hf-transformers-bot commited on
Commit
d852ac1
·
verified ·
1 Parent(s): ccd74c1

Upload benchmark results for run 18472460805

Browse files
2025-10-13/18472460805/benchmark_results/Llama-2-7b-hf/Llama-2-7b-hf_benchmark_20251013_164918.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": "2025-10-13T16:45:43.505903",
8
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
9
+ "hardware_info": {
10
+ "gpu_name": "unknown",
11
+ "gpu_memory_total_mb": 0,
12
+ "cpu_count": 160,
13
+ "memory_total_mb": 1285717,
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
+ 1.253781982421875,
42
+ 1.247297607421875,
43
+ 1.2510565185546876,
44
+ 1.249962890625,
45
+ 1.2520672607421874
46
+ ],
47
+ "mean": 1.2508332519531247,
48
+ "median": 1.2510565185546876,
49
+ "std": 0.002169144906964286,
50
+ "min": 1.247297607421875,
51
+ "max": 1.253781982421875,
52
+ "p25": 1.249962890625,
53
+ "p75": 1.2520672607421874,
54
+ "p90": 1.25309609375,
55
+ "p95": 1.2534390380859375,
56
+ "p99": 1.2537133935546876,
57
+ "unit": "seconds"
58
+ },
59
+ "time_to_first_token_seconds": {
60
+ "name": "time_to_first_token_seconds",
61
+ "measurements": [
62
+ 0.013386244773864747,
63
+ 0.013458804130554199,
64
+ 0.013526283264160156,
65
+ 0.013448885917663574,
66
+ 0.013419803619384765
67
+ ],
68
+ "mean": 0.013448004341125488,
69
+ "median": 0.013448885917663574,
70
+ "std": 4.659297645632036e-05,
71
+ "min": 0.013386244773864747,
72
+ "max": 0.013526283264160156,
73
+ "p25": 0.013419803619384765,
74
+ "p75": 0.013458804130554199,
75
+ "p90": 0.013499291610717773,
76
+ "p95": 0.013512787437438964,
77
+ "p99": 0.013523584098815917,
78
+ "unit": "seconds"
79
+ },
80
+ "tokens_per_second": {
81
+ "name": "tokens_per_second",
82
+ "measurements": [
83
+ 79.7586832495666,
84
+ 80.17332784490533,
85
+ 79.93243991528644,
86
+ 80.00237507050991,
87
+ 79.86791375786237
88
+ ],
89
+ "mean": 79.94694796762613,
90
+ "median": 79.93243991528644,
91
+ "std": 0.1387231102107666,
92
+ "min": 79.7586832495666,
93
+ "max": 80.17332784490533,
94
+ "p25": 79.86791375786237,
95
+ "p75": 80.00237507050991,
96
+ "p90": 80.10494673514717,
97
+ "p95": 80.13913729002624,
98
+ "p99": 80.16648973392951,
99
+ "unit": "tokens/sec"
100
+ },
101
+ "time_per_output_token_seconds": {
102
+ "name": "time_per_output_token_seconds",
103
+ "measurements": [
104
+ 0.01253781982421875,
105
+ 0.01247297607421875,
106
+ 0.012510565185546876,
107
+ 0.012499628906249999,
108
+ 0.012520672607421875
109
+ ],
110
+ "mean": 0.012508332519531252,
111
+ "median": 0.012510565185546876,
112
+ "std": 2.1691449069642654e-05,
113
+ "min": 0.01247297607421875,
114
+ "max": 0.01253781982421875,
115
+ "p25": 0.012499628906249999,
116
+ "p75": 0.012520672607421875,
117
+ "p90": 0.0125309609375,
118
+ "p95": 0.012534390380859375,
119
+ "p99": 0.012537133935546875,
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": "2025-10-13T16:45:59.336358",
132
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
133
+ "hardware_info": {
134
+ "gpu_name": "unknown",
135
+ "gpu_memory_total_mb": 0,
136
+ "cpu_count": 160,
137
+ "memory_total_mb": 1285717,
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
+ 2.80608251953125,
166
+ 2.804099853515625,
167
+ 2.86517333984375,
168
+ 2.8766630859375,
169
+ 2.90892919921875
170
+ ],
171
+ "mean": 2.852189599609375,
172
+ "median": 2.86517333984375,
173
+ "std": 0.04104972038894004,
174
+ "min": 2.804099853515625,
175
+ "max": 2.90892919921875,
176
+ "p25": 2.80608251953125,
177
+ "p75": 2.8766630859375,
178
+ "p90": 2.89602275390625,
179
+ "p95": 2.9024759765625,
180
+ "p99": 2.9076385546875,
181
+ "unit": "seconds"
182
+ },
183
+ "time_to_first_token_seconds": {
184
+ "name": "time_to_first_token_seconds",
185
+ "measurements": [
186
+ 0.016007278442382814,
187
+ 0.016087999343872072,
188
+ 0.016185037612915038,
189
+ 0.016229835510253907,
190
+ 0.016210075378417968
191
+ ],
192
+ "mean": 0.016144045257568358,
193
+ "median": 0.016185037612915038,
194
+ "std": 8.396025689596366e-05,
195
+ "min": 0.016007278442382814,
196
+ "max": 0.016229835510253907,
197
+ "p25": 0.016087999343872072,
198
+ "p75": 0.016210075378417968,
199
+ "p90": 0.01622193145751953,
200
+ "p95": 0.01622588348388672,
201
+ "p99": 0.01622904510498047,
202
+ "unit": "seconds"
203
+ },
204
+ "tokens_per_second": {
205
+ "name": "tokens_per_second",
206
+ "measurements": [
207
+ 35.63687072777346,
208
+ 35.66206812308254,
209
+ 34.901902307053234,
210
+ 34.76249981753082,
211
+ 34.37691093576872
212
+ ],
213
+ "mean": 35.06805038224176,
214
+ "median": 34.901902307053234,
215
+ "std": 0.5049858863135633,
216
+ "min": 34.37691093576872,
217
+ "max": 35.66206812308254,
218
+ "p25": 34.76249981753082,
219
+ "p75": 35.63687072777346,
220
+ "p90": 35.65198916495891,
221
+ "p95": 35.65702864402073,
222
+ "p99": 35.66106022727018,
223
+ "unit": "tokens/sec"
224
+ },
225
+ "time_per_output_token_seconds": {
226
+ "name": "time_per_output_token_seconds",
227
+ "measurements": [
228
+ 0.0280608251953125,
229
+ 0.02804099853515625,
230
+ 0.0286517333984375,
231
+ 0.028766630859375,
232
+ 0.029089291992187502
233
+ ],
234
+ "mean": 0.028521895996093748,
235
+ "median": 0.0286517333984375,
236
+ "std": 0.00041049720388940125,
237
+ "min": 0.02804099853515625,
238
+ "max": 0.029089291992187502,
239
+ "p25": 0.0280608251953125,
240
+ "p75": 0.028766630859375,
241
+ "p90": 0.0289602275390625,
242
+ "p95": 0.029024759765625,
243
+ "p99": 0.029076385546875003,
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": "2025-10-13T16:46:39.263431",
256
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
257
+ "hardware_info": {
258
+ "gpu_name": "unknown",
259
+ "gpu_memory_total_mb": 0,
260
+ "cpu_count": 160,
261
+ "memory_total_mb": 1285717,
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
+ 1.3963861083984375,
290
+ 1.4380765380859375,
291
+ 1.3660684814453126,
292
+ 1.3686383056640625,
293
+ 1.37243115234375
294
+ ],
295
+ "mean": 1.3883201171875,
296
+ "median": 1.37243115234375,
297
+ "std": 0.02711357533299664,
298
+ "min": 1.3660684814453126,
299
+ "max": 1.4380765380859375,
300
+ "p25": 1.3686383056640625,
301
+ "p75": 1.3963861083984375,
302
+ "p90": 1.4214003662109376,
303
+ "p95": 1.4297384521484375,
304
+ "p99": 1.4364089208984374,
305
+ "unit": "seconds"
306
+ },
307
+ "time_to_first_token_seconds": {
308
+ "name": "time_to_first_token_seconds",
309
+ "measurements": [
310
+ 0.01498841667175293,
311
+ 0.015222852706909179,
312
+ 0.014707303047180176,
313
+ 0.014793701171875,
314
+ 0.014913898468017578
315
+ ],
316
+ "mean": 0.01492523441314697,
317
+ "median": 0.014913898468017578,
318
+ "std": 0.00017747801219763451,
319
+ "min": 0.014707303047180176,
320
+ "max": 0.015222852706909179,
321
+ "p25": 0.014793701171875,
322
+ "p75": 0.01498841667175293,
323
+ "p90": 0.01512907829284668,
324
+ "p95": 0.015175965499877929,
325
+ "p99": 0.01521347526550293,
326
+ "unit": "seconds"
327
+ },
328
+ "tokens_per_second": {
329
+ "name": "tokens_per_second",
330
+ "measurements": [
331
+ 71.61343084019461,
332
+ 69.53732805703012,
333
+ 73.20277230479626,
334
+ 73.06532309241487,
335
+ 72.86339998128607
336
+ ],
337
+ "mean": 72.05645085514439,
338
+ "median": 72.86339998128607,
339
+ "std": 1.38023833846717,
340
+ "min": 69.53732805703012,
341
+ "max": 73.20277230479626,
342
+ "p25": 71.61343084019461,
343
+ "p75": 73.06532309241487,
344
+ "p90": 73.1477926198437,
345
+ "p95": 73.17528246231998,
346
+ "p99": 73.197274336301,
347
+ "unit": "tokens/sec"
348
+ },
349
+ "time_per_output_token_seconds": {
350
+ "name": "time_per_output_token_seconds",
351
+ "measurements": [
352
+ 0.013963861083984374,
353
+ 0.014380765380859376,
354
+ 0.013660684814453126,
355
+ 0.013686383056640626,
356
+ 0.0137243115234375
357
+ ],
358
+ "mean": 0.013883201171875002,
359
+ "median": 0.0137243115234375,
360
+ "std": 0.0002711357533299666,
361
+ "min": 0.013660684814453126,
362
+ "max": 0.014380765380859376,
363
+ "p25": 0.013686383056640626,
364
+ "p75": 0.013963861083984374,
365
+ "p90": 0.014214003662109375,
366
+ "p95": 0.014297384521484375,
367
+ "p99": 0.014364089208984375,
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": "2025-10-13T16:46:58.391919",
380
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
381
+ "hardware_info": {
382
+ "gpu_name": "unknown",
383
+ "gpu_memory_total_mb": 0,
384
+ "cpu_count": 160,
385
+ "memory_total_mb": 1285717,
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
+ 1.7653189697265625,
414
+ 1.7578612060546874,
415
+ 1.666230224609375,
416
+ 1.5366112060546875,
417
+ 1.4908321533203126
418
+ ],
419
+ "mean": 1.643370751953125,
420
+ "median": 1.666230224609375,
421
+ "std": 0.11239894314620968,
422
+ "min": 1.4908321533203126,
423
+ "max": 1.7653189697265625,
424
+ "p25": 1.5366112060546875,
425
+ "p75": 1.7578612060546874,
426
+ "p90": 1.7623358642578124,
427
+ "p95": 1.7638274169921875,
428
+ "p99": 1.7650206591796875,
429
+ "unit": "seconds"
430
+ },
431
+ "time_to_first_token_seconds": {
432
+ "name": "time_to_first_token_seconds",
433
+ "measurements": [
434
+ 0.01977033233642578,
435
+ 0.019453741073608397,
436
+ 0.02034576416015625,
437
+ 0.01625735855102539,
438
+ 0.019544620513916016
439
+ ],
440
+ "mean": 0.019074363327026368,
441
+ "median": 0.019544620513916016,
442
+ "std": 0.001442318837494893,
443
+ "min": 0.01625735855102539,
444
+ "max": 0.02034576416015625,
445
+ "p25": 0.019453741073608397,
446
+ "p75": 0.01977033233642578,
447
+ "p90": 0.020115591430664062,
448
+ "p95": 0.020230677795410157,
449
+ "p99": 0.020322746887207033,
450
+ "unit": "seconds"
451
+ },
452
+ "tokens_per_second": {
453
+ "name": "tokens_per_second",
454
+ "measurements": [
455
+ 56.646986587069534,
456
+ 56.88731263626793,
457
+ 60.01571602954426,
458
+ 65.07827068159558,
459
+ 67.07663218645011
460
+ ],
461
+ "mean": 61.14098362418548,
462
+ "median": 60.01571602954426,
463
+ "std": 4.249473064505088,
464
+ "min": 56.646986587069534,
465
+ "max": 67.07663218645011,
466
+ "p25": 56.88731263626793,
467
+ "p75": 65.07827068159558,
468
+ "p90": 66.2772875845083,
469
+ "p95": 66.6769598854792,
470
+ "p99": 66.99669772625593,
471
+ "unit": "tokens/sec"
472
+ },
473
+ "time_per_output_token_seconds": {
474
+ "name": "time_per_output_token_seconds",
475
+ "measurements": [
476
+ 0.017653189697265627,
477
+ 0.017578612060546873,
478
+ 0.01666230224609375,
479
+ 0.015366112060546874,
480
+ 0.014908321533203126
481
+ ],
482
+ "mean": 0.01643370751953125,
483
+ "median": 0.01666230224609375,
484
+ "std": 0.001123989431462097,
485
+ "min": 0.014908321533203126,
486
+ "max": 0.017653189697265627,
487
+ "p25": 0.015366112060546874,
488
+ "p75": 0.017578612060546873,
489
+ "p90": 0.017623358642578125,
490
+ "p95": 0.017638274169921878,
491
+ "p99": 0.017650206591796877,
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": "2025-10-13T16:47:19.936944",
504
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
505
+ "hardware_info": {
506
+ "gpu_name": "unknown",
507
+ "gpu_memory_total_mb": 0,
508
+ "cpu_count": 160,
509
+ "memory_total_mb": 1285717,
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
+ 1.297906005859375,
538
+ 1.039819580078125,
539
+ 1.1393956298828125,
540
+ 1.209964599609375,
541
+ 1.204503173828125
542
+ ],
543
+ "mean": 1.1783177978515627,
544
+ "median": 1.204503173828125,
545
+ "std": 0.0856565525302743,
546
+ "min": 1.039819580078125,
547
+ "max": 1.297906005859375,
548
+ "p25": 1.1393956298828125,
549
+ "p75": 1.209964599609375,
550
+ "p90": 1.262729443359375,
551
+ "p95": 1.280317724609375,
552
+ "p99": 1.294388349609375,
553
+ "unit": "seconds"
554
+ },
555
+ "time_to_first_token_seconds": {
556
+ "name": "time_to_first_token_seconds",
557
+ "measurements": [
558
+ 0.01493561840057373,
559
+ 0.013307167053222655,
560
+ 0.011511198043823242,
561
+ 0.013460644721984863,
562
+ 0.013498442649841309
563
+ ],
564
+ "mean": 0.01334261417388916,
565
+ "median": 0.013460644721984863,
566
+ "std": 0.001089148082701835,
567
+ "min": 0.011511198043823242,
568
+ "max": 0.01493561840057373,
569
+ "p25": 0.013307167053222655,
570
+ "p75": 0.013498442649841309,
571
+ "p90": 0.014360748100280762,
572
+ "p95": 0.014648183250427245,
573
+ "p99": 0.014878131370544433,
574
+ "unit": "seconds"
575
+ },
576
+ "tokens_per_second": {
577
+ "name": "tokens_per_second",
578
+ "measurements": [
579
+ 77.04718180557889,
580
+ 96.17052988412344,
581
+ 87.76582723095494,
582
+ 82.647046064227,
583
+ 83.02178206984895
584
+ ],
585
+ "mean": 85.33047341094665,
586
+ "median": 83.02178206984895,
587
+ "std": 6.396585502022612,
588
+ "min": 77.04718180557889,
589
+ "max": 96.17052988412344,
590
+ "p25": 82.647046064227,
591
+ "p75": 87.76582723095494,
592
+ "p90": 92.80864882285604,
593
+ "p95": 94.48958935348973,
594
+ "p99": 95.8343417779967,
595
+ "unit": "tokens/sec"
596
+ },
597
+ "time_per_output_token_seconds": {
598
+ "name": "time_per_output_token_seconds",
599
+ "measurements": [
600
+ 0.01297906005859375,
601
+ 0.01039819580078125,
602
+ 0.011393956298828126,
603
+ 0.01209964599609375,
604
+ 0.012045031738281252
605
+ ],
606
+ "mean": 0.011783177978515625,
607
+ "median": 0.012045031738281252,
608
+ "std": 0.0008565655253027427,
609
+ "min": 0.01039819580078125,
610
+ "max": 0.01297906005859375,
611
+ "p25": 0.011393956298828126,
612
+ "p75": 0.01209964599609375,
613
+ "p90": 0.01262729443359375,
614
+ "p95": 0.01280317724609375,
615
+ "p99": 0.012943883496093751,
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": "2025-10-13T16:47:38.750325",
628
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
629
+ "hardware_info": {
630
+ "gpu_name": "unknown",
631
+ "gpu_memory_total_mb": 0,
632
+ "cpu_count": 160,
633
+ "memory_total_mb": 1285717,
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
+ 1.4030408935546874,
662
+ 1.391091796875,
663
+ 1.400947509765625,
664
+ 1.3868839111328124,
665
+ 1.388552978515625
666
+ ],
667
+ "mean": 1.39410341796875,
668
+ "median": 1.391091796875,
669
+ "std": 0.006613903250918552,
670
+ "min": 1.3868839111328124,
671
+ "max": 1.4030408935546874,
672
+ "p25": 1.388552978515625,
673
+ "p75": 1.400947509765625,
674
+ "p90": 1.4022035400390624,
675
+ "p95": 1.402622216796875,
676
+ "p99": 1.4029571582031248,
677
+ "unit": "seconds"
678
+ },
679
+ "time_to_first_token_seconds": {
680
+ "name": "time_to_first_token_seconds",
681
+ "measurements": [
682
+ 0.015062058448791504,
683
+ 0.01526289176940918,
684
+ 0.015385890007019043,
685
+ 0.01542836856842041,
686
+ 0.015391451835632324
687
+ ],
688
+ "mean": 0.015306132125854494,
689
+ "median": 0.015385890007019043,
690
+ "std": 0.00013418212148879482,
691
+ "min": 0.015062058448791504,
692
+ "max": 0.01542836856842041,
693
+ "p25": 0.01526289176940918,
694
+ "p75": 0.015391451835632324,
695
+ "p90": 0.015413601875305175,
696
+ "p95": 0.015420985221862792,
697
+ "p99": 0.015426891899108886,
698
+ "unit": "seconds"
699
+ },
700
+ "tokens_per_second": {
701
+ "name": "tokens_per_second",
702
+ "measurements": [
703
+ 71.27376005887047,
704
+ 71.88598209308954,
705
+ 71.38026178920133,
706
+ 72.10408830708808,
707
+ 72.01741780634171
708
+ ],
709
+ "mean": 71.73230201091823,
710
+ "median": 71.88598209308954,
711
+ "std": 0.33980190522423454,
712
+ "min": 71.27376005887047,
713
+ "max": 72.10408830708808,
714
+ "p25": 71.38026178920133,
715
+ "p75": 72.01741780634171,
716
+ "p90": 72.06942010678954,
717
+ "p95": 72.0867542069388,
718
+ "p99": 72.10062148705822,
719
+ "unit": "tokens/sec"
720
+ },
721
+ "time_per_output_token_seconds": {
722
+ "name": "time_per_output_token_seconds",
723
+ "measurements": [
724
+ 0.014030408935546874,
725
+ 0.01391091796875,
726
+ 0.01400947509765625,
727
+ 0.013868839111328125,
728
+ 0.01388552978515625
729
+ ],
730
+ "mean": 0.0139410341796875,
731
+ "median": 0.01391091796875,
732
+ "std": 6.61390325091853e-05,
733
+ "min": 0.013868839111328125,
734
+ "max": 0.014030408935546874,
735
+ "p25": 0.01388552978515625,
736
+ "p75": 0.01400947509765625,
737
+ "p90": 0.014022035400390624,
738
+ "p95": 0.01402622216796875,
739
+ "p99": 0.014029571582031249,
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": "2025-10-13T16:47:57.710135",
752
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
753
+ "hardware_info": {
754
+ "gpu_name": "unknown",
755
+ "gpu_memory_total_mb": 0,
756
+ "cpu_count": 160,
757
+ "memory_total_mb": 1285717,
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
+ 1.6522928466796876,
786
+ 1.69812158203125,
787
+ 1.6752822265625,
788
+ 1.6810009765625,
789
+ 1.659888671875
790
+ ],
791
+ "mean": 1.6733172607421873,
792
+ "median": 1.6752822265625,
793
+ "std": 0.016127665173414,
794
+ "min": 1.6522928466796876,
795
+ "max": 1.69812158203125,
796
+ "p25": 1.659888671875,
797
+ "p75": 1.6810009765625,
798
+ "p90": 1.69127333984375,
799
+ "p95": 1.6946974609375,
800
+ "p99": 1.6974367578125,
801
+ "unit": "seconds"
802
+ },
803
+ "time_to_first_token_seconds": {
804
+ "name": "time_to_first_token_seconds",
805
+ "measurements": [
806
+ 0.017624011993408205,
807
+ 0.018345598220825196,
808
+ 0.017044540405273437,
809
+ 0.017554851531982423,
810
+ 0.017010902404785155
811
+ ],
812
+ "mean": 0.017515980911254884,
813
+ "median": 0.017554851531982423,
814
+ "std": 0.0004855526683241233,
815
+ "min": 0.017010902404785155,
816
+ "max": 0.018345598220825196,
817
+ "p25": 0.017044540405273437,
818
+ "p75": 0.017624011993408205,
819
+ "p90": 0.018056963729858398,
820
+ "p95": 0.018201280975341797,
821
+ "p99": 0.018316734771728516,
822
+ "unit": "seconds"
823
+ },
824
+ "tokens_per_second": {
825
+ "name": "tokens_per_second",
826
+ "measurements": [
827
+ 60.521959046758454,
828
+ 58.888598471484315,
829
+ 59.69143492030552,
830
+ 59.488365202792004,
831
+ 60.245004194793744
832
+ ],
833
+ "mean": 59.76707236722681,
834
+ "median": 59.69143492030552,
835
+ "std": 0.5750457065522424,
836
+ "min": 58.888598471484315,
837
+ "max": 60.521959046758454,
838
+ "p25": 59.488365202792004,
839
+ "p75": 60.245004194793744,
840
+ "p90": 60.41117710597257,
841
+ "p95": 60.46656807636551,
842
+ "p99": 60.51088085267987,
843
+ "unit": "tokens/sec"
844
+ },
845
+ "time_per_output_token_seconds": {
846
+ "name": "time_per_output_token_seconds",
847
+ "measurements": [
848
+ 0.016522928466796875,
849
+ 0.0169812158203125,
850
+ 0.016752822265625,
851
+ 0.016810009765625,
852
+ 0.01659888671875
853
+ ],
854
+ "mean": 0.016733172607421874,
855
+ "median": 0.016752822265625,
856
+ "std": 0.0001612766517341401,
857
+ "min": 0.016522928466796875,
858
+ "max": 0.0169812158203125,
859
+ "p25": 0.01659888671875,
860
+ "p75": 0.016810009765625,
861
+ "p90": 0.0169127333984375,
862
+ "p95": 0.016946974609375,
863
+ "p99": 0.016974367578125,
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": "2025-10-13T16:48:34.616781",
876
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
877
+ "hardware_info": {
878
+ "gpu_name": "unknown",
879
+ "gpu_memory_total_mb": 0,
880
+ "cpu_count": 160,
881
+ "memory_total_mb": 1285717,
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
+ 1.992040771484375,
910
+ 1.982639892578125,
911
+ 2.022150634765625,
912
+ 1.9947305908203126,
913
+ 2.0063343505859375
914
+ ],
915
+ "mean": 1.999579248046875,
916
+ "median": 1.9947305908203126,
917
+ "std": 0.013582185230671328,
918
+ "min": 1.982639892578125,
919
+ "max": 2.022150634765625,
920
+ "p25": 1.992040771484375,
921
+ "p75": 2.0063343505859375,
922
+ "p90": 2.01582412109375,
923
+ "p95": 2.018987377929687,
924
+ "p99": 2.021517983398437,
925
+ "unit": "seconds"
926
+ },
927
+ "time_to_first_token_seconds": {
928
+ "name": "time_to_first_token_seconds",
929
+ "measurements": [
930
+ 0.020558080673217773,
931
+ 0.021703859329223633,
932
+ 0.021735740661621095,
933
+ 0.020920034408569335,
934
+ 0.020917991638183593
935
+ ],
936
+ "mean": 0.021167141342163086,
937
+ "median": 0.020920034408569335,
938
+ "std": 0.00047020499652847676,
939
+ "min": 0.020558080673217773,
940
+ "max": 0.021735740661621095,
941
+ "p25": 0.020917991638183593,
942
+ "p75": 0.021703859329223633,
943
+ "p90": 0.02172298812866211,
944
+ "p95": 0.0217293643951416,
945
+ "p99": 0.021734465408325195,
946
+ "unit": "seconds"
947
+ },
948
+ "tokens_per_second": {
949
+ "name": "tokens_per_second",
950
+ "measurements": [
951
+ 50.19977574328697,
952
+ 50.43780283769285,
953
+ 49.452300081289636,
954
+ 50.13208322978394,
955
+ 49.84214120183688
956
+ ],
957
+ "mean": 50.01282061877806,
958
+ "median": 50.13208322978394,
959
+ "std": 0.3385558777134175,
960
+ "min": 49.452300081289636,
961
+ "max": 50.43780283769285,
962
+ "p25": 49.84214120183688,
963
+ "p75": 50.19977574328697,
964
+ "p90": 50.3425919999305,
965
+ "p95": 50.390197418811674,
966
+ "p99": 50.428281753916615,
967
+ "unit": "tokens/sec"
968
+ },
969
+ "time_per_output_token_seconds": {
970
+ "name": "time_per_output_token_seconds",
971
+ "measurements": [
972
+ 0.01992040771484375,
973
+ 0.01982639892578125,
974
+ 0.020221506347656248,
975
+ 0.019947305908203126,
976
+ 0.020063343505859376
977
+ ],
978
+ "mean": 0.01999579248046875,
979
+ "median": 0.019947305908203126,
980
+ "std": 0.00013582185230671346,
981
+ "min": 0.01982639892578125,
982
+ "max": 0.020221506347656248,
983
+ "p25": 0.01992040771484375,
984
+ "p75": 0.020063343505859376,
985
+ "p90": 0.0201582412109375,
986
+ "p95": 0.020189873779296874,
987
+ "p99": 0.020215179833984374,
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": "2025-10-13T16:49:04.189554",
1000
+ "commit_id": "b84c0b31c674436b076bba818885226a0e0ddecc",
1001
+ "hardware_info": {
1002
+ "gpu_name": "unknown",
1003
+ "gpu_memory_total_mb": 0,
1004
+ "cpu_count": 160,
1005
+ "memory_total_mb": 1285717,
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
+ 1.6099449462890625,
1034
+ 1.610724365234375,
1035
+ 1.6092164306640624,
1036
+ 1.6082889404296874,
1037
+ 1.4541551513671875
1038
+ ],
1039
+ "mean": 1.578465966796875,
1040
+ "median": 1.6092164306640624,
1041
+ "std": 0.06216061459922433,
1042
+ "min": 1.4541551513671875,
1043
+ "max": 1.610724365234375,
1044
+ "p25": 1.6082889404296874,
1045
+ "p75": 1.6099449462890625,
1046
+ "p90": 1.61041259765625,
1047
+ "p95": 1.6105684814453125,
1048
+ "p99": 1.6106931884765625,
1049
+ "unit": "seconds"
1050
+ },
1051
+ "time_to_first_token_seconds": {
1052
+ "name": "time_to_first_token_seconds",
1053
+ "measurements": [
1054
+ 0.01663454818725586,
1055
+ 0.016967382431030274,
1056
+ 0.016576028823852538,
1057
+ 0.017461572647094727,
1058
+ 0.016870744705200197
1059
+ ],
1060
+ "mean": 0.016902055358886718,
1061
+ "median": 0.016870744705200197,
1062
+ "std": 0.00031501141808293413,
1063
+ "min": 0.016576028823852538,
1064
+ "max": 0.017461572647094727,
1065
+ "p25": 0.01663454818725586,
1066
+ "p75": 0.016967382431030274,
1067
+ "p90": 0.017263896560668946,
1068
+ "p95": 0.017362734603881835,
1069
+ "p99": 0.01744180503845215,
1070
+ "unit": "seconds"
1071
+ },
1072
+ "tokens_per_second": {
1073
+ "name": "tokens_per_second",
1074
+ "measurements": [
1075
+ 62.11392521868583,
1076
+ 62.08386869807429,
1077
+ 62.142045093793755,
1078
+ 62.177882024907134,
1079
+ 68.76845287518366
1080
+ ],
1081
+ "mean": 63.457234782128936,
1082
+ "median": 62.142045093793755,
1083
+ "std": 2.655790654768381,
1084
+ "min": 62.08386869807429,
1085
+ "max": 68.76845287518366,
1086
+ "p25": 62.11392521868583,
1087
+ "p75": 62.177882024907134,
1088
+ "p90": 66.13222453507305,
1089
+ "p95": 67.45033870512835,
1090
+ "p99": 68.5048300411726,
1091
+ "unit": "tokens/sec"
1092
+ },
1093
+ "time_per_output_token_seconds": {
1094
+ "name": "time_per_output_token_seconds",
1095
+ "measurements": [
1096
+ 0.016099449462890623,
1097
+ 0.01610724365234375,
1098
+ 0.016092164306640625,
1099
+ 0.016082889404296873,
1100
+ 0.014541551513671876
1101
+ ],
1102
+ "mean": 0.01578465966796875,
1103
+ "median": 0.016092164306640625,
1104
+ "std": 0.000621606145992243,
1105
+ "min": 0.014541551513671876,
1106
+ "max": 0.01610724365234375,
1107
+ "p25": 0.016082889404296873,
1108
+ "p75": 0.016099449462890623,
1109
+ "p90": 0.016104125976562498,
1110
+ "p95": 0.016105684814453125,
1111
+ "p99": 0.016106931884765623,
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
+ }