_seconds": {
411
+ "name": "latency_seconds",
412
+ "measurements": [
413
+ 1.2477208251953125,
414
+ 1.2004837646484374,
415
+ 1.1963392333984375,
416
+ 1.1896580810546875,
417
+ 1.186978515625
418
+ ],
419
+ "mean": 1.2042360839843749,
420
+ "median": 1.1963392333984375,
421
+ "std": 0.0222607488771342,
422
+ "min": 1.186978515625,
423
+ "max": 1.2477208251953125,
424
+ "p25": 1.1896580810546875,
425
+ "p75": 1.2004837646484374,
426
+ "p90": 1.2288260009765624,
427
+ "p95": 1.2382734130859374,
428
+ "p99": 1.2458313427734375,
429
+ "unit": "seconds"
430
+ },
431
+ "time_to_first_token_seconds": {
432
+ "name": "time_to_first_token_seconds",
433
+ "measurements": [
434
+ 0.015897880554199217,
435
+ 0.014037492752075196,
436
+ 0.014376593589782714,
437
+ 0.014052876472473145,
438
+ 0.014019386291503906
439
+ ],
440
+ "mean": 0.014476845932006835,
441
+ "median": 0.014052876472473145,
442
+ "std": 0.0007226950354905272,
443
+ "min": 0.014019386291503906,
444
+ "max": 0.015897880554199217,
445
+ "p25": 0.014037492752075196,
446
+ "p75": 0.014376593589782714,
447
+ "p90": 0.015289365768432615,
448
+ "p95": 0.015593623161315916,
449
+ "p99": 0.015837029075622556,
450
+ "unit": "seconds"
451
+ },
452
+ "tokens_per_second": {
453
+ "name": "tokens_per_second",
454
+ "measurements": [
455
+ 80.14613363878611,
456
+ 83.29975210392377,
457
+ 83.58833114243882,
458
+ 84.05776549791965,
459
+ 84.24752317218253
460
+ ],
461
+ "mean": 83.06790111105018,
462
+ "median": 83.58833114243882,
463
+ "std": 1.498844032843351,
464
+ "min": 80.14613363878611,
465
+ "max": 84.24752317218253,
466
+ "p25": 83.29975210392377,
467
+ "p75": 84.05776549791965,
468
+ "p90": 84.17162010247738,
469
+ "p95": 84.20957163732994,
470
+ "p99": 84.23993286521201,
471
+ "unit": "tokens/sec"
472
+ },
473
+ "time_per_output_token_seconds": {
474
+ "name": "time_per_output_token_seconds",
475
+ "measurements": [
476
+ 0.012477208251953125,
477
+ 0.012004837646484374,
478
+ 0.011963392333984375,
479
+ 0.011896580810546875,
480
+ 0.01186978515625
481
+ ],
482
+ "mean": 0.012042360839843749,
483
+ "median": 0.011963392333984375,
484
+ "std": 0.0002226074887713421,
485
+ "min": 0.01186978515625,
486
+ "max": 0.012477208251953125,
487
+ "p25": 0.011896580810546875,
488
+ "p75": 0.012004837646484374,
489
+ "p90": 0.012288260009765625,
490
+ "p95": 0.012382734130859375,
491
+ "p99": 0.012458313427734374,
492
+ "unit": "seconds/token"
493
+ }
494
+ },
495
+ "gpu_metrics": {
496
+ "gpu_monitoring_status": "disabled",
497
+ "gpu_monitoring_reason": "no_gpus_available"
498
+ }
499
+ },
500
+ {
501
+ "scenario_name": "eager_sdpa_flash_attention",
502
+ "metadata": {
503
+ "timestamp": "2025-10-14T16:46:49.285898",
504
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
505
+ "hardware_info": {
506
+ "gpu_name": "unknown",
507
+ "gpu_memory_total_mb": 0,
508
+ "cpu_count": 256,
509
+ "memory_total_mb": 3091588,
510
+ "python_version": "3.12.10",
511
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
512
+ "cuda_version": null
513
+ },
514
+ "config": {
515
+ "name": "eager",
516
+ "model_id": "meta-llama/Llama-2-7b-hf",
517
+ "variant": "eager",
518
+ "warmup_iterations": 3,
519
+ "measurement_iterations": 5,
520
+ "num_tokens_to_generate": 100,
521
+ "device": "cuda",
522
+ "torch_dtype": "float16",
523
+ "compile_mode": null,
524
+ "compile_options": {},
525
+ "use_cache": true,
526
+ "batch_size": 1,
527
+ "sequence_length": null,
528
+ "attn_implementation": "sdpa",
529
+ "sdpa_backend": "flash_attention",
530
+ "custom_params": {}
531
+ }
532
+ },
533
+ "measurements": {
534
+ "latency_seconds": {
535
+ "name": "latency_seconds",
536
+ "measurements": [
537
+ 0.97755126953125,
538
+ 0.9843341674804688,
539
+ 0.9782413940429687,
540
+ 1.2665042724609374,
541
+ 1.27641259765625
542
+ ],
543
+ "mean": 1.096608740234375,
544
+ "median": 0.9843341674804688,
545
+ "std": 0.14281807215897202,
546
+ "min": 0.97755126953125,
547
+ "max": 1.27641259765625,
548
+ "p25": 0.9782413940429687,
549
+ "p75": 1.2665042724609374,
550
+ "p90": 1.272449267578125,
551
+ "p95": 1.2744309326171874,
552
+ "p99": 1.2760162646484374,
553
+ "unit": "seconds"
554
+ },
555
+ "time_to_first_token_seconds": {
556
+ "name": "time_to_first_token_seconds",
557
+ "measurements": [
558
+ 0.010833080291748048,
559
+ 0.010851507186889648,
560
+ 0.01093106460571289,
561
+ 0.010935432434082031,
562
+ 0.014071904182434081
563
+ ],
564
+ "mean": 0.01152459774017334,
565
+ "median": 0.01093106460571289,
566
+ "std": 0.0012743166580394029,
567
+ "min": 0.010833080291748048,
568
+ "max": 0.014071904182434081,
569
+ "p25": 0.010851507186889648,
570
+ "p75": 0.010935432434082031,
571
+ "p90": 0.01281731548309326,
572
+ "p95": 0.01344460983276367,
573
+ "p99": 0.0139464453125,
574
+ "unit": "seconds"
575
+ },
576
+ "tokens_per_second": {
577
+ "name": "tokens_per_second",
578
+ "measurements": [
579
+ 102.2964248698193,
580
+ 101.59151566988983,
581
+ 102.22425733459359,
582
+ 78.9574912413762,
583
+ 78.34457305076752
584
+ ],
585
+ "mean": 92.68285243328928,
586
+ "median": 101.59151566988983,
587
+ "std": 11.461197656841312,
588
+ "min": 78.34457305076752,
589
+ "max": 102.2964248698193,
590
+ "p25": 78.9574912413762,
591
+ "p75": 102.22425733459359,
592
+ "p90": 102.26755785572902,
593
+ "p95": 102.28199136277416,
594
+ "p99": 102.29353816841028,
595
+ "unit": "tokens/sec"
596
+ },
597
+ "time_per_output_token_seconds": {
598
+ "name": "time_per_output_token_seconds",
599
+ "measurements": [
600
+ 0.0097755126953125,
601
+ 0.009843341674804688,
602
+ 0.009782413940429687,
603
+ 0.012665042724609374,
604
+ 0.012764125976562499
605
+ ],
606
+ "mean": 0.01096608740234375,
607
+ "median": 0.009843341674804688,
608
+ "std": 0.0014281807215897199,
609
+ "min": 0.0097755126953125,
610
+ "max": 0.012764125976562499,
611
+ "p25": 0.009782413940429687,
612
+ "p75": 0.012665042724609374,
613
+ "p90": 0.012724492675781249,
614
+ "p95": 0.012744309326171873,
615
+ "p99": 0.012760162646484373,
616
+ "unit": "seconds/token"
617
+ }
618
+ },
619
+ "gpu_metrics": {
620
+ "gpu_monitoring_status": "disabled",
621
+ "gpu_monitoring_reason": "no_gpus_available"
622
+ }
623
+ },
624
+ {
625
+ "scenario_name": "eager_sdpa_efficient_attention",
626
+ "metadata": {
627
+ "timestamp": "2025-10-14T16:47:02.894966",
628
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
629
+ "hardware_info": {
630
+ "gpu_name": "unknown",
631
+ "gpu_memory_total_mb": 0,
632
+ "cpu_count": 256,
633
+ "memory_total_mb": 3091588,
634
+ "python_version": "3.12.10",
635
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
636
+ "cuda_version": null
637
+ },
638
+ "config": {
639
+ "name": "eager",
640
+ "model_id": "meta-llama/Llama-2-7b-hf",
641
+ "variant": "eager",
642
+ "warmup_iterations": 3,
643
+ "measurement_iterations": 5,
644
+ "num_tokens_to_generate": 100,
645
+ "device": "cuda",
646
+ "torch_dtype": "float16",
647
+ "compile_mode": null,
648
+ "compile_options": {},
649
+ "use_cache": true,
650
+ "batch_size": 1,
651
+ "sequence_length": null,
652
+ "attn_implementation": "sdpa",
653
+ "sdpa_backend": "efficient_attention",
654
+ "custom_params": {}
655
+ }
656
+ },
657
+ "measurements": {
658
+ "latency_seconds": {
659
+ "name": "latency_seconds",
660
+ "measurements": [
661
+ 0.9612239379882812,
662
+ 0.9601256713867188,
663
+ 0.961915283203125,
664
+ 0.9606069946289062,
665
+ 0.9779444580078125
666
+ ],
667
+ "mean": 0.9643632690429687,
668
+ "median": 0.9612239379882812,
669
+ "std": 0.006817089690062623,
670
+ "min": 0.9601256713867188,
671
+ "max": 0.9779444580078125,
672
+ "p25": 0.9606069946289062,
673
+ "p75": 0.961915283203125,
674
+ "p90": 0.9715327880859375,
675
+ "p95": 0.974738623046875,
676
+ "p99": 0.977303291015625,
677
+ "unit": "seconds"
678
+ },
679
+ "time_to_first_token_seconds": {
680
+ "name": "time_to_first_token_seconds",
681
+ "measurements": [
682
+ 0.010618803977966308,
683
+ 0.01059160327911377,
684
+ 0.010604421615600585,
685
+ 0.01064348030090332,
686
+ 0.010587117195129395
687
+ ],
688
+ "mean": 0.010609085273742677,
689
+ "median": 0.010604421615600585,
690
+ "std": 2.0432510499684284e-05,
691
+ "min": 0.010587117195129395,
692
+ "max": 0.01064348030090332,
693
+ "p25": 0.01059160327911377,
694
+ "p75": 0.010618803977966308,
695
+ "p90": 0.010633609771728515,
696
+ "p95": 0.010638545036315918,
697
+ "p99": 0.01064249324798584,
698
+ "unit": "seconds"
699
+ },
700
+ "tokens_per_second": {
701
+ "name": "tokens_per_second",
702
+ "measurements": [
703
+ 104.0340299985529,
704
+ 104.15303223334195,
705
+ 103.95925893495058,
706
+ 104.10084515221668,
707
+ 102.2552959742844
708
+ ],
709
+ "mean": 103.7004924586693,
710
+ "median": 104.0340299985529,
711
+ "std": 0.7255170273850966,
712
+ "min": 102.2552959742844,
713
+ "max": 104.15303223334195,
714
+ "p25": 103.95925893495058,
715
+ "p75": 104.10084515221668,
716
+ "p90": 104.13215740089184,
717
+ "p95": 104.14259481711689,
718
+ "p99": 104.15094475009694,
719
+ "unit": "tokens/sec"
720
+ },
721
+ "time_per_output_token_seconds": {
722
+ "name": "time_per_output_token_seconds",
723
+ "measurements": [
724
+ 0.009612239379882813,
725
+ 0.009601256713867188,
726
+ 0.00961915283203125,
727
+ 0.009606069946289063,
728
+ 0.009779444580078124
729
+ ],
730
+ "mean": 0.009643632690429688,
731
+ "median": 0.009612239379882813,
732
+ "std": 6.817089690062585e-05,
733
+ "min": 0.009601256713867188,
734
+ "max": 0.009779444580078124,
735
+ "p25": 0.009606069946289063,
736
+ "p75": 0.00961915283203125,
737
+ "p90": 0.009715327880859374,
738
+ "p95": 0.00974738623046875,
739
+ "p99": 0.00977303291015625,
740
+ "unit": "seconds/token"
741
+ }
742
+ },
743
+ "gpu_metrics": {
744
+ "gpu_monitoring_status": "disabled",
745
+ "gpu_monitoring_reason": "no_gpus_available"
746
+ }
747
+ },
748
+ {
749
+ "scenario_name": "compiled_compile_max-autotune_sdpa_default",
750
+ "metadata": {
751
+ "timestamp": "2025-10-14T16:47:16.442234",
752
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
753
+ "hardware_info": {
754
+ "gpu_name": "unknown",
755
+ "gpu_memory_total_mb": 0,
756
+ "cpu_count": 256,
757
+ "memory_total_mb": 3091588,
758
+ "python_version": "3.12.10",
759
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
760
+ "cuda_version": null
761
+ },
762
+ "config": {
763
+ "name": "compiled",
764
+ "model_id": "meta-llama/Llama-2-7b-hf",
765
+ "variant": "compiled",
766
+ "warmup_iterations": 3,
767
+ "measurement_iterations": 5,
768
+ "num_tokens_to_generate": 100,
769
+ "device": "cuda",
770
+ "torch_dtype": "float16",
771
+ "compile_mode": "max-autotune",
772
+ "compile_options": {},
773
+ "use_cache": true,
774
+ "batch_size": 1,
775
+ "sequence_length": null,
776
+ "attn_implementation": "sdpa",
777
+ "sdpa_backend": null,
778
+ "custom_params": {}
779
+ }
780
+ },
781
+ "measurements": {
782
+ "latency_seconds": {
783
+ "name": "latency_seconds",
784
+ "measurements": [
785
+ 1.19472998046875,
786
+ 1.1893704833984375,
787
+ 1.1986197509765626,
788
+ 1.1894735107421874,
789
+ 1.19451953125
790
+ ],
791
+ "mean": 1.1933426513671876,
792
+ "median": 1.19451953125,
793
+ "std": 0.0035186894187997546,
794
+ "min": 1.1893704833984375,
795
+ "max": 1.1986197509765626,
796
+ "p25": 1.1894735107421874,
797
+ "p75": 1.19472998046875,
798
+ "p90": 1.1970638427734375,
799
+ "p95": 1.1978417968750001,
800
+ "p99": 1.19846416015625,
801
+ "unit": "seconds"
802
+ },
803
+ "time_to_first_token_seconds": {
804
+ "name": "time_to_first_token_seconds",
805
+ "measurements": [
806
+ 0.012587512016296386,
807
+ 0.01252906608581543,
808
+ 0.012641151428222657,
809
+ 0.012648000717163085,
810
+ 0.012591958045959473
811
+ ],
812
+ "mean": 0.012599537658691406,
813
+ "median": 0.012591958045959473,
814
+ "std": 4.3008564353678446e-05,
815
+ "min": 0.01252906608581543,
816
+ "max": 0.012648000717163085,
817
+ "p25": 0.012587512016296386,
818
+ "p75": 0.012641151428222657,
819
+ "p90": 0.012645261001586914,
820
+ "p95": 0.012646630859374999,
821
+ "p99": 0.012647726745605468,
822
+ "unit": "seconds"
823
+ },
824
+ "tokens_per_second": {
825
+ "name": "tokens_per_second",
826
+ "measurements": [
827
+ 83.70092124143834,
828
+ 84.07809122206048,
829
+ 83.42929433502665,
830
+ 84.07080872074545,
831
+ 83.71566758339682
832
+ ],
833
+ "mean": 83.79895662053354,
834
+ "median": 83.71566758339682,
835
+ "std": 0.2469890259695373,
836
+ "min": 83.42929433502665,
837
+ "max": 84.07809122206048,
838
+ "p25": 83.70092124143834,
839
+ "p75": 84.07080872074545,
840
+ "p90": 84.07517822153447,
841
+ "p95": 84.07663472179748,
842
+ "p99": 84.07779992200788,
843
+ "unit": "tokens/sec"
844
+ },
845
+ "time_per_output_token_seconds": {
846
+ "name": "time_per_output_token_seconds",
847
+ "measurements": [
848
+ 0.011947299804687499,
849
+ 0.011893704833984376,
850
+ 0.011986197509765626,
851
+ 0.011894735107421874,
852
+ 0.0119451953125
853
+ ],
854
+ "mean": 0.011933426513671875,
855
+ "median": 0.0119451953125,
856
+ "std": 3.5186894187997543e-05,
857
+ "min": 0.011893704833984376,
858
+ "max": 0.011986197509765626,
859
+ "p25": 0.011894735107421874,
860
+ "p75": 0.011947299804687499,
861
+ "p90": 0.011970638427734374,
862
+ "p95": 0.01197841796875,
863
+ "p99": 0.0119846416015625,
864
+ "unit": "seconds/token"
865
+ }
866
+ },
867
+ "gpu_metrics": {
868
+ "gpu_monitoring_status": "disabled",
869
+ "gpu_monitoring_reason": "no_gpus_available"
870
+ }
871
+ },
872
+ {
873
+ "scenario_name": "compiled_compile_max-autotune_sdpa_math",
874
+ "metadata": {
875
+ "timestamp": "2025-10-14T16:47:31.252992",
876
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
877
+ "hardware_info": {
878
+ "gpu_name": "unknown",
879
+ "gpu_memory_total_mb": 0,
880
+ "cpu_count": 256,
881
+ "memory_total_mb": 3091588,
882
+ "python_version": "3.12.10",
883
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
884
+ "cuda_version": null
885
+ },
886
+ "config": {
887
+ "name": "compiled",
888
+ "model_id": "meta-llama/Llama-2-7b-hf",
889
+ "variant": "compiled",
890
+ "warmup_iterations": 3,
891
+ "measurement_iterations": 5,
892
+ "num_tokens_to_generate": 100,
893
+ "device": "cuda",
894
+ "torch_dtype": "float16",
895
+ "compile_mode": "max-autotune",
896
+ "compile_options": {},
897
+ "use_cache": true,
898
+ "batch_size": 1,
899
+ "sequence_length": null,
900
+ "attn_implementation": "sdpa",
901
+ "sdpa_backend": "math",
902
+ "custom_params": {}
903
+ }
904
+ },
905
+ "measurements": {
906
+ "latency_seconds": {
907
+ "name": "latency_seconds",
908
+ "measurements": [
909
+ 1.4327379150390624,
910
+ 1.4171070556640626,
911
+ 1.415954345703125,
912
+ 1.4109727783203125,
913
+ 1.6265574951171875
914
+ ],
915
+ "mean": 1.46066591796875,
916
+ "median": 1.4171070556640626,
917
+ "std": 0.08326571446195447,
918
+ "min": 1.4109727783203125,
919
+ "max": 1.6265574951171875,
920
+ "p25": 1.415954345703125,
921
+ "p75": 1.4327379150390624,
922
+ "p90": 1.5490296630859375,
923
+ "p95": 1.5877935791015625,
924
+ "p99": 1.6188047119140625,
925
+ "unit": "seconds"
926
+ },
927
+ "time_to_first_token_seconds": {
928
+ "name": "time_to_first_token_seconds",
929
+ "measurements": [
930
+ 0.014985213279724121,
931
+ 0.015277523994445801,
932
+ 0.015312576293945312,
933
+ 0.015211627960205079,
934
+ 0.015109796524047852
935
+ ],
936
+ "mean": 0.015179347610473634,
937
+ "median": 0.015211627960205079,
938
+ "std": 0.00011912682103271219,
939
+ "min": 0.014985213279724121,
940
+ "max": 0.015312576293945312,
941
+ "p25": 0.015109796524047852,
942
+ "p75": 0.015277523994445801,
943
+ "p90": 0.015298555374145508,
944
+ "p95": 0.01530556583404541,
945
+ "p99": 0.015311174201965332,
946
+ "unit": "seconds"
947
+ },
948
+ "tokens_per_second": {
949
+ "name": "tokens_per_second",
950
+ "measurements": [
951
+ 69.79643586613228,
952
+ 70.56629885533917,
953
+ 70.62374595866132,
954
+ 70.87308950002894,
955
+ 61.47953595258271
956
+ ],
957
+ "mean": 68.66782122654888,
958
+ "median": 70.56629885533917,
959
+ "std": 3.612154016852685,
960
+ "min": 61.47953595258271,
961
+ "max": 70.87308950002894,
962
+ "p25": 69.79643586613228,
963
+ "p75": 70.62374595866132,
964
+ "p90": 70.7733520834819,
965
+ "p95": 70.82322079175542,
966
+ "p99": 70.86311575837424,
967
+ "unit": "tokens/sec"
968
+ },
969
+ "time_per_output_token_seconds": {
970
+ "name": "time_per_output_token_seconds",
971
+ "measurements": [
972
+ 0.014327379150390625,
973
+ 0.014171070556640625,
974
+ 0.01415954345703125,
975
+ 0.014109727783203125,
976
+ 0.016265574951171875
977
+ ],
978
+ "mean": 0.014606659179687499,
979
+ "median": 0.014171070556640625,
980
+ "std": 0.0008326571446195448,
981
+ "min": 0.014109727783203125,
982
+ "max": 0.016265574951171875,
983
+ "p25": 0.01415954345703125,
984
+ "p75": 0.014327379150390625,
985
+ "p90": 0.015490296630859376,
986
+ "p95": 0.015877935791015624,
987
+ "p99": 0.016188047119140625,
988
+ "unit": "seconds/token"
989
+ }
990
+ },
991
+ "gpu_metrics": {
992
+ "gpu_monitoring_status": "disabled",
993
+ "gpu_monitoring_reason": "no_gpus_available"
994
+ }
995
+ },
996
+ {
997
+ "scenario_name": "compiled_compile_max-autotune_sdpa_efficient_attention",
998
+ "metadata": {
999
+ "timestamp": "2025-10-14T16:47:53.262145",
1000
+ "commit_id": "9e4199ede396f136b3dff1e918816fcc3a65f0a0",
1001
+ "hardware_info": {
1002
+ "gpu_name": "unknown",
1003
+ "gpu_memory_total_mb": 0,
1004
+ "cpu_count": 256,
1005
+ "memory_total_mb": 3091588,
1006
+ "python_version": "3.12.10",
1007
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
1008
+ "cuda_version": null
1009
+ },
1010
+ "config": {
1011
+ "name": "compiled",
1012
+ "model_id": "meta-llama/Llama-2-7b-hf",
1013
+ "variant": "compiled",
1014
+ "warmup_iterations": 3,
1015
+ "measurement_iterations": 5,
1016
+ "num_tokens_to_generate": 100,
1017
+ "device": "cuda",
1018
+ "torch_dtype": "float16",
1019
+ "compile_mode": "max-autotune",
1020
+ "compile_options": {},
1021
+ "use_cache": true,
1022
+ "batch_size": 1,
1023
+ "sequence_length": null,
1024
+ "attn_implementation": "sdpa",
1025
+ "sdpa_backend": "efficient_attention",
1026
+ "custom_params": {}
1027
+ }
1028
+ },
1029
+ "measurements": {
1030
+ "latency_seconds": {
1031
+ "name": "latency_seconds",
1032
+ "measurements": [
1033
+ 1.520528564453125,
1034
+ 1.5196619873046875,
1035
+ 1.5177569580078125,
1036
+ 1.5210328369140624,
1037
+ 1.52907958984375
1038
+ ],
1039
+ "mean": 1.5216119873046876,
1040
+ "median": 1.520528564453125,
1041
+ "std": 0.0038971286561114435,
1042
+ "min": 1.5177569580078125,
1043
+ "max": 1.52907958984375,
1044
+ "p25": 1.5196619873046875,
1045
+ "p75": 1.5210328369140624,
1046
+ "p90": 1.525860888671875,
1047
+ "p95": 1.5274702392578123,
1048
+ "p99": 1.5287577197265625,
1049
+ "unit": "seconds"
1050
+ },
1051
+ "time_to_first_token_seconds": {
1052
+ "name": "time_to_first_token_seconds",
1053
+ "measurements": [
1054
+ 0.015781990051269533,
1055
+ 0.015999311447143554,
1056
+ 0.015809990882873534,
1057
+ 0.015766006469726562,
1058
+ 0.015670265197753906
1059
+ ],
1060
+ "mean": 0.015805512809753418,
1061
+ "median": 0.015781990051269533,
1062
+ "std": 0.00010768839548783476,
1063
+ "min": 0.015670265197753906,
1064
+ "max": 0.015999311447143554,
1065
+ "p25": 0.015766006469726562,
1066
+ "p75": 0.015809990882873534,
1067
+ "p90": 0.015923583221435545,
1068
+ "p95": 0.01596144733428955,
1069
+ "p99": 0.015991738624572753,
1070
+ "unit": "seconds"
1071
+ },
1072
+ "tokens_per_second": {
1073
+ "name": "tokens_per_second",
1074
+ "measurements": [
1075
+ 65.76660402033691,
1076
+ 65.80410698918818,
1077
+ 65.88670173600038,
1078
+ 65.74480022593355,
1079
+ 65.39881943635031
1080
+ ],
1081
+ "mean": 65.72020648156186,
1082
+ "median": 65.76660402033691,
1083
+ "std": 0.16781369528552512,
1084
+ "min": 65.39881943635031,
1085
+ "max": 65.88670173600038,
1086
+ "p25": 65.74480022593355,
1087
+ "p75": 65.80410698918818,
1088
+ "p90": 65.8536638372755,
1089
+ "p95": 65.87018278663794,
1090
+ "p99": 65.88339794612789,
1091
+ "unit": "tokens/sec"
1092
+ },
1093
+ "time_per_output_token_seconds": {
1094
+ "name": "time_per_output_token_seconds",
1095
+ "measurements": [
1096
+ 0.01520528564453125,
1097
+ 0.015196619873046874,
1098
+ 0.015177569580078125,
1099
+ 0.015210328369140625,
1100
+ 0.015290795898437499
1101
+ ],
1102
+ "mean": 0.015216119873046875,
1103
+ "median": 0.01520528564453125,
1104
+ "std": 3.897128656111459e-05,
1105
+ "min": 0.015177569580078125,
1106
+ "max": 0.015290795898437499,
1107
+ "p25": 0.015196619873046874,
1108
+ "p75": 0.015210328369140625,
1109
+ "p90": 0.015258608886718749,
1110
+ "p95": 0.015274702392578124,
1111
+ "p99": 0.015287577197265624,
1112
+ "unit": "seconds/token"
1113
+ }
1114
+ },
1115
+ "gpu_metrics": {
1116
+ "gpu_monitoring_status": "disabled",
1117
+ "gpu_monitoring_reason": "no_gpus_available"
1118
+ }
1119
+ }
1120
+ ]
1121
+ }