hf-transformers-bot commited on
Commit
19f9be8
·
verified ·
1 Parent(s): 562683c

Upload benchmark results for run 18137006858

Browse files
2025-09-30/18137006858/benchmark_results/Llama-2-7b-hf/Llama-2-7b-hf_benchmark_20250930_164857.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": "2025-09-30T16:45:03.415241",
8
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 1.8427708740234374,
42
+ 2.485162353515625,
43
+ 2.66862841796875,
44
+ 2.649970947265625,
45
+ 2.317434814453125
46
+ ],
47
+ "mean": 2.3927934814453122,
48
+ "median": 2.485162353515625,
49
+ "std": 0.3029722650759261,
50
+ "min": 1.8427708740234374,
51
+ "max": 2.66862841796875,
52
+ "p25": 2.317434814453125,
53
+ "p75": 2.649970947265625,
54
+ "p90": 2.6611654296875,
55
+ "p95": 2.664896923828125,
56
+ "p99": 2.667882119140625,
57
+ "unit": "seconds"
58
+ },
59
+ "time_to_first_token_seconds": {
60
+ "name": "time_to_first_token_seconds",
61
+ "measurements": [
62
+ 0.017335275650024413,
63
+ 0.025133834838867188,
64
+ 0.0318666877746582,
65
+ 0.031823768615722656,
66
+ 0.0278397216796875
67
+ ],
68
+ "mean": 0.02679985771179199,
69
+ "median": 0.0278397216796875,
70
+ "std": 0.005373038442917878,
71
+ "min": 0.017335275650024413,
72
+ "max": 0.0318666877746582,
73
+ "p25": 0.025133834838867188,
74
+ "p75": 0.031823768615722656,
75
+ "p90": 0.03184952011108398,
76
+ "p95": 0.03185810394287109,
77
+ "p99": 0.03186497100830078,
78
+ "unit": "seconds"
79
+ },
80
+ "tokens_per_second": {
81
+ "name": "tokens_per_second",
82
+ "measurements": [
83
+ 54.266106226035426,
84
+ 40.238819752977264,
85
+ 37.4724331520519,
86
+ 37.7362627704221,
87
+ 43.15115979803656
88
+ ],
89
+ "mean": 42.57295633990465,
90
+ "median": 40.238819752977264,
91
+ "std": 6.195501732383686,
92
+ "min": 37.4724331520519,
93
+ "max": 54.266106226035426,
94
+ "p25": 37.7362627704221,
95
+ "p75": 43.15115979803656,
96
+ "p90": 49.82012765483588,
97
+ "p95": 52.04311694043565,
98
+ "p99": 53.821508368915474,
99
+ "unit": "tokens/sec"
100
+ },
101
+ "time_per_output_token_seconds": {
102
+ "name": "time_per_output_token_seconds",
103
+ "measurements": [
104
+ 0.018427708740234375,
105
+ 0.024851623535156252,
106
+ 0.0266862841796875,
107
+ 0.02649970947265625,
108
+ 0.02317434814453125
109
+ ],
110
+ "mean": 0.023927934814453124,
111
+ "median": 0.024851623535156252,
112
+ "std": 0.0030297226507592606,
113
+ "min": 0.018427708740234375,
114
+ "max": 0.0266862841796875,
115
+ "p25": 0.02317434814453125,
116
+ "p75": 0.02649970947265625,
117
+ "p90": 0.026611654296875,
118
+ "p95": 0.02664896923828125,
119
+ "p99": 0.02667882119140625,
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": "2025-09-30T16:45:27.125756",
132
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 3.661329833984375,
166
+ 3.230316650390625,
167
+ 3.4840419921875,
168
+ 4.2126806640625,
169
+ 4.12465869140625
170
+ ],
171
+ "mean": 3.74260556640625,
172
+ "median": 3.661329833984375,
173
+ "std": 0.3749229221652764,
174
+ "min": 3.230316650390625,
175
+ "max": 4.2126806640625,
176
+ "p25": 3.4840419921875,
177
+ "p75": 4.12465869140625,
178
+ "p90": 4.177471875,
179
+ "p95": 4.19507626953125,
180
+ "p99": 4.20915978515625,
181
+ "unit": "seconds"
182
+ },
183
+ "time_to_first_token_seconds": {
184
+ "name": "time_to_first_token_seconds",
185
+ "measurements": [
186
+ 0.025784397125244142,
187
+ 0.018329458236694338,
188
+ 0.020504295349121095,
189
+ 0.01940707778930664,
190
+ 0.02257065963745117
191
+ ],
192
+ "mean": 0.021319177627563475,
193
+ "median": 0.020504295349121095,
194
+ "std": 0.0026367773261639035,
195
+ "min": 0.018329458236694338,
196
+ "max": 0.025784397125244142,
197
+ "p25": 0.01940707778930664,
198
+ "p75": 0.02257065963745117,
199
+ "p90": 0.024498902130126954,
200
+ "p95": 0.025141649627685548,
201
+ "p99": 0.025655847625732423,
202
+ "unit": "seconds"
203
+ },
204
+ "tokens_per_second": {
205
+ "name": "tokens_per_second",
206
+ "measurements": [
207
+ 27.312480583366845,
208
+ 30.956717505668532,
209
+ 28.70229469800785,
210
+ 23.737854343691215,
211
+ 24.244430262399792
212
+ ],
213
+ "mean": 26.990755478626845,
214
+ "median": 27.312480583366845,
215
+ "std": 2.7160779196991385,
216
+ "min": 23.737854343691215,
217
+ "max": 30.956717505668532,
218
+ "p25": 24.244430262399792,
219
+ "p75": 28.70229469800785,
220
+ "p90": 30.05494838260426,
221
+ "p95": 30.505832944136394,
222
+ "p99": 30.866540593362103,
223
+ "unit": "tokens/sec"
224
+ },
225
+ "time_per_output_token_seconds": {
226
+ "name": "time_per_output_token_seconds",
227
+ "measurements": [
228
+ 0.03661329833984375,
229
+ 0.03230316650390625,
230
+ 0.034840419921875004,
231
+ 0.042126806640625,
232
+ 0.041246586914062494
233
+ ],
234
+ "mean": 0.0374260556640625,
235
+ "median": 0.03661329833984375,
236
+ "std": 0.003749229221652763,
237
+ "min": 0.03230316650390625,
238
+ "max": 0.042126806640625,
239
+ "p25": 0.034840419921875004,
240
+ "p75": 0.041246586914062494,
241
+ "p90": 0.041774718749999995,
242
+ "p95": 0.0419507626953125,
243
+ "p99": 0.0420915978515625,
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": "2025-09-30T16:46:19.858181",
256
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 2.073455322265625,
290
+ 1.7209776611328125,
291
+ 1.94188916015625,
292
+ 1.9943028564453125,
293
+ 1.87789111328125
294
+ ],
295
+ "mean": 1.9217032226562498,
296
+ "median": 1.94188916015625,
297
+ "std": 0.11909419012840199,
298
+ "min": 1.7209776611328125,
299
+ "max": 2.073455322265625,
300
+ "p25": 1.87789111328125,
301
+ "p75": 1.9943028564453125,
302
+ "p90": 2.0417943359375,
303
+ "p95": 2.057624829101562,
304
+ "p99": 2.0702892236328125,
305
+ "unit": "seconds"
306
+ },
307
+ "time_to_first_token_seconds": {
308
+ "name": "time_to_first_token_seconds",
309
+ "measurements": [
310
+ 0.020682775497436523,
311
+ 0.01753339195251465,
312
+ 0.02114176559448242,
313
+ 0.019893510818481445,
314
+ 0.02112776756286621
315
+ ],
316
+ "mean": 0.02007584228515625,
317
+ "median": 0.020682775497436523,
318
+ "std": 0.00134969668245188,
319
+ "min": 0.01753339195251465,
320
+ "max": 0.02114176559448242,
321
+ "p25": 0.019893510818481445,
322
+ "p75": 0.02112776756286621,
323
+ "p90": 0.021136166381835936,
324
+ "p95": 0.02113896598815918,
325
+ "p99": 0.021141205673217773,
326
+ "unit": "seconds"
327
+ },
328
+ "tokens_per_second": {
329
+ "name": "tokens_per_second",
330
+ "measurements": [
331
+ 48.22867361845633,
332
+ 58.106506701647845,
333
+ 51.49624502355928,
334
+ 50.14283546594428,
335
+ 53.25122382909061
336
+ ],
337
+ "mean": 52.24509692773967,
338
+ "median": 51.49624502355928,
339
+ "std": 3.3609627938073805,
340
+ "min": 48.22867361845633,
341
+ "max": 58.106506701647845,
342
+ "p25": 50.14283546594428,
343
+ "p75": 53.25122382909061,
344
+ "p90": 56.16439355262495,
345
+ "p95": 57.135450127136394,
346
+ "p99": 57.912295386745555,
347
+ "unit": "tokens/sec"
348
+ },
349
+ "time_per_output_token_seconds": {
350
+ "name": "time_per_output_token_seconds",
351
+ "measurements": [
352
+ 0.02073455322265625,
353
+ 0.017209776611328123,
354
+ 0.019418891601562502,
355
+ 0.019943028564453125,
356
+ 0.0187789111328125
357
+ ],
358
+ "mean": 0.019217032226562496,
359
+ "median": 0.019418891601562502,
360
+ "std": 0.0011909419012840208,
361
+ "min": 0.017209776611328123,
362
+ "max": 0.02073455322265625,
363
+ "p25": 0.0187789111328125,
364
+ "p75": 0.019943028564453125,
365
+ "p90": 0.020417943359375,
366
+ "p95": 0.020576248291015627,
367
+ "p99": 0.020702892236328124,
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": "2025-09-30T16:46:43.305678",
380
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 2.43696435546875,
414
+ 2.415430419921875,
415
+ 2.273930419921875,
416
+ 2.350267333984375,
417
+ 2.308427734375
418
+ ],
419
+ "mean": 2.357004052734375,
420
+ "median": 2.350267333984375,
421
+ "std": 0.06182808576955217,
422
+ "min": 2.273930419921875,
423
+ "max": 2.43696435546875,
424
+ "p25": 2.308427734375,
425
+ "p75": 2.415430419921875,
426
+ "p90": 2.42835078125,
427
+ "p95": 2.432657568359375,
428
+ "p99": 2.436102998046875,
429
+ "unit": "seconds"
430
+ },
431
+ "time_to_first_token_seconds": {
432
+ "name": "time_to_first_token_seconds",
433
+ "measurements": [
434
+ 0.02654995346069336,
435
+ 0.02732185173034668,
436
+ 0.02729901123046875,
437
+ 0.024365665435791016,
438
+ 0.02632731056213379
439
+ ],
440
+ "mean": 0.026372758483886715,
441
+ "median": 0.02654995346069336,
442
+ "std": 0.0010789446897907168,
443
+ "min": 0.024365665435791016,
444
+ "max": 0.02732185173034668,
445
+ "p25": 0.02632731056213379,
446
+ "p75": 0.02729901123046875,
447
+ "p90": 0.027312715530395507,
448
+ "p95": 0.027317283630371093,
449
+ "p99": 0.027320938110351564,
450
+ "unit": "seconds"
451
+ },
452
+ "tokens_per_second": {
453
+ "name": "tokens_per_second",
454
+ "measurements": [
455
+ 41.03465845759775,
456
+ 41.40048878047765,
457
+ 43.97671939471027,
458
+ 42.54835122542056,
459
+ 43.31952805404788
460
+ ],
461
+ "mean": 42.45594918245082,
462
+ "median": 42.54835122542056,
463
+ "std": 1.1136527850453408,
464
+ "min": 41.03465845759775,
465
+ "max": 43.97671939471027,
466
+ "p25": 41.40048878047765,
467
+ "p75": 43.31952805404788,
468
+ "p90": 43.71384285844532,
469
+ "p95": 43.845281126577795,
470
+ "p99": 43.950431741083776,
471
+ "unit": "tokens/sec"
472
+ },
473
+ "time_per_output_token_seconds": {
474
+ "name": "time_per_output_token_seconds",
475
+ "measurements": [
476
+ 0.0243696435546875,
477
+ 0.024154304199218747,
478
+ 0.02273930419921875,
479
+ 0.02350267333984375,
480
+ 0.02308427734375
481
+ ],
482
+ "mean": 0.02357004052734375,
483
+ "median": 0.02350267333984375,
484
+ "std": 0.0006182808576955211,
485
+ "min": 0.02273930419921875,
486
+ "max": 0.0243696435546875,
487
+ "p25": 0.02308427734375,
488
+ "p75": 0.024154304199218747,
489
+ "p90": 0.0242835078125,
490
+ "p95": 0.02432657568359375,
491
+ "p99": 0.02436102998046875,
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": "2025-09-30T16:47:08.379815",
504
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 1.68916748046875,
538
+ 1.4298231201171876,
539
+ 1.350528564453125,
540
+ 1.3439833984375,
541
+ 1.3375166015625
542
+ ],
543
+ "mean": 1.4302038330078126,
544
+ "median": 1.350528564453125,
545
+ "std": 0.13374255140255836,
546
+ "min": 1.3375166015625,
547
+ "max": 1.68916748046875,
548
+ "p25": 1.3439833984375,
549
+ "p75": 1.4298231201171876,
550
+ "p90": 1.5854297363281251,
551
+ "p95": 1.6372986083984375,
552
+ "p99": 1.6787937060546876,
553
+ "unit": "seconds"
554
+ },
555
+ "time_to_first_token_seconds": {
556
+ "name": "time_to_first_token_seconds",
557
+ "measurements": [
558
+ 0.020855411529541016,
559
+ 0.015023483276367187,
560
+ 0.015214675903320313,
561
+ 0.014908242225646972,
562
+ 0.015557709693908692
563
+ ],
564
+ "mean": 0.016311904525756836,
565
+ "median": 0.015214675903320313,
566
+ "std": 0.0022823879405721014,
567
+ "min": 0.014908242225646972,
568
+ "max": 0.020855411529541016,
569
+ "p25": 0.015023483276367187,
570
+ "p75": 0.015557709693908692,
571
+ "p90": 0.018736330795288088,
572
+ "p95": 0.01979587116241455,
573
+ "p99": 0.020643503456115722,
574
+ "unit": "seconds"
575
+ },
576
+ "tokens_per_second": {
577
+ "name": "tokens_per_second",
578
+ "measurements": [
579
+ 59.20076082227776,
580
+ 69.93872080611206,
581
+ 74.04508326004448,
582
+ 74.40568099000247,
583
+ 74.76542712305704
584
+ ],
585
+ "mean": 70.47113460029875,
586
+ "median": 74.04508326004448,
587
+ "std": 5.899144041821492,
588
+ "min": 59.20076082227776,
589
+ "max": 74.76542712305704,
590
+ "p25": 69.93872080611206,
591
+ "p75": 74.40568099000247,
592
+ "p90": 74.62152866983521,
593
+ "p95": 74.69347789644613,
594
+ "p99": 74.75103727773485,
595
+ "unit": "tokens/sec"
596
+ },
597
+ "time_per_output_token_seconds": {
598
+ "name": "time_per_output_token_seconds",
599
+ "measurements": [
600
+ 0.0168916748046875,
601
+ 0.014298231201171876,
602
+ 0.01350528564453125,
603
+ 0.013439833984375,
604
+ 0.013375166015625
605
+ ],
606
+ "mean": 0.014302038330078124,
607
+ "median": 0.01350528564453125,
608
+ "std": 0.0013374255140255835,
609
+ "min": 0.013375166015625,
610
+ "max": 0.0168916748046875,
611
+ "p25": 0.013439833984375,
612
+ "p75": 0.014298231201171876,
613
+ "p90": 0.01585429736328125,
614
+ "p95": 0.016372986083984374,
615
+ "p99": 0.016787937060546876,
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": "2025-09-30T16:47:27.395794",
628
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 1.222987548828125,
662
+ 1.0896163330078126,
663
+ 1.2504635009765626,
664
+ 1.3217215576171875,
665
+ 1.3213828125
666
+ ],
667
+ "mean": 1.2412343505859376,
668
+ "median": 1.2504635009765626,
669
+ "std": 0.08521515438995989,
670
+ "min": 1.0896163330078126,
671
+ "max": 1.3217215576171875,
672
+ "p25": 1.222987548828125,
673
+ "p75": 1.3213828125,
674
+ "p90": 1.3215860595703124,
675
+ "p95": 1.32165380859375,
676
+ "p99": 1.3217080078124999,
677
+ "unit": "seconds"
678
+ },
679
+ "time_to_first_token_seconds": {
680
+ "name": "time_to_first_token_seconds",
681
+ "measurements": [
682
+ 0.013826903343200683,
683
+ 0.014057819366455077,
684
+ 0.011890419006347656,
685
+ 0.011616104125976563,
686
+ 0.014308053970336914
687
+ ],
688
+ "mean": 0.01313985996246338,
689
+ "median": 0.013826903343200683,
690
+ "std": 0.001145625581633928,
691
+ "min": 0.011616104125976563,
692
+ "max": 0.014308053970336914,
693
+ "p25": 0.011890419006347656,
694
+ "p75": 0.014057819366455077,
695
+ "p90": 0.01420796012878418,
696
+ "p95": 0.014258007049560546,
697
+ "p99": 0.01429804458618164,
698
+ "unit": "seconds"
699
+ },
700
+ "tokens_per_second": {
701
+ "name": "tokens_per_second",
702
+ "measurements": [
703
+ 81.76698127124898,
704
+ 91.77542311976613,
705
+ 79.97034693288045,
706
+ 75.65890063886147,
707
+ 75.67829629235472
708
+ ],
709
+ "mean": 80.96998965102235,
710
+ "median": 79.97034693288045,
711
+ "std": 5.909341058889003,
712
+ "min": 75.65890063886147,
713
+ "max": 91.77542311976613,
714
+ "p25": 75.67829629235472,
715
+ "p75": 81.76698127124898,
716
+ "p90": 87.77204638035927,
717
+ "p95": 89.7737347500627,
718
+ "p99": 91.37508544582545,
719
+ "unit": "tokens/sec"
720
+ },
721
+ "time_per_output_token_seconds": {
722
+ "name": "time_per_output_token_seconds",
723
+ "measurements": [
724
+ 0.01222987548828125,
725
+ 0.010896163330078125,
726
+ 0.012504635009765626,
727
+ 0.013217215576171875,
728
+ 0.013213828125
729
+ ],
730
+ "mean": 0.012412343505859376,
731
+ "median": 0.012504635009765626,
732
+ "std": 0.0008521515438995994,
733
+ "min": 0.010896163330078125,
734
+ "max": 0.013217215576171875,
735
+ "p25": 0.01222987548828125,
736
+ "p75": 0.013213828125,
737
+ "p90": 0.013215860595703125,
738
+ "p95": 0.0132165380859375,
739
+ "p99": 0.013217080078125,
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": "2025-09-30T16:47:44.235835",
752
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 1.6860185546875,
786
+ 1.7887880859375,
787
+ 1.9566868896484375,
788
+ 1.9313388671875,
789
+ 1.7233868408203126
790
+ ],
791
+ "mean": 1.8172438476562498,
792
+ "median": 1.7887880859375,
793
+ "std": 0.10890452797179835,
794
+ "min": 1.6860185546875,
795
+ "max": 1.9566868896484375,
796
+ "p25": 1.7233868408203126,
797
+ "p75": 1.9313388671875,
798
+ "p90": 1.9465476806640625,
799
+ "p95": 1.95161728515625,
800
+ "p99": 1.95567296875,
801
+ "unit": "seconds"
802
+ },
803
+ "time_to_first_token_seconds": {
804
+ "name": "time_to_first_token_seconds",
805
+ "measurements": [
806
+ 0.017306955337524414,
807
+ 0.017060441970825194,
808
+ 0.020885492324829103,
809
+ 0.0195371150970459,
810
+ 0.016946044921875
811
+ ],
812
+ "mean": 0.018347209930419918,
813
+ "median": 0.017306955337524414,
814
+ "std": 0.0015849237584427379,
815
+ "min": 0.016946044921875,
816
+ "max": 0.020885492324829103,
817
+ "p25": 0.017060441970825194,
818
+ "p75": 0.0195371150970459,
819
+ "p90": 0.020346141433715822,
820
+ "p95": 0.020615816879272462,
821
+ "p99": 0.020831557235717776,
822
+ "unit": "seconds"
823
+ },
824
+ "tokens_per_second": {
825
+ "name": "tokens_per_second",
826
+ "measurements": [
827
+ 59.31132828993972,
828
+ 55.903771266225874,
829
+ 51.10679717283087,
830
+ 51.77755271172291,
831
+ 58.025277686582044
832
+ ],
833
+ "mean": 55.224945425460284,
834
+ "median": 55.903771266225874,
835
+ "std": 3.2816109386472307,
836
+ "min": 51.10679717283087,
837
+ "max": 59.31132828993972,
838
+ "p25": 51.77755271172291,
839
+ "p75": 58.025277686582044,
840
+ "p90": 58.79690804859665,
841
+ "p95": 59.054118169268186,
842
+ "p99": 59.25988626580541,
843
+ "unit": "tokens/sec"
844
+ },
845
+ "time_per_output_token_seconds": {
846
+ "name": "time_per_output_token_seconds",
847
+ "measurements": [
848
+ 0.016860185546875,
849
+ 0.017887880859375,
850
+ 0.019566868896484374,
851
+ 0.019313388671875,
852
+ 0.017233868408203127
853
+ ],
854
+ "mean": 0.0181724384765625,
855
+ "median": 0.017887880859375,
856
+ "std": 0.001089045279717983,
857
+ "min": 0.016860185546875,
858
+ "max": 0.019566868896484374,
859
+ "p25": 0.017233868408203127,
860
+ "p75": 0.019313388671875,
861
+ "p90": 0.019465476806640625,
862
+ "p95": 0.0195161728515625,
863
+ "p99": 0.0195567296875,
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": "2025-09-30T16:48:04.805570",
876
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 1.9962694091796875,
910
+ 1.9163814697265624,
911
+ 2.242566162109375,
912
+ 2.282380126953125,
913
+ 2.317659423828125
914
+ ],
915
+ "mean": 2.151051318359375,
916
+ "median": 2.242566162109375,
917
+ "std": 0.1627315953758108,
918
+ "min": 1.9163814697265624,
919
+ "max": 2.317659423828125,
920
+ "p25": 1.9962694091796875,
921
+ "p75": 2.282380126953125,
922
+ "p90": 2.3035477050781252,
923
+ "p95": 2.310603564453125,
924
+ "p99": 2.316248251953125,
925
+ "unit": "seconds"
926
+ },
927
+ "time_to_first_token_seconds": {
928
+ "name": "time_to_first_token_seconds",
929
+ "measurements": [
930
+ 0.01912412452697754,
931
+ 0.023650800704956055,
932
+ 0.01922660255432129,
933
+ 0.021211803436279295,
934
+ 0.02663654327392578
935
+ ],
936
+ "mean": 0.02196997489929199,
937
+ "median": 0.021211803436279295,
938
+ "std": 0.0028566349404919645,
939
+ "min": 0.01912412452697754,
940
+ "max": 0.02663654327392578,
941
+ "p25": 0.01922660255432129,
942
+ "p75": 0.023650800704956055,
943
+ "p90": 0.02544224624633789,
944
+ "p95": 0.026039394760131833,
945
+ "p99": 0.02651711357116699,
946
+ "unit": "seconds"
947
+ },
948
+ "tokens_per_second": {
949
+ "name": "tokens_per_second",
950
+ "measurements": [
951
+ 50.09343906196122,
952
+ 52.18167759379787,
953
+ 44.59177244783683,
954
+ 43.81391110931881,
955
+ 43.146977926044016
956
+ ],
957
+ "mean": 46.765555627791755,
958
+ "median": 44.59177244783683,
959
+ "std": 3.6589842460045054,
960
+ "min": 43.146977926044016,
961
+ "max": 52.18167759379787,
962
+ "p25": 43.81391110931881,
963
+ "p75": 50.09343906196122,
964
+ "p90": 51.34638218106321,
965
+ "p95": 51.76402988743054,
966
+ "p99": 52.09814805252441,
967
+ "unit": "tokens/sec"
968
+ },
969
+ "time_per_output_token_seconds": {
970
+ "name": "time_per_output_token_seconds",
971
+ "measurements": [
972
+ 0.019962694091796875,
973
+ 0.019163814697265625,
974
+ 0.02242566162109375,
975
+ 0.022823801269531253,
976
+ 0.02317659423828125
977
+ ],
978
+ "mean": 0.02151051318359375,
979
+ "median": 0.02242566162109375,
980
+ "std": 0.0016273159537581086,
981
+ "min": 0.019163814697265625,
982
+ "max": 0.02317659423828125,
983
+ "p25": 0.019962694091796875,
984
+ "p75": 0.022823801269531253,
985
+ "p90": 0.023035477050781253,
986
+ "p95": 0.023106035644531252,
987
+ "p99": 0.02316248251953125,
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": "2025-09-30T16:48:35.498835",
1000
+ "commit_id": "16a141765c3f44a77359482c54d2c6506847e834",
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
+ 2.21717236328125,
1034
+ 2.26595654296875,
1035
+ 2.116620361328125,
1036
+ 2.89249951171875,
1037
+ 3.493482177734375
1038
+ ],
1039
+ "mean": 2.5971461914062504,
1040
+ "median": 2.26595654296875,
1041
+ "std": 0.5245233170929247,
1042
+ "min": 2.116620361328125,
1043
+ "max": 3.493482177734375,
1044
+ "p25": 2.21717236328125,
1045
+ "p75": 2.89249951171875,
1046
+ "p90": 3.253089111328125,
1047
+ "p95": 3.37328564453125,
1048
+ "p99": 3.46944287109375,
1049
+ "unit": "seconds"
1050
+ },
1051
+ "time_to_first_token_seconds": {
1052
+ "name": "time_to_first_token_seconds",
1053
+ "measurements": [
1054
+ 0.024572307586669923,
1055
+ 0.024493383407592773,
1056
+ 0.02217302894592285,
1057
+ 0.025545244216918946,
1058
+ 0.033832527160644534
1059
+ ],
1060
+ "mean": 0.026123298263549803,
1061
+ "median": 0.024572307586669923,
1062
+ "std": 0.004010814851569426,
1063
+ "min": 0.02217302894592285,
1064
+ "max": 0.033832527160644534,
1065
+ "p25": 0.024493383407592773,
1066
+ "p75": 0.025545244216918946,
1067
+ "p90": 0.0305176139831543,
1068
+ "p95": 0.03217507057189942,
1069
+ "p99": 0.03350103584289551,
1070
+ "unit": "seconds"
1071
+ },
1072
+ "tokens_per_second": {
1073
+ "name": "tokens_per_second",
1074
+ "measurements": [
1075
+ 45.1024925513718,
1076
+ 44.13147300211887,
1077
+ 47.24512804802301,
1078
+ 34.57217523973896,
1079
+ 28.624734552060293
1080
+ ],
1081
+ "mean": 39.93520067866258,
1082
+ "median": 44.13147300211887,
1083
+ "std": 7.1334890581714125,
1084
+ "min": 28.624734552060293,
1085
+ "max": 47.24512804802301,
1086
+ "p25": 34.57217523973896,
1087
+ "p75": 45.1024925513718,
1088
+ "p90": 46.38807384936253,
1089
+ "p95": 46.81660094869277,
1090
+ "p99": 47.15942262815697,
1091
+ "unit": "tokens/sec"
1092
+ },
1093
+ "time_per_output_token_seconds": {
1094
+ "name": "time_per_output_token_seconds",
1095
+ "measurements": [
1096
+ 0.0221717236328125,
1097
+ 0.0226595654296875,
1098
+ 0.02116620361328125,
1099
+ 0.0289249951171875,
1100
+ 0.03493482177734375
1101
+ ],
1102
+ "mean": 0.0259714619140625,
1103
+ "median": 0.0226595654296875,
1104
+ "std": 0.005245233170929247,
1105
+ "min": 0.02116620361328125,
1106
+ "max": 0.03493482177734375,
1107
+ "p25": 0.0221717236328125,
1108
+ "p75": 0.0289249951171875,
1109
+ "p90": 0.03253089111328125,
1110
+ "p95": 0.0337328564453125,
1111
+ "p99": 0.0346944287109375,
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
+ }