hf-transformers-bot commited on
Commit
9183ca4
·
verified ·
1 Parent(s): a394c90

Upload benchmark results for run 18287926826

Browse files
2025-10-06/18287926826/benchmark_results/Llama-2-7b-hf/Llama-2-7b-hf_benchmark_20251006_165028.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": "2025-10-06T16:47:21.098862",
8
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
9
+ "hardware_info": {
10
+ "gpu_name": "unknown",
11
+ "gpu_memory_total_mb": 0,
12
+ "cpu_count": 160,
13
+ "memory_total_mb": 1290069,
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
+ 1.660790771484375,
42
+ 1.44341064453125,
43
+ 1.38679638671875,
44
+ 1.3855726318359376,
45
+ 1.5558170166015626
46
+ ],
47
+ "mean": 1.4864774902343751,
48
+ "median": 1.44341064453125,
49
+ "std": 0.10692610156765815,
50
+ "min": 1.3855726318359376,
51
+ "max": 1.660790771484375,
52
+ "p25": 1.38679638671875,
53
+ "p75": 1.5558170166015626,
54
+ "p90": 1.6188012695312501,
55
+ "p95": 1.6397960205078126,
56
+ "p99": 1.6565918212890625,
57
+ "unit": "seconds"
58
+ },
59
+ "time_to_first_token_seconds": {
60
+ "name": "time_to_first_token_seconds",
61
+ "measurements": [
62
+ 0.017014352798461915,
63
+ 0.01779473114013672,
64
+ 0.015285110473632813,
65
+ 0.015057990074157715,
66
+ 0.014874834060668946
67
+ ],
68
+ "mean": 0.01600540370941162,
69
+ "median": 0.015285110473632813,
70
+ "std": 0.0011759484334844417,
71
+ "min": 0.014874834060668946,
72
+ "max": 0.01779473114013672,
73
+ "p25": 0.015057990074157715,
74
+ "p75": 0.017014352798461915,
75
+ "p90": 0.017482579803466798,
76
+ "p95": 0.017638655471801757,
77
+ "p99": 0.017763516006469728,
78
+ "unit": "seconds"
79
+ },
80
+ "tokens_per_second": {
81
+ "name": "tokens_per_second",
82
+ "measurements": [
83
+ 60.21228062980046,
84
+ 69.28035370868085,
85
+ 72.10863898816932,
86
+ 72.17232622983907,
87
+ 64.27491082366117
88
+ ],
89
+ "mean": 67.60970207603017,
90
+ "median": 69.28035370868085,
91
+ "std": 4.6838387277109135,
92
+ "min": 60.21228062980046,
93
+ "max": 72.17232622983907,
94
+ "p25": 64.27491082366117,
95
+ "p75": 72.10863898816932,
96
+ "p90": 72.14685133317117,
97
+ "p95": 72.15958878150512,
98
+ "p99": 72.16977874017228,
99
+ "unit": "tokens/sec"
100
+ },
101
+ "time_per_output_token_seconds": {
102
+ "name": "time_per_output_token_seconds",
103
+ "measurements": [
104
+ 0.016607907714843752,
105
+ 0.0144341064453125,
106
+ 0.013867963867187501,
107
+ 0.013855726318359376,
108
+ 0.015558170166015626
109
+ ],
110
+ "mean": 0.014864774902343753,
111
+ "median": 0.0144341064453125,
112
+ "std": 0.0010692610156765817,
113
+ "min": 0.013855726318359376,
114
+ "max": 0.016607907714843752,
115
+ "p25": 0.013867963867187501,
116
+ "p75": 0.015558170166015626,
117
+ "p90": 0.0161880126953125,
118
+ "p95": 0.016397960205078128,
119
+ "p99": 0.016565918212890628,
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": "2025-10-06T16:47:40.424184",
132
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
133
+ "hardware_info": {
134
+ "gpu_name": "unknown",
135
+ "gpu_memory_total_mb": 0,
136
+ "cpu_count": 160,
137
+ "memory_total_mb": 1290069,
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
+ 3.048676025390625,
166
+ 3.0587001953125,
167
+ 3.08124169921875,
168
+ 3.12808203125,
169
+ 3.162422607421875
170
+ ],
171
+ "mean": 3.0958245117187495,
172
+ "median": 3.08124169921875,
173
+ "std": 0.04310384430573788,
174
+ "min": 3.048676025390625,
175
+ "max": 3.162422607421875,
176
+ "p25": 3.0587001953125,
177
+ "p75": 3.12808203125,
178
+ "p90": 3.148686376953125,
179
+ "p95": 3.1555544921875,
180
+ "p99": 3.161048984375,
181
+ "unit": "seconds"
182
+ },
183
+ "time_to_first_token_seconds": {
184
+ "name": "time_to_first_token_seconds",
185
+ "measurements": [
186
+ 0.017091384887695314,
187
+ 0.01681911087036133,
188
+ 0.01682522964477539,
189
+ 0.01683926773071289,
190
+ 0.01666111373901367
191
+ ],
192
+ "mean": 0.01684722137451172,
193
+ "median": 0.01682522964477539,
194
+ "std": 0.00013826711122381436,
195
+ "min": 0.01666111373901367,
196
+ "max": 0.017091384887695314,
197
+ "p25": 0.01681911087036133,
198
+ "p75": 0.01683926773071289,
199
+ "p90": 0.016990538024902344,
200
+ "p95": 0.01704096145629883,
201
+ "p99": 0.017081300201416016,
202
+ "unit": "seconds"
203
+ },
204
+ "tokens_per_second": {
205
+ "name": "tokens_per_second",
206
+ "measurements": [
207
+ 32.801123886945994,
208
+ 32.693625924257425,
209
+ 32.454448485931835,
210
+ 31.96847109538218,
211
+ 31.621327195584318
212
+ ],
213
+ "mean": 32.30779931762035,
214
+ "median": 32.454448485931835,
215
+ "std": 0.44724152231111824,
216
+ "min": 31.621327195584318,
217
+ "max": 32.801123886945994,
218
+ "p25": 31.96847109538218,
219
+ "p75": 32.693625924257425,
220
+ "p90": 32.758124701870564,
221
+ "p95": 32.77962429440828,
222
+ "p99": 32.79682396843845,
223
+ "unit": "tokens/sec"
224
+ },
225
+ "time_per_output_token_seconds": {
226
+ "name": "time_per_output_token_seconds",
227
+ "measurements": [
228
+ 0.030486760253906252,
229
+ 0.030587001953124998,
230
+ 0.030812416992187498,
231
+ 0.0312808203125,
232
+ 0.03162422607421875
233
+ ],
234
+ "mean": 0.030958245117187498,
235
+ "median": 0.030812416992187498,
236
+ "std": 0.0004310384430573785,
237
+ "min": 0.030486760253906252,
238
+ "max": 0.03162422607421875,
239
+ "p25": 0.030587001953124998,
240
+ "p75": 0.0312808203125,
241
+ "p90": 0.03148686376953125,
242
+ "p95": 0.031555544921875,
243
+ "p99": 0.03161048984375,
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": "2025-10-06T16:48:21.392148",
256
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
257
+ "hardware_info": {
258
+ "gpu_name": "unknown",
259
+ "gpu_memory_total_mb": 0,
260
+ "cpu_count": 160,
261
+ "memory_total_mb": 1290069,
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
+ 1.2696163330078125,
290
+ 1.26688427734375,
291
+ 1.2686103515625,
292
+ 1.279510498046875,
293
+ 1.2729949951171875
294
+ ],
295
+ "mean": 1.271523291015625,
296
+ "median": 1.2696163330078125,
297
+ "std": 0.004463262082190109,
298
+ "min": 1.26688427734375,
299
+ "max": 1.279510498046875,
300
+ "p25": 1.2686103515625,
301
+ "p75": 1.2729949951171875,
302
+ "p90": 1.276904296875,
303
+ "p95": 1.2782073974609376,
304
+ "p99": 1.2792498779296875,
305
+ "unit": "seconds"
306
+ },
307
+ "time_to_first_token_seconds": {
308
+ "name": "time_to_first_token_seconds",
309
+ "measurements": [
310
+ 0.013801494598388671,
311
+ 0.014049090385437012,
312
+ 0.013923731803894043,
313
+ 0.013882253646850586,
314
+ 0.013997930526733399
315
+ ],
316
+ "mean": 0.013930900192260742,
317
+ "median": 0.013923731803894043,
318
+ "std": 8.674752722707646e-05,
319
+ "min": 0.013801494598388671,
320
+ "max": 0.014049090385437012,
321
+ "p25": 0.013882253646850586,
322
+ "p75": 0.013997930526733399,
323
+ "p90": 0.014028626441955567,
324
+ "p95": 0.01403885841369629,
325
+ "p99": 0.014047043991088868,
326
+ "unit": "seconds"
327
+ },
328
+ "tokens_per_second": {
329
+ "name": "tokens_per_second",
330
+ "measurements": [
331
+ 78.76395206975072,
332
+ 78.93380775840704,
333
+ 78.82641023450087,
334
+ 78.154888258163,
335
+ 78.55490428758075
336
+ ],
337
+ "mean": 78.64679252168048,
338
+ "median": 78.76395206975072,
339
+ "std": 0.27523803538344627,
340
+ "min": 78.154888258163,
341
+ "max": 78.93380775840704,
342
+ "p25": 78.55490428758075,
343
+ "p75": 78.82641023450087,
344
+ "p90": 78.89084874884458,
345
+ "p95": 78.9123282536258,
346
+ "p99": 78.92951185745079,
347
+ "unit": "tokens/sec"
348
+ },
349
+ "time_per_output_token_seconds": {
350
+ "name": "time_per_output_token_seconds",
351
+ "measurements": [
352
+ 0.012696163330078124,
353
+ 0.012668842773437501,
354
+ 0.012686103515625,
355
+ 0.01279510498046875,
356
+ 0.012729949951171875
357
+ ],
358
+ "mean": 0.012715232910156249,
359
+ "median": 0.012696163330078124,
360
+ "std": 4.46326208219009e-05,
361
+ "min": 0.012668842773437501,
362
+ "max": 0.01279510498046875,
363
+ "p25": 0.012686103515625,
364
+ "p75": 0.012729949951171875,
365
+ "p90": 0.01276904296875,
366
+ "p95": 0.012782073974609374,
367
+ "p99": 0.012792498779296875,
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": "2025-10-06T16:48:37.203719",
380
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
381
+ "hardware_info": {
382
+ "gpu_name": "unknown",
383
+ "gpu_memory_total_mb": 0,
384
+ "cpu_count": 160,
385
+ "memory_total_mb": 1290069,
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
+ 1.7220289306640626,
414
+ 1.7221925048828124,
415
+ 1.72422119140625,
416
+ 1.725779541015625,
417
+ 1.738879150390625
418
+ ],
419
+ "mean": 1.7266202636718748,
420
+ "median": 1.72422119140625,
421
+ "std": 0.006283757674931698,
422
+ "min": 1.7220289306640626,
423
+ "max": 1.738879150390625,
424
+ "p25": 1.7221925048828124,
425
+ "p75": 1.725779541015625,
426
+ "p90": 1.7336393066406248,
427
+ "p95": 1.7362592285156249,
428
+ "p99": 1.738355166015625,
429
+ "unit": "seconds"
430
+ },
431
+ "time_to_first_token_seconds": {
432
+ "name": "time_to_first_token_seconds",
433
+ "measurements": [
434
+ 0.019119218826293947,
435
+ 0.0191378173828125,
436
+ 0.01924737548828125,
437
+ 0.019010099411010742,
438
+ 0.019249975204467775
439
+ ],
440
+ "mean": 0.019152897262573243,
441
+ "median": 0.0191378173828125,
442
+ "std": 8.955794570008222e-05,
443
+ "min": 0.019010099411010742,
444
+ "max": 0.019249975204467775,
445
+ "p25": 0.019119218826293947,
446
+ "p75": 0.01924737548828125,
447
+ "p90": 0.019248935317993165,
448
+ "p95": 0.01924945526123047,
449
+ "p99": 0.019249871215820315,
450
+ "unit": "seconds"
451
+ },
452
+ "tokens_per_second": {
453
+ "name": "tokens_per_second",
454
+ "measurements": [
455
+ 58.07103366227256,
456
+ 58.06551806286287,
457
+ 57.99719925634451,
458
+ 57.944828770626046,
459
+ 57.50830929080714
460
+ ],
461
+ "mean": 57.91737780858263,
462
+ "median": 57.99719925634451,
463
+ "std": 0.20977022066363163,
464
+ "min": 57.50830929080714,
465
+ "max": 58.07103366227256,
466
+ "p25": 57.944828770626046,
467
+ "p75": 58.06551806286287,
468
+ "p90": 58.06882742250868,
469
+ "p95": 58.06993054239062,
470
+ "p99": 58.07081303829617,
471
+ "unit": "tokens/sec"
472
+ },
473
+ "time_per_output_token_seconds": {
474
+ "name": "time_per_output_token_seconds",
475
+ "measurements": [
476
+ 0.017220289306640626,
477
+ 0.017221925048828125,
478
+ 0.0172422119140625,
479
+ 0.01725779541015625,
480
+ 0.01738879150390625
481
+ ],
482
+ "mean": 0.01726620263671875,
483
+ "median": 0.0172422119140625,
484
+ "std": 6.283757674931757e-05,
485
+ "min": 0.017220289306640626,
486
+ "max": 0.01738879150390625,
487
+ "p25": 0.017221925048828125,
488
+ "p75": 0.01725779541015625,
489
+ "p90": 0.01733639306640625,
490
+ "p95": 0.01736259228515625,
491
+ "p99": 0.01738355166015625,
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": "2025-10-06T16:48:57.147148",
504
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
505
+ "hardware_info": {
506
+ "gpu_name": "unknown",
507
+ "gpu_memory_total_mb": 0,
508
+ "cpu_count": 160,
509
+ "memory_total_mb": 1290069,
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
+ 1.258722412109375,
538
+ 1.2679189453125,
539
+ 1.2024837646484374,
540
+ 1.037858642578125,
541
+ 1.21246875
542
+ ],
543
+ "mean": 1.1958905029296876,
544
+ "median": 1.21246875,
545
+ "std": 0.08297990550181483,
546
+ "min": 1.037858642578125,
547
+ "max": 1.2679189453125,
548
+ "p25": 1.2024837646484374,
549
+ "p75": 1.258722412109375,
550
+ "p90": 1.2642403320312499,
551
+ "p95": 1.266079638671875,
552
+ "p99": 1.2675510839843749,
553
+ "unit": "seconds"
554
+ },
555
+ "time_to_first_token_seconds": {
556
+ "name": "time_to_first_token_seconds",
557
+ "measurements": [
558
+ 0.013544422149658203,
559
+ 0.013700497627258302,
560
+ 0.013858653068542481,
561
+ 0.011628024101257324,
562
+ 0.01140906810760498
563
+ ],
564
+ "mean": 0.012828133010864256,
565
+ "median": 0.013544422149658203,
566
+ "std": 0.0010761104346811096,
567
+ "min": 0.01140906810760498,
568
+ "max": 0.013858653068542481,
569
+ "p25": 0.011628024101257324,
570
+ "p75": 0.013700497627258302,
571
+ "p90": 0.01379539089202881,
572
+ "p95": 0.013827021980285644,
573
+ "p99": 0.013852326850891114,
574
+ "unit": "seconds"
575
+ },
576
+ "tokens_per_second": {
577
+ "name": "tokens_per_second",
578
+ "measurements": [
579
+ 79.44563395230198,
580
+ 78.86939490075473,
581
+ 83.1612059471226,
582
+ 96.35223516720147,
583
+ 82.47635248331143
584
+ ],
585
+ "mean": 84.06096449013845,
586
+ "median": 82.47635248331143,
587
+ "std": 6.366310804219967,
588
+ "min": 78.86939490075473,
589
+ "max": 96.35223516720147,
590
+ "p25": 79.44563395230198,
591
+ "p75": 83.1612059471226,
592
+ "p90": 91.07582347916993,
593
+ "p95": 93.7140293231857,
594
+ "p99": 95.82459399839831,
595
+ "unit": "tokens/sec"
596
+ },
597
+ "time_per_output_token_seconds": {
598
+ "name": "time_per_output_token_seconds",
599
+ "measurements": [
600
+ 0.012587224121093749,
601
+ 0.012679189453125,
602
+ 0.012024837646484375,
603
+ 0.010378586425781251,
604
+ 0.0121246875
605
+ ],
606
+ "mean": 0.011958905029296876,
607
+ "median": 0.0121246875,
608
+ "std": 0.0008297990550181484,
609
+ "min": 0.010378586425781251,
610
+ "max": 0.012679189453125,
611
+ "p25": 0.012024837646484375,
612
+ "p75": 0.012587224121093749,
613
+ "p90": 0.012642403320312499,
614
+ "p95": 0.01266079638671875,
615
+ "p99": 0.01267551083984375,
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": "2025-10-06T16:49:12.559660",
628
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
629
+ "hardware_info": {
630
+ "gpu_name": "unknown",
631
+ "gpu_memory_total_mb": 0,
632
+ "cpu_count": 160,
633
+ "memory_total_mb": 1290069,
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
+ 1.413236083984375,
662
+ 1.4222109375,
663
+ 1.4179783935546875,
664
+ 1.4124739990234374,
665
+ 1.420943115234375
666
+ ],
667
+ "mean": 1.417368505859375,
668
+ "median": 1.4179783935546875,
669
+ "std": 0.003940356509222913,
670
+ "min": 1.4124739990234374,
671
+ "max": 1.4222109375,
672
+ "p25": 1.413236083984375,
673
+ "p75": 1.420943115234375,
674
+ "p90": 1.4217038085937501,
675
+ "p95": 1.421957373046875,
676
+ "p99": 1.422160224609375,
677
+ "unit": "seconds"
678
+ },
679
+ "time_to_first_token_seconds": {
680
+ "name": "time_to_first_token_seconds",
681
+ "measurements": [
682
+ 0.015315542221069336,
683
+ 0.015408140182495117,
684
+ 0.01541914176940918,
685
+ 0.015681412696838377,
686
+ 0.015814130783081054
687
+ ],
688
+ "mean": 0.015527673530578611,
689
+ "median": 0.01541914176940918,
690
+ "std": 0.00018802156563235215,
691
+ "min": 0.015315542221069336,
692
+ "max": 0.015814130783081054,
693
+ "p25": 0.015408140182495117,
694
+ "p75": 0.015681412696838377,
695
+ "p90": 0.015761043548583983,
696
+ "p95": 0.01578758716583252,
697
+ "p99": 0.01580882205963135,
698
+ "unit": "seconds"
699
+ },
700
+ "tokens_per_second": {
701
+ "name": "tokens_per_second",
702
+ "measurements": [
703
+ 70.75958584220923,
704
+ 70.31305790390182,
705
+ 70.52293635399691,
706
+ 70.7977634060085,
707
+ 70.3757940257205
708
+ ],
709
+ "mean": 70.5538275063674,
710
+ "median": 70.52293635399691,
711
+ "std": 0.19619449579413517,
712
+ "min": 70.31305790390182,
713
+ "max": 70.7977634060085,
714
+ "p25": 70.3757940257205,
715
+ "p75": 70.75958584220923,
716
+ "p90": 70.78249238048879,
717
+ "p95": 70.79012789324865,
718
+ "p99": 70.79623630345652,
719
+ "unit": "tokens/sec"
720
+ },
721
+ "time_per_output_token_seconds": {
722
+ "name": "time_per_output_token_seconds",
723
+ "measurements": [
724
+ 0.01413236083984375,
725
+ 0.014222109375,
726
+ 0.014179783935546874,
727
+ 0.014124739990234374,
728
+ 0.01420943115234375
729
+ ],
730
+ "mean": 0.01417368505859375,
731
+ "median": 0.014179783935546874,
732
+ "std": 3.940356509222878e-05,
733
+ "min": 0.014124739990234374,
734
+ "max": 0.014222109375,
735
+ "p25": 0.01413236083984375,
736
+ "p75": 0.01420943115234375,
737
+ "p90": 0.0142170380859375,
738
+ "p95": 0.01421957373046875,
739
+ "p99": 0.01422160224609375,
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": "2025-10-06T16:49:30.120669",
752
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
753
+ "hardware_info": {
754
+ "gpu_name": "unknown",
755
+ "gpu_memory_total_mb": 0,
756
+ "cpu_count": 160,
757
+ "memory_total_mb": 1290069,
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
+ 1.5785823974609374,
786
+ 1.537088134765625,
787
+ 1.51745263671875,
788
+ 1.5229188232421875,
789
+ 1.5191055908203126
790
+ ],
791
+ "mean": 1.5350295166015626,
792
+ "median": 1.5229188232421875,
793
+ "std": 0.022848537599779548,
794
+ "min": 1.51745263671875,
795
+ "max": 1.5785823974609374,
796
+ "p25": 1.5191055908203126,
797
+ "p75": 1.537088134765625,
798
+ "p90": 1.5619846923828125,
799
+ "p95": 1.5702835449218748,
800
+ "p99": 1.576922626953125,
801
+ "unit": "seconds"
802
+ },
803
+ "time_to_first_token_seconds": {
804
+ "name": "time_to_first_token_seconds",
805
+ "measurements": [
806
+ 0.016804347991943358,
807
+ 0.016540395736694337,
808
+ 0.015838370323181153,
809
+ 0.017777488708496093,
810
+ 0.015661375045776367
811
+ ],
812
+ "mean": 0.016524395561218264,
813
+ "median": 0.016540395736694337,
814
+ "std": 0.0007568760588281651,
815
+ "min": 0.015661375045776367,
816
+ "max": 0.017777488708496093,
817
+ "p25": 0.015838370323181153,
818
+ "p75": 0.016804347991943358,
819
+ "p90": 0.017388232421875,
820
+ "p95": 0.017582860565185546,
821
+ "p99": 0.017738563079833982,
822
+ "unit": "seconds"
823
+ },
824
+ "tokens_per_second": {
825
+ "name": "tokens_per_second",
826
+ "measurements": [
827
+ 63.34797610871911,
828
+ 65.05807815324005,
829
+ 65.89991514742371,
830
+ 65.6633817074419,
831
+ 65.82820878567124
832
+ ],
833
+ "mean": 65.1595119804992,
834
+ "median": 65.6633817074419,
835
+ "std": 0.9530150056139661,
836
+ "min": 63.34797610871911,
837
+ "max": 65.89991514742371,
838
+ "p25": 65.05807815324005,
839
+ "p75": 65.82820878567124,
840
+ "p90": 65.87123260272273,
841
+ "p95": 65.88557387507322,
842
+ "p99": 65.89704689295361,
843
+ "unit": "tokens/sec"
844
+ },
845
+ "time_per_output_token_seconds": {
846
+ "name": "time_per_output_token_seconds",
847
+ "measurements": [
848
+ 0.015785823974609373,
849
+ 0.015370881347656249,
850
+ 0.0151745263671875,
851
+ 0.015229188232421875,
852
+ 0.015191055908203126
853
+ ],
854
+ "mean": 0.015350295166015627,
855
+ "median": 0.015229188232421875,
856
+ "std": 0.0002284853759977952,
857
+ "min": 0.0151745263671875,
858
+ "max": 0.015785823974609373,
859
+ "p25": 0.015191055908203126,
860
+ "p75": 0.015370881347656249,
861
+ "p90": 0.015619846923828123,
862
+ "p95": 0.015702835449218747,
863
+ "p99": 0.01576922626953125,
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": "2025-10-06T16:49:47.772173",
876
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
877
+ "hardware_info": {
878
+ "gpu_name": "unknown",
879
+ "gpu_memory_total_mb": 0,
880
+ "cpu_count": 160,
881
+ "memory_total_mb": 1290069,
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
+ 1.799255126953125,
910
+ 1.7842337646484374,
911
+ 1.7990965576171876,
912
+ 1.780177978515625,
913
+ 1.773841064453125
914
+ ],
915
+ "mean": 1.7873208984375002,
916
+ "median": 1.7842337646484374,
917
+ "std": 0.010230831673598336,
918
+ "min": 1.773841064453125,
919
+ "max": 1.799255126953125,
920
+ "p25": 1.780177978515625,
921
+ "p75": 1.7990965576171876,
922
+ "p90": 1.79919169921875,
923
+ "p95": 1.7992234130859375,
924
+ "p99": 1.7992487841796874,
925
+ "unit": "seconds"
926
+ },
927
+ "time_to_first_token_seconds": {
928
+ "name": "time_to_first_token_seconds",
929
+ "measurements": [
930
+ 0.01945689010620117,
931
+ 0.018858943939208984,
932
+ 0.01884702491760254,
933
+ 0.018989700317382813,
934
+ 0.018868984222412108
935
+ ],
936
+ "mean": 0.019004308700561523,
937
+ "median": 0.018868984222412108,
938
+ "std": 0.00023204531893629556,
939
+ "min": 0.01884702491760254,
940
+ "max": 0.01945689010620117,
941
+ "p25": 0.018858943939208984,
942
+ "p75": 0.018989700317382813,
943
+ "p90": 0.019270014190673826,
944
+ "p95": 0.019363452148437498,
945
+ "p99": 0.019438202514648437,
946
+ "unit": "seconds"
947
+ },
948
+ "tokens_per_second": {
949
+ "name": "tokens_per_second",
950
+ "measurements": [
951
+ 55.57855498199464,
952
+ 56.04646766658619,
953
+ 55.58345358208285,
954
+ 56.17415854305956,
955
+ 56.37483650815694
956
+ ],
957
+ "mean": 55.95149425637603,
958
+ "median": 56.04646766658619,
959
+ "std": 0.3201111168530252,
960
+ "min": 55.57855498199464,
961
+ "max": 56.37483650815694,
962
+ "p25": 55.58345358208285,
963
+ "p75": 56.17415854305956,
964
+ "p90": 56.29456532211799,
965
+ "p95": 56.33470091513746,
966
+ "p99": 56.36680938955304,
967
+ "unit": "tokens/sec"
968
+ },
969
+ "time_per_output_token_seconds": {
970
+ "name": "time_per_output_token_seconds",
971
+ "measurements": [
972
+ 0.01799255126953125,
973
+ 0.017842337646484373,
974
+ 0.017990965576171875,
975
+ 0.01780177978515625,
976
+ 0.01773841064453125
977
+ ],
978
+ "mean": 0.017873208984375,
979
+ "median": 0.017842337646484373,
980
+ "std": 0.00010230831673598374,
981
+ "min": 0.01773841064453125,
982
+ "max": 0.01799255126953125,
983
+ "p25": 0.01780177978515625,
984
+ "p75": 0.017990965576171875,
985
+ "p90": 0.0179919169921875,
986
+ "p95": 0.017992234130859375,
987
+ "p99": 0.017992487841796877,
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": "2025-10-06T16:50:14.058667",
1000
+ "commit_id": "505387c05b9e7dfcd73d8567959f7848525d8a2f",
1001
+ "hardware_info": {
1002
+ "gpu_name": "unknown",
1003
+ "gpu_memory_total_mb": 0,
1004
+ "cpu_count": 160,
1005
+ "memory_total_mb": 1290069,
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
+ 1.51914599609375,
1034
+ 1.52213525390625,
1035
+ 1.5182034912109375,
1036
+ 1.52373046875,
1037
+ 1.50556298828125
1038
+ ],
1039
+ "mean": 1.5177556396484375,
1040
+ "median": 1.51914599609375,
1041
+ "std": 0.006413641918114025,
1042
+ "min": 1.50556298828125,
1043
+ "max": 1.52373046875,
1044
+ "p25": 1.5182034912109375,
1045
+ "p75": 1.52213525390625,
1046
+ "p90": 1.5230923828125,
1047
+ "p95": 1.5234114257812499,
1048
+ "p99": 1.5236666601562499,
1049
+ "unit": "seconds"
1050
+ },
1051
+ "time_to_first_token_seconds": {
1052
+ "name": "time_to_first_token_seconds",
1053
+ "measurements": [
1054
+ 0.01562629508972168,
1055
+ 0.01588336753845215,
1056
+ 0.015721213340759276,
1057
+ 0.015683133125305174,
1058
+ 0.01561705493927002
1059
+ ],
1060
+ "mean": 0.015706212806701658,
1061
+ "median": 0.015683133125305174,
1062
+ "std": 9.641312515338897e-05,
1063
+ "min": 0.01561705493927002,
1064
+ "max": 0.01588336753845215,
1065
+ "p25": 0.01562629508972168,
1066
+ "p75": 0.015721213340759276,
1067
+ "p90": 0.015818505859375002,
1068
+ "p95": 0.015850936698913576,
1069
+ "p99": 0.015876881370544434,
1070
+ "unit": "seconds"
1071
+ },
1072
+ "tokens_per_second": {
1073
+ "name": "tokens_per_second",
1074
+ "measurements": [
1075
+ 65.82645792908292,
1076
+ 65.69718409935672,
1077
+ 65.86732317433864,
1078
+ 65.62840479394988,
1079
+ 66.42033629835704
1080
+ ],
1081
+ "mean": 65.88794125901703,
1082
+ "median": 65.82645792908292,
1083
+ "std": 0.2797846507050255,
1084
+ "min": 65.62840479394988,
1085
+ "max": 66.42033629835704,
1086
+ "p25": 65.69718409935672,
1087
+ "p75": 65.86732317433864,
1088
+ "p90": 66.19913104874968,
1089
+ "p95": 66.30973367355337,
1090
+ "p99": 66.3982157733963,
1091
+ "unit": "tokens/sec"
1092
+ },
1093
+ "time_per_output_token_seconds": {
1094
+ "name": "time_per_output_token_seconds",
1095
+ "measurements": [
1096
+ 0.015191459960937499,
1097
+ 0.0152213525390625,
1098
+ 0.015182034912109375,
1099
+ 0.0152373046875,
1100
+ 0.0150556298828125
1101
+ ],
1102
+ "mean": 0.015177556396484373,
1103
+ "median": 0.015191459960937499,
1104
+ "std": 6.413641918114052e-05,
1105
+ "min": 0.0150556298828125,
1106
+ "max": 0.0152373046875,
1107
+ "p25": 0.015182034912109375,
1108
+ "p75": 0.0152213525390625,
1109
+ "p90": 0.015230923828125,
1110
+ "p95": 0.0152341142578125,
1111
+ "p99": 0.0152366666015625,
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
+ }