hf-transformers-bot commited on
Commit
e0a81b1
·
verified ·
1 Parent(s): d2bcf75

Upload benchmark results for run 18351700409

Browse files
2025-10-08/18351700409/benchmark_results/Llama-2-7b-hf/Llama-2-7b-hf_benchmark_20251008_165037.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": "2025-10-08T16:47:33.797316",
8
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 1.5103232421875,
42
+ 1.486981689453125,
43
+ 1.491109375,
44
+ 1.47263916015625,
45
+ 1.4851337890625
46
+ ],
47
+ "mean": 1.489237451171875,
48
+ "median": 1.486981689453125,
49
+ "std": 0.01221103148674987,
50
+ "min": 1.47263916015625,
51
+ "max": 1.5103232421875,
52
+ "p25": 1.4851337890625,
53
+ "p75": 1.491109375,
54
+ "p90": 1.5026376953125,
55
+ "p95": 1.50648046875,
56
+ "p99": 1.5095546874999999,
57
+ "unit": "seconds"
58
+ },
59
+ "time_to_first_token_seconds": {
60
+ "name": "time_to_first_token_seconds",
61
+ "measurements": [
62
+ 0.015771682739257813,
63
+ 0.01584728240966797,
64
+ 0.01586856174468994,
65
+ 0.01582520294189453,
66
+ 0.015639124870300294
67
+ ],
68
+ "mean": 0.01579037094116211,
69
+ "median": 0.01582520294189453,
70
+ "std": 8.220779539727114e-05,
71
+ "min": 0.015639124870300294,
72
+ "max": 0.01586856174468994,
73
+ "p25": 0.015771682739257813,
74
+ "p75": 0.01584728240966797,
75
+ "p90": 0.015860050010681154,
76
+ "p95": 0.015864305877685548,
77
+ "p99": 0.01586771057128906,
78
+ "unit": "seconds"
79
+ },
80
+ "tokens_per_second": {
81
+ "name": "tokens_per_second",
82
+ "measurements": [
83
+ 66.21099192988878,
84
+ 67.25032373248492,
85
+ 67.06416154079912,
86
+ 67.9052973094847,
87
+ 67.33400097450185
88
+ ],
89
+ "mean": 67.15295509743189,
90
+ "median": 67.25032373248492,
91
+ "std": 0.5483531652940425,
92
+ "min": 66.21099192988878,
93
+ "max": 67.9052973094847,
94
+ "p25": 67.06416154079912,
95
+ "p75": 67.33400097450185,
96
+ "p90": 67.67677877549156,
97
+ "p95": 67.79103804248813,
98
+ "p99": 67.88244545608538,
99
+ "unit": "tokens/sec"
100
+ },
101
+ "time_per_output_token_seconds": {
102
+ "name": "time_per_output_token_seconds",
103
+ "measurements": [
104
+ 0.015103232421875,
105
+ 0.014869816894531249,
106
+ 0.01491109375,
107
+ 0.014726391601562501,
108
+ 0.014851337890625
109
+ ],
110
+ "mean": 0.01489237451171875,
111
+ "median": 0.014869816894531249,
112
+ "std": 0.00012211031486749875,
113
+ "min": 0.014726391601562501,
114
+ "max": 0.015103232421875,
115
+ "p25": 0.014851337890625,
116
+ "p75": 0.01491109375,
117
+ "p90": 0.015026376953125,
118
+ "p95": 0.015064804687499999,
119
+ "p99": 0.015095546875,
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": "2025-10-08T16:47:50.349982",
132
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 2.896669921875,
166
+ 2.915219970703125,
167
+ 2.969078125,
168
+ 2.97758740234375,
169
+ 2.999339599609375
170
+ ],
171
+ "mean": 2.95157900390625,
172
+ "median": 2.969078125,
173
+ "std": 0.03898901669761709,
174
+ "min": 2.896669921875,
175
+ "max": 2.999339599609375,
176
+ "p25": 2.915219970703125,
177
+ "p75": 2.97758740234375,
178
+ "p90": 2.990638720703125,
179
+ "p95": 2.99498916015625,
180
+ "p99": 2.99846951171875,
181
+ "unit": "seconds"
182
+ },
183
+ "time_to_first_token_seconds": {
184
+ "name": "time_to_first_token_seconds",
185
+ "measurements": [
186
+ 0.017231666564941406,
187
+ 0.016756750106811522,
188
+ 0.01737946319580078,
189
+ 0.016743072509765623,
190
+ 0.016818151473999024
191
+ ],
192
+ "mean": 0.016985820770263668,
193
+ "median": 0.016818151473999024,
194
+ "std": 0.00026642357015946717,
195
+ "min": 0.016743072509765623,
196
+ "max": 0.01737946319580078,
197
+ "p25": 0.016756750106811522,
198
+ "p75": 0.017231666564941406,
199
+ "p90": 0.01732034454345703,
200
+ "p95": 0.017349903869628903,
201
+ "p99": 0.017373551330566405,
202
+ "unit": "seconds"
203
+ },
204
+ "tokens_per_second": {
205
+ "name": "tokens_per_second",
206
+ "measurements": [
207
+ 34.52240079023933,
208
+ 34.302728783749686,
209
+ 33.68048794606912,
210
+ 33.58423666129395,
211
+ 33.34067273109844
212
+ ],
213
+ "mean": 33.8861053824901,
214
+ "median": 33.68048794606912,
215
+ "std": 0.4492987728896976,
216
+ "min": 33.34067273109844,
217
+ "max": 34.52240079023933,
218
+ "p25": 33.58423666129395,
219
+ "p75": 34.302728783749686,
220
+ "p90": 34.434531987643474,
221
+ "p95": 34.4784663889414,
222
+ "p99": 34.513613909979746,
223
+ "unit": "tokens/sec"
224
+ },
225
+ "time_per_output_token_seconds": {
226
+ "name": "time_per_output_token_seconds",
227
+ "measurements": [
228
+ 0.02896669921875,
229
+ 0.02915219970703125,
230
+ 0.02969078125,
231
+ 0.0297758740234375,
232
+ 0.02999339599609375
233
+ ],
234
+ "mean": 0.0295157900390625,
235
+ "median": 0.02969078125,
236
+ "std": 0.0003898901669761716,
237
+ "min": 0.02896669921875,
238
+ "max": 0.02999339599609375,
239
+ "p25": 0.02915219970703125,
240
+ "p75": 0.0297758740234375,
241
+ "p90": 0.02990638720703125,
242
+ "p95": 0.0299498916015625,
243
+ "p99": 0.0299846951171875,
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": "2025-10-08T16:48:29.641788",
256
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 1.2435693359375,
290
+ 1.240337158203125,
291
+ 1.2349161376953126,
292
+ 1.255384521484375,
293
+ 1.2356710205078125
294
+ ],
295
+ "mean": 1.241975634765625,
296
+ "median": 1.240337158203125,
297
+ "std": 0.007410902667461378,
298
+ "min": 1.2349161376953126,
299
+ "max": 1.255384521484375,
300
+ "p25": 1.2356710205078125,
301
+ "p75": 1.2435693359375,
302
+ "p90": 1.250658447265625,
303
+ "p95": 1.253021484375,
304
+ "p99": 1.2549119140625,
305
+ "unit": "seconds"
306
+ },
307
+ "time_to_first_token_seconds": {
308
+ "name": "time_to_first_token_seconds",
309
+ "measurements": [
310
+ 0.013400632858276368,
311
+ 0.013522631645202636,
312
+ 0.01343767261505127,
313
+ 0.013448633193969727,
314
+ 0.013552230834960937
315
+ ],
316
+ "mean": 0.013472360229492187,
317
+ "median": 0.013448633193969727,
318
+ "std": 5.6245084135940905e-05,
319
+ "min": 0.013400632858276368,
320
+ "max": 0.013552230834960937,
321
+ "p25": 0.01343767261505127,
322
+ "p75": 0.013522631645202636,
323
+ "p90": 0.013540391159057617,
324
+ "p95": 0.013546310997009278,
325
+ "p99": 0.013551046867370605,
326
+ "unit": "seconds"
327
+ },
328
+ "tokens_per_second": {
329
+ "name": "tokens_per_second",
330
+ "measurements": [
331
+ 80.41369074496531,
332
+ 80.62323968820694,
333
+ 80.97715864870553,
334
+ 79.65686870327136,
335
+ 80.92768895632423
336
+ ],
337
+ "mean": 80.51972934829467,
338
+ "median": 80.62323968820694,
339
+ "std": 0.47794553565293224,
340
+ "min": 79.65686870327136,
341
+ "max": 80.97715864870553,
342
+ "p25": 80.41369074496531,
343
+ "p75": 80.92768895632423,
344
+ "p90": 80.957370771753,
345
+ "p95": 80.96726471022927,
346
+ "p99": 80.97517986101028,
347
+ "unit": "tokens/sec"
348
+ },
349
+ "time_per_output_token_seconds": {
350
+ "name": "time_per_output_token_seconds",
351
+ "measurements": [
352
+ 0.012435693359375,
353
+ 0.012403371582031251,
354
+ 0.012349161376953126,
355
+ 0.012553845214843751,
356
+ 0.012356710205078125
357
+ ],
358
+ "mean": 0.012419756347656252,
359
+ "median": 0.012403371582031251,
360
+ "std": 7.410902667461418e-05,
361
+ "min": 0.012349161376953126,
362
+ "max": 0.012553845214843751,
363
+ "p25": 0.012356710205078125,
364
+ "p75": 0.012435693359375,
365
+ "p90": 0.01250658447265625,
366
+ "p95": 0.012530214843750002,
367
+ "p99": 0.012549119140625002,
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": "2025-10-08T16:48:44.889111",
380
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 1.5058968505859376,
414
+ 1.50697265625,
415
+ 1.51368310546875,
416
+ 1.5143343505859375,
417
+ 1.51178759765625
418
+ ],
419
+ "mean": 1.510534912109375,
420
+ "median": 1.51178759765625,
421
+ "std": 0.003467487757653589,
422
+ "min": 1.5058968505859376,
423
+ "max": 1.5143343505859375,
424
+ "p25": 1.50697265625,
425
+ "p75": 1.51368310546875,
426
+ "p90": 1.5140738525390625,
427
+ "p95": 1.5142041015625,
428
+ "p99": 1.51430830078125,
429
+ "unit": "seconds"
430
+ },
431
+ "time_to_first_token_seconds": {
432
+ "name": "time_to_first_token_seconds",
433
+ "measurements": [
434
+ 0.01723542594909668,
435
+ 0.017408224105834962,
436
+ 0.01726866340637207,
437
+ 0.017369104385375977,
438
+ 0.017894216537475586
439
+ ],
440
+ "mean": 0.017435126876831054,
441
+ "median": 0.017369104385375977,
442
+ "std": 0.00023809098786435997,
443
+ "min": 0.01723542594909668,
444
+ "max": 0.017894216537475586,
445
+ "p25": 0.01726866340637207,
446
+ "p75": 0.017408224105834962,
447
+ "p90": 0.017699819564819336,
448
+ "p95": 0.017797018051147463,
449
+ "p99": 0.017874776840209963,
450
+ "unit": "seconds"
451
+ },
452
+ "tokens_per_second": {
453
+ "name": "tokens_per_second",
454
+ "measurements": [
455
+ 66.40561069046028,
456
+ 66.35820469950879,
457
+ 66.06402597658146,
458
+ 66.03561489660936,
459
+ 66.1468582987661
460
+ ],
461
+ "mean": 66.2020629123852,
462
+ "median": 66.1468582987661,
463
+ "std": 0.1520648826074468,
464
+ "min": 66.03561489660936,
465
+ "max": 66.40561069046028,
466
+ "p25": 66.06402597658146,
467
+ "p75": 66.35820469950879,
468
+ "p90": 66.38664829407968,
469
+ "p95": 66.39612949226998,
470
+ "p99": 66.40371445082222,
471
+ "unit": "tokens/sec"
472
+ },
473
+ "time_per_output_token_seconds": {
474
+ "name": "time_per_output_token_seconds",
475
+ "measurements": [
476
+ 0.015058968505859376,
477
+ 0.015069726562500002,
478
+ 0.015136831054687501,
479
+ 0.015143343505859375,
480
+ 0.0151178759765625
481
+ ],
482
+ "mean": 0.01510534912109375,
483
+ "median": 0.0151178759765625,
484
+ "std": 3.4674877576535916e-05,
485
+ "min": 0.015058968505859376,
486
+ "max": 0.015143343505859375,
487
+ "p25": 0.015069726562500002,
488
+ "p75": 0.015136831054687501,
489
+ "p90": 0.015140738525390626,
490
+ "p95": 0.015142041015625,
491
+ "p99": 0.0151430830078125,
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": "2025-10-08T16:49:02.444036",
504
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 1.3873543701171875,
538
+ 1.3830389404296874,
539
+ 1.3685765380859376,
540
+ 1.3873699951171874,
541
+ 1.3700758056640625
542
+ ],
543
+ "mean": 1.3792831298828125,
544
+ "median": 1.3830389404296874,
545
+ "std": 0.008295231742128028,
546
+ "min": 1.3685765380859376,
547
+ "max": 1.3873699951171874,
548
+ "p25": 1.3700758056640625,
549
+ "p75": 1.3873543701171875,
550
+ "p90": 1.3873637451171874,
551
+ "p95": 1.3873668701171875,
552
+ "p99": 1.3873693701171874,
553
+ "unit": "seconds"
554
+ },
555
+ "time_to_first_token_seconds": {
556
+ "name": "time_to_first_token_seconds",
557
+ "measurements": [
558
+ 0.014968173027038574,
559
+ 0.014970053672790527,
560
+ 0.014916854858398438,
561
+ 0.014887934684753417,
562
+ 0.015169891357421875
563
+ ],
564
+ "mean": 0.014982581520080567,
565
+ "median": 0.014968173027038574,
566
+ "std": 9.871934388721647e-05,
567
+ "min": 0.014887934684753417,
568
+ "max": 0.015169891357421875,
569
+ "p25": 0.014916854858398438,
570
+ "p75": 0.014970053672790527,
571
+ "p90": 0.015089956283569336,
572
+ "p95": 0.015129923820495605,
573
+ "p99": 0.015161897850036622,
574
+ "unit": "seconds"
575
+ },
576
+ "tokens_per_second": {
577
+ "name": "tokens_per_second",
578
+ "measurements": [
579
+ 72.07963744083148,
580
+ 72.30454405638908,
581
+ 73.06862072899328,
582
+ 72.07882565714078,
583
+ 72.98866207737386
584
+ ],
585
+ "mean": 72.50405799214569,
586
+ "median": 72.30454405638908,
587
+ "std": 0.4368829035174882,
588
+ "min": 72.07882565714078,
589
+ "max": 73.06862072899328,
590
+ "p25": 72.07963744083148,
591
+ "p75": 72.98866207737386,
592
+ "p90": 73.03663726834552,
593
+ "p95": 73.05262899866939,
594
+ "p99": 73.0654223829285,
595
+ "unit": "tokens/sec"
596
+ },
597
+ "time_per_output_token_seconds": {
598
+ "name": "time_per_output_token_seconds",
599
+ "measurements": [
600
+ 0.013873543701171876,
601
+ 0.013830389404296874,
602
+ 0.013685765380859376,
603
+ 0.013873699951171874,
604
+ 0.013700758056640625
605
+ ],
606
+ "mean": 0.013792831298828126,
607
+ "median": 0.013830389404296874,
608
+ "std": 8.295231742128034e-05,
609
+ "min": 0.013685765380859376,
610
+ "max": 0.013873699951171874,
611
+ "p25": 0.013700758056640625,
612
+ "p75": 0.013873543701171876,
613
+ "p90": 0.013873637451171875,
614
+ "p95": 0.013873668701171874,
615
+ "p99": 0.013873693701171873,
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": "2025-10-08T16:49:19.661869",
628
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 1.3659605712890626,
662
+ 1.368508544921875,
663
+ 1.44200830078125,
664
+ 1.3304757080078125,
665
+ 1.2310838623046876
666
+ ],
667
+ "mean": 1.3476073974609375,
668
+ "median": 1.3659605712890626,
669
+ "std": 0.06863884691815322,
670
+ "min": 1.2310838623046876,
671
+ "max": 1.44200830078125,
672
+ "p25": 1.3304757080078125,
673
+ "p75": 1.368508544921875,
674
+ "p90": 1.4126083984375,
675
+ "p95": 1.427308349609375,
676
+ "p99": 1.4390683105468751,
677
+ "unit": "seconds"
678
+ },
679
+ "time_to_first_token_seconds": {
680
+ "name": "time_to_first_token_seconds",
681
+ "measurements": [
682
+ 0.015998961448669434,
683
+ 0.014891414642333985,
684
+ 0.015130412101745606,
685
+ 0.015309049606323242,
686
+ 0.013614394187927247
687
+ ],
688
+ "mean": 0.014988846397399901,
689
+ "median": 0.015130412101745606,
690
+ "std": 0.000779937841558215,
691
+ "min": 0.013614394187927247,
692
+ "max": 0.015998961448669434,
693
+ "p25": 0.014891414642333985,
694
+ "p75": 0.015309049606323242,
695
+ "p90": 0.01572299671173096,
696
+ "p95": 0.015860979080200194,
697
+ "p99": 0.015971364974975585,
698
+ "unit": "seconds"
699
+ },
700
+ "tokens_per_second": {
701
+ "name": "tokens_per_second",
702
+ "measurements": [
703
+ 73.20855528474705,
704
+ 73.07225108025085,
705
+ 69.3477284047686,
706
+ 75.16108666856833,
707
+ 81.22923471094161
708
+ ],
709
+ "mean": 74.40377122985528,
710
+ "median": 73.20855528474705,
711
+ "std": 3.8968072705411108,
712
+ "min": 69.3477284047686,
713
+ "max": 81.22923471094161,
714
+ "p25": 73.07225108025085,
715
+ "p75": 75.16108666856833,
716
+ "p90": 78.8019754939923,
717
+ "p95": 80.01560510246695,
718
+ "p99": 80.98650878924668,
719
+ "unit": "tokens/sec"
720
+ },
721
+ "time_per_output_token_seconds": {
722
+ "name": "time_per_output_token_seconds",
723
+ "measurements": [
724
+ 0.013659605712890625,
725
+ 0.01368508544921875,
726
+ 0.0144200830078125,
727
+ 0.013304757080078125,
728
+ 0.012310838623046876
729
+ ],
730
+ "mean": 0.013476073974609376,
731
+ "median": 0.013659605712890625,
732
+ "std": 0.0006863884691815323,
733
+ "min": 0.012310838623046876,
734
+ "max": 0.0144200830078125,
735
+ "p25": 0.013304757080078125,
736
+ "p75": 0.01368508544921875,
737
+ "p90": 0.014126083984375001,
738
+ "p95": 0.01427308349609375,
739
+ "p99": 0.014390683105468751,
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": "2025-10-08T16:49:36.055607",
752
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 1.639004638671875,
786
+ 1.64076220703125,
787
+ 1.5574283447265624,
788
+ 1.345704345703125,
789
+ 1.3005045166015625
790
+ ],
791
+ "mean": 1.496680810546875,
792
+ "median": 1.5574283447265624,
793
+ "std": 0.1455917683227924,
794
+ "min": 1.3005045166015625,
795
+ "max": 1.64076220703125,
796
+ "p25": 1.345704345703125,
797
+ "p75": 1.639004638671875,
798
+ "p90": 1.6400591796875,
799
+ "p95": 1.640410693359375,
800
+ "p99": 1.640691904296875,
801
+ "unit": "seconds"
802
+ },
803
+ "time_to_first_token_seconds": {
804
+ "name": "time_to_first_token_seconds",
805
+ "measurements": [
806
+ 0.016977428436279297,
807
+ 0.016975387573242188,
808
+ 0.017003187179565428,
809
+ 0.018475610733032225,
810
+ 0.013728590011596679
811
+ ],
812
+ "mean": 0.016632040786743164,
813
+ "median": 0.016977428436279297,
814
+ "std": 0.0015622871592453965,
815
+ "min": 0.013728590011596679,
816
+ "max": 0.018475610733032225,
817
+ "p25": 0.016975387573242188,
818
+ "p75": 0.017003187179565428,
819
+ "p90": 0.017886641311645506,
820
+ "p95": 0.018181126022338866,
821
+ "p99": 0.018416713790893553,
822
+ "unit": "seconds"
823
+ },
824
+ "tokens_per_second": {
825
+ "name": "tokens_per_second",
826
+ "measurements": [
827
+ 61.012640013656345,
828
+ 60.94728387298562,
829
+ 64.20841147433784,
830
+ 74.3105276573588,
831
+ 76.89323545089782
832
+ ],
833
+ "mean": 67.4744196938473,
834
+ "median": 64.20841147433784,
835
+ "std": 6.789276407736781,
836
+ "min": 60.94728387298562,
837
+ "max": 76.89323545089782,
838
+ "p25": 61.012640013656345,
839
+ "p75": 74.3105276573588,
840
+ "p90": 75.86015233348222,
841
+ "p95": 76.37669389219002,
842
+ "p99": 76.78992713915626,
843
+ "unit": "tokens/sec"
844
+ },
845
+ "time_per_output_token_seconds": {
846
+ "name": "time_per_output_token_seconds",
847
+ "measurements": [
848
+ 0.01639004638671875,
849
+ 0.0164076220703125,
850
+ 0.015574283447265625,
851
+ 0.01345704345703125,
852
+ 0.013005045166015625
853
+ ],
854
+ "mean": 0.014966808105468748,
855
+ "median": 0.015574283447265625,
856
+ "std": 0.0014559176832279236,
857
+ "min": 0.013005045166015625,
858
+ "max": 0.0164076220703125,
859
+ "p25": 0.01345704345703125,
860
+ "p75": 0.01639004638671875,
861
+ "p90": 0.016400591796875,
862
+ "p95": 0.01640410693359375,
863
+ "p99": 0.01640691904296875,
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": "2025-10-08T16:49:54.811322",
876
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 1.9403372802734375,
910
+ 1.9268399658203126,
911
+ 1.9812059326171876,
912
+ 2.024867919921875,
913
+ 1.6080335693359376
914
+ ],
915
+ "mean": 1.8962569335937502,
916
+ "median": 1.9403372802734375,
917
+ "std": 0.14812712430006952,
918
+ "min": 1.6080335693359376,
919
+ "max": 2.024867919921875,
920
+ "p25": 1.9268399658203126,
921
+ "p75": 1.9812059326171876,
922
+ "p90": 2.007403125,
923
+ "p95": 2.0161355224609374,
924
+ "p99": 2.0231214404296876,
925
+ "unit": "seconds"
926
+ },
927
+ "time_to_first_token_seconds": {
928
+ "name": "time_to_first_token_seconds",
929
+ "measurements": [
930
+ 0.020655384063720705,
931
+ 0.020297708511352538,
932
+ 0.020569103240966797,
933
+ 0.020394426345825195,
934
+ 0.01685442924499512
935
+ ],
936
+ "mean": 0.01975421028137207,
937
+ "median": 0.020394426345825195,
938
+ "std": 0.0014553461248580514,
939
+ "min": 0.01685442924499512,
940
+ "max": 0.020655384063720705,
941
+ "p25": 0.020297708511352538,
942
+ "p75": 0.020569103240966797,
943
+ "p90": 0.020620871734619142,
944
+ "p95": 0.02063812789916992,
945
+ "p99": 0.020651932830810547,
946
+ "unit": "seconds"
947
+ },
948
+ "tokens_per_second": {
949
+ "name": "tokens_per_second",
950
+ "measurements": [
951
+ 51.53743167059478,
952
+ 51.89844604319646,
953
+ 50.47430878015758,
954
+ 49.385937233801535,
955
+ 62.18775646661192
956
+ ],
957
+ "mean": 53.09677603887245,
958
+ "median": 51.53743167059478,
959
+ "std": 4.629495305866719,
960
+ "min": 49.385937233801535,
961
+ "max": 62.18775646661192,
962
+ "p25": 50.47430878015758,
963
+ "p75": 51.89844604319646,
964
+ "p90": 58.07203229724574,
965
+ "p95": 60.12989438192883,
966
+ "p99": 61.7761840496753,
967
+ "unit": "tokens/sec"
968
+ },
969
+ "time_per_output_token_seconds": {
970
+ "name": "time_per_output_token_seconds",
971
+ "measurements": [
972
+ 0.019403372802734374,
973
+ 0.019268399658203126,
974
+ 0.019812059326171874,
975
+ 0.02024867919921875,
976
+ 0.016080335693359375
977
+ ],
978
+ "mean": 0.018962569335937502,
979
+ "median": 0.019403372802734374,
980
+ "std": 0.0014812712430006955,
981
+ "min": 0.016080335693359375,
982
+ "max": 0.02024867919921875,
983
+ "p25": 0.019268399658203126,
984
+ "p75": 0.019812059326171874,
985
+ "p90": 0.02007403125,
986
+ "p95": 0.020161355224609375,
987
+ "p99": 0.020231214404296875,
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": "2025-10-08T16:50:22.220162",
1000
+ "commit_id": "bef73bf8d7421e2158a83ebcc24a9fc78e885dfe",
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
+ 1.6417381591796876,
1034
+ 1.3114473876953125,
1035
+ 1.3418922119140626,
1036
+ 1.66994873046875,
1037
+ 1.6428519287109375
1038
+ ],
1039
+ "mean": 1.5215756835937502,
1040
+ "median": 1.6417381591796876,
1041
+ "std": 0.1597507808680023,
1042
+ "min": 1.3114473876953125,
1043
+ "max": 1.66994873046875,
1044
+ "p25": 1.3418922119140626,
1045
+ "p75": 1.6428519287109375,
1046
+ "p90": 1.659110009765625,
1047
+ "p95": 1.6645293701171875,
1048
+ "p99": 1.6688648583984373,
1049
+ "unit": "seconds"
1050
+ },
1051
+ "time_to_first_token_seconds": {
1052
+ "name": "time_to_first_token_seconds",
1053
+ "measurements": [
1054
+ 0.01717442512512207,
1055
+ 0.017789539337158204,
1056
+ 0.01361779022216797,
1057
+ 0.016827352523803712,
1058
+ 0.017435184478759767
1059
+ ],
1060
+ "mean": 0.01656885833740234,
1061
+ "median": 0.01717442512512207,
1062
+ "std": 0.0015088350795172856,
1063
+ "min": 0.01361779022216797,
1064
+ "max": 0.017789539337158204,
1065
+ "p25": 0.016827352523803712,
1066
+ "p75": 0.017435184478759767,
1067
+ "p90": 0.01764779739379883,
1068
+ "p95": 0.017718668365478515,
1069
+ "p99": 0.017775365142822266,
1070
+ "unit": "seconds"
1071
+ },
1072
+ "tokens_per_second": {
1073
+ "name": "tokens_per_second",
1074
+ "measurements": [
1075
+ 60.91105298421406,
1076
+ 76.25162925958942,
1077
+ 74.52163378857452,
1078
+ 59.882077919799535,
1079
+ 60.8697584075425
1080
+ ],
1081
+ "mean": 66.487230471944,
1082
+ "median": 60.91105298421406,
1083
+ "std": 7.296203236540946,
1084
+ "min": 59.882077919799535,
1085
+ "max": 76.25162925958942,
1086
+ "p25": 60.8697584075425,
1087
+ "p75": 74.52163378857452,
1088
+ "p90": 75.55963107118346,
1089
+ "p95": 75.90563016538644,
1090
+ "p99": 76.18242944074882,
1091
+ "unit": "tokens/sec"
1092
+ },
1093
+ "time_per_output_token_seconds": {
1094
+ "name": "time_per_output_token_seconds",
1095
+ "measurements": [
1096
+ 0.016417381591796877,
1097
+ 0.013114473876953126,
1098
+ 0.013418922119140626,
1099
+ 0.0166994873046875,
1100
+ 0.016428519287109377
1101
+ ],
1102
+ "mean": 0.015215756835937502,
1103
+ "median": 0.016417381591796877,
1104
+ "std": 0.0015975078086800233,
1105
+ "min": 0.013114473876953126,
1106
+ "max": 0.0166994873046875,
1107
+ "p25": 0.013418922119140626,
1108
+ "p75": 0.016428519287109377,
1109
+ "p90": 0.016591100097656252,
1110
+ "p95": 0.016645293701171876,
1111
+ "p99": 0.016688648583984375,
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
+ }