hf-transformers-bot commited on
Commit
106675a
·
verified ·
1 Parent(s): c542503

Upload benchmark results for run 18319645219

Browse files
2025-10-07/18319645219/benchmark_results/Llama-2-7b-hf/Llama-2-7b-hf_benchmark_20251007_164739.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": "2025-10-07T16:44:51.064763",
8
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 1.45562744140625,
42
+ 1.4441689453125,
43
+ 1.4354913330078125,
44
+ 1.38665673828125,
45
+ 1.2561416015625
46
+ ],
47
+ "mean": 1.3956172119140624,
48
+ "median": 1.4354913330078125,
49
+ "std": 0.07359644008151782,
50
+ "min": 1.2561416015625,
51
+ "max": 1.45562744140625,
52
+ "p25": 1.38665673828125,
53
+ "p75": 1.4441689453125,
54
+ "p90": 1.45104404296875,
55
+ "p95": 1.4533357421875,
56
+ "p99": 1.4551691015625,
57
+ "unit": "seconds"
58
+ },
59
+ "time_to_first_token_seconds": {
60
+ "name": "time_to_first_token_seconds",
61
+ "measurements": [
62
+ 0.015133640289306641,
63
+ 0.015530954360961914,
64
+ 0.01528835678100586,
65
+ 0.015322834968566894,
66
+ 0.013661869049072265
67
+ ],
68
+ "mean": 0.014987531089782716,
69
+ "median": 0.01528835678100586,
70
+ "std": 0.0006748275903744328,
71
+ "min": 0.013661869049072265,
72
+ "max": 0.015530954360961914,
73
+ "p25": 0.015133640289306641,
74
+ "p75": 0.015322834968566894,
75
+ "p90": 0.015447706604003906,
76
+ "p95": 0.01548933048248291,
77
+ "p99": 0.015522629585266113,
78
+ "unit": "seconds"
79
+ },
80
+ "tokens_per_second": {
81
+ "name": "tokens_per_second",
82
+ "measurements": [
83
+ 68.69889722839532,
84
+ 69.24397614599118,
85
+ 69.66255922316722,
86
+ 72.11590095754282,
87
+ 79.60885928434435
88
+ ],
89
+ "mean": 71.86603856788818,
90
+ "median": 69.66255922316722,
91
+ "std": 4.044132062438307,
92
+ "min": 68.69889722839532,
93
+ "max": 79.60885928434435,
94
+ "p25": 69.24397614599118,
95
+ "p75": 72.11590095754282,
96
+ "p90": 76.61167595362373,
97
+ "p95": 78.11026761898404,
98
+ "p99": 79.30914095127228,
99
+ "unit": "tokens/sec"
100
+ },
101
+ "time_per_output_token_seconds": {
102
+ "name": "time_per_output_token_seconds",
103
+ "measurements": [
104
+ 0.0145562744140625,
105
+ 0.014441689453125,
106
+ 0.014354913330078125,
107
+ 0.0138665673828125,
108
+ 0.012561416015625
109
+ ],
110
+ "mean": 0.013956172119140626,
111
+ "median": 0.014354913330078125,
112
+ "std": 0.0007359644008151783,
113
+ "min": 0.012561416015625,
114
+ "max": 0.0145562744140625,
115
+ "p25": 0.0138665673828125,
116
+ "p75": 0.014441689453125,
117
+ "p90": 0.0145104404296875,
118
+ "p95": 0.014533357421875,
119
+ "p99": 0.014551691015625,
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": "2025-10-07T16:45:06.709036",
132
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 2.981670166015625,
166
+ 2.93972705078125,
167
+ 3.061232666015625,
168
+ 3.0037060546875,
169
+ 3.031892822265625
170
+ ],
171
+ "mean": 3.003645751953125,
172
+ "median": 3.0037060546875,
173
+ "std": 0.041671765083079335,
174
+ "min": 2.93972705078125,
175
+ "max": 3.061232666015625,
176
+ "p25": 2.981670166015625,
177
+ "p75": 3.031892822265625,
178
+ "p90": 3.0494967285156247,
179
+ "p95": 3.0553646972656248,
180
+ "p99": 3.060059072265625,
181
+ "unit": "seconds"
182
+ },
183
+ "time_to_first_token_seconds": {
184
+ "name": "time_to_first_token_seconds",
185
+ "measurements": [
186
+ 0.018348104476928712,
187
+ 0.01657129669189453,
188
+ 0.016540777206420897,
189
+ 0.016198104858398437,
190
+ 0.016286062240600585
191
+ ],
192
+ "mean": 0.016788869094848634,
193
+ "median": 0.016540777206420897,
194
+ "std": 0.0007927066832200513,
195
+ "min": 0.016198104858398437,
196
+ "max": 0.018348104476928712,
197
+ "p25": 0.016286062240600585,
198
+ "p75": 0.01657129669189453,
199
+ "p90": 0.01763738136291504,
200
+ "p95": 0.017992742919921877,
201
+ "p99": 0.018277032165527344,
202
+ "unit": "seconds"
203
+ },
204
+ "tokens_per_second": {
205
+ "name": "tokens_per_second",
206
+ "measurements": [
207
+ 33.53825018601201,
208
+ 34.01676355409405,
209
+ 32.66657941755074,
210
+ 33.292205754934905,
211
+ 32.982696243620374
212
+ ],
213
+ "mean": 33.29929903124242,
214
+ "median": 33.292205754934905,
215
+ "std": 0.4631359941022515,
216
+ "min": 32.66657941755074,
217
+ "max": 34.01676355409405,
218
+ "p25": 32.982696243620374,
219
+ "p75": 33.53825018601201,
220
+ "p90": 33.82535820686123,
221
+ "p95": 33.92106088047764,
222
+ "p99": 33.997623019370764,
223
+ "unit": "tokens/sec"
224
+ },
225
+ "time_per_output_token_seconds": {
226
+ "name": "time_per_output_token_seconds",
227
+ "measurements": [
228
+ 0.02981670166015625,
229
+ 0.0293972705078125,
230
+ 0.030612326660156248,
231
+ 0.030037060546875,
232
+ 0.03031892822265625
233
+ ],
234
+ "mean": 0.03003645751953125,
235
+ "median": 0.030037060546875,
236
+ "std": 0.0004167176508307928,
237
+ "min": 0.0293972705078125,
238
+ "max": 0.030612326660156248,
239
+ "p25": 0.02981670166015625,
240
+ "p75": 0.03031892822265625,
241
+ "p90": 0.030494967285156247,
242
+ "p95": 0.030553646972656247,
243
+ "p99": 0.030600590722656248,
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": "2025-10-07T16:45:46.870322",
256
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 1.062322509765625,
290
+ 1.1968133544921875,
291
+ 1.21119482421875,
292
+ 1.214175537109375,
293
+ 1.19860546875
294
+ ],
295
+ "mean": 1.1766223388671875,
296
+ "median": 1.19860546875,
297
+ "std": 0.05755151689688297,
298
+ "min": 1.062322509765625,
299
+ "max": 1.214175537109375,
300
+ "p25": 1.1968133544921875,
301
+ "p75": 1.21119482421875,
302
+ "p90": 1.212983251953125,
303
+ "p95": 1.21357939453125,
304
+ "p99": 1.21405630859375,
305
+ "unit": "seconds"
306
+ },
307
+ "time_to_first_token_seconds": {
308
+ "name": "time_to_first_token_seconds",
309
+ "measurements": [
310
+ 0.011842226028442383,
311
+ 0.013398957252502442,
312
+ 0.013412515640258789,
313
+ 0.013431636810302735,
314
+ 0.013555193901062012
315
+ ],
316
+ "mean": 0.013128105926513672,
317
+ "median": 0.013412515640258789,
318
+ "std": 0.0006453327163167625,
319
+ "min": 0.011842226028442383,
320
+ "max": 0.013555193901062012,
321
+ "p25": 0.013398957252502442,
322
+ "p75": 0.013431636810302735,
323
+ "p90": 0.013505771064758301,
324
+ "p95": 0.013530482482910156,
325
+ "p99": 0.01355025161743164,
326
+ "unit": "seconds"
327
+ },
328
+ "tokens_per_second": {
329
+ "name": "tokens_per_second",
330
+ "measurements": [
331
+ 94.13337200400893,
332
+ 83.55521738176992,
333
+ 82.56310050243356,
334
+ 82.36041407824199,
335
+ 83.43028845370434
336
+ ],
337
+ "mean": 85.20847848403176,
338
+ "median": 83.43028845370434,
339
+ "std": 4.486835230408838,
340
+ "min": 82.36041407824199,
341
+ "max": 94.13337200400893,
342
+ "p25": 82.56310050243356,
343
+ "p75": 83.55521738176992,
344
+ "p90": 89.90211015511333,
345
+ "p95": 92.01774107956113,
346
+ "p99": 93.71024581911938,
347
+ "unit": "tokens/sec"
348
+ },
349
+ "time_per_output_token_seconds": {
350
+ "name": "time_per_output_token_seconds",
351
+ "measurements": [
352
+ 0.010623225097656251,
353
+ 0.011968133544921876,
354
+ 0.0121119482421875,
355
+ 0.01214175537109375,
356
+ 0.0119860546875
357
+ ],
358
+ "mean": 0.011766223388671875,
359
+ "median": 0.0119860546875,
360
+ "std": 0.0005755151689688293,
361
+ "min": 0.010623225097656251,
362
+ "max": 0.01214175537109375,
363
+ "p25": 0.011968133544921876,
364
+ "p75": 0.0121119482421875,
365
+ "p90": 0.01212983251953125,
366
+ "p95": 0.0121357939453125,
367
+ "p99": 0.0121405630859375,
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": "2025-10-07T16:46:00.800348",
380
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 1.2528175048828125,
414
+ 1.265097900390625,
415
+ 1.4556885986328125,
416
+ 1.466048828125,
417
+ 1.4553988037109375
418
+ ],
419
+ "mean": 1.3790103271484375,
420
+ "median": 1.4553988037109375,
421
+ "std": 0.0981744637702043,
422
+ "min": 1.2528175048828125,
423
+ "max": 1.466048828125,
424
+ "p25": 1.265097900390625,
425
+ "p75": 1.4556885986328125,
426
+ "p90": 1.461904736328125,
427
+ "p95": 1.4639767822265624,
428
+ "p99": 1.4656344189453123,
429
+ "unit": "seconds"
430
+ },
431
+ "time_to_first_token_seconds": {
432
+ "name": "time_to_first_token_seconds",
433
+ "measurements": [
434
+ 0.0165382194519043,
435
+ 0.015063767433166503,
436
+ 0.017294845581054687,
437
+ 0.01683037567138672,
438
+ 0.0168732967376709
439
+ ],
440
+ "mean": 0.016520100975036624,
441
+ "median": 0.01683037567138672,
442
+ "std": 0.0007671352557237882,
443
+ "min": 0.015063767433166503,
444
+ "max": 0.017294845581054687,
445
+ "p25": 0.0165382194519043,
446
+ "p75": 0.0168732967376709,
447
+ "p90": 0.017126226043701172,
448
+ "p95": 0.017210535812377928,
449
+ "p99": 0.017277983627319336,
450
+ "unit": "seconds"
451
+ },
452
+ "tokens_per_second": {
453
+ "name": "tokens_per_second",
454
+ "measurements": [
455
+ 79.82008521612565,
456
+ 79.04526595856568,
457
+ 68.69601101081669,
458
+ 68.21055211912334,
459
+ 68.709689567576
460
+ ],
461
+ "mean": 72.89632077444148,
462
+ "median": 68.709689567576,
463
+ "std": 5.345558002400832,
464
+ "min": 68.21055211912334,
465
+ "max": 79.82008521612565,
466
+ "p25": 68.69601101081669,
467
+ "p75": 79.04526595856568,
468
+ "p90": 79.51015751310166,
469
+ "p95": 79.66512136461365,
470
+ "p99": 79.78909244582324,
471
+ "unit": "tokens/sec"
472
+ },
473
+ "time_per_output_token_seconds": {
474
+ "name": "time_per_output_token_seconds",
475
+ "measurements": [
476
+ 0.012528175048828125,
477
+ 0.01265097900390625,
478
+ 0.014556885986328125,
479
+ 0.01466048828125,
480
+ 0.014553988037109375
481
+ ],
482
+ "mean": 0.013790103271484378,
483
+ "median": 0.014553988037109375,
484
+ "std": 0.0009817446377020434,
485
+ "min": 0.012528175048828125,
486
+ "max": 0.01466048828125,
487
+ "p25": 0.01265097900390625,
488
+ "p75": 0.014556885986328125,
489
+ "p90": 0.01461904736328125,
490
+ "p95": 0.014639767822265625,
491
+ "p99": 0.014656344189453125,
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": "2025-10-07T16:46:16.258305",
504
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 1.2086533203125,
538
+ 1.2092880859375,
539
+ 1.2239691162109374,
540
+ 1.209033203125,
541
+ 1.21711669921875
542
+ ],
543
+ "mean": 1.2136120849609375,
544
+ "median": 1.2092880859375,
545
+ "std": 0.006063050233509511,
546
+ "min": 1.2086533203125,
547
+ "max": 1.2239691162109374,
548
+ "p25": 1.209033203125,
549
+ "p75": 1.21711669921875,
550
+ "p90": 1.2212281494140624,
551
+ "p95": 1.2225986328125,
552
+ "p99": 1.22369501953125,
553
+ "unit": "seconds"
554
+ },
555
+ "time_to_first_token_seconds": {
556
+ "name": "time_to_first_token_seconds",
557
+ "measurements": [
558
+ 0.013097923278808593,
559
+ 0.01406294536590576,
560
+ 0.013600313186645507,
561
+ 0.01371675205230713,
562
+ 0.013695672035217285
563
+ ],
564
+ "mean": 0.013634721183776854,
565
+ "median": 0.013695672035217285,
566
+ "std": 0.0003108555680337801,
567
+ "min": 0.013097923278808593,
568
+ "max": 0.01406294536590576,
569
+ "p25": 0.013600313186645507,
570
+ "p75": 0.01371675205230713,
571
+ "p90": 0.013924468040466308,
572
+ "p95": 0.013993706703186035,
573
+ "p99": 0.014049097633361816,
574
+ "unit": "seconds"
575
+ },
576
+ "tokens_per_second": {
577
+ "name": "tokens_per_second",
578
+ "measurements": [
579
+ 82.73671061785093,
580
+ 82.69328141315066,
581
+ 81.70140788320849,
582
+ 82.71071442995033,
583
+ 82.16139016430272
584
+ ],
585
+ "mean": 82.40070090169262,
586
+ "median": 82.69328141315066,
587
+ "std": 0.4100962393076474,
588
+ "min": 81.70140788320849,
589
+ "max": 82.73671061785093,
590
+ "p25": 82.16139016430272,
591
+ "p75": 82.71071442995033,
592
+ "p90": 82.72631214269069,
593
+ "p95": 82.73151138027082,
594
+ "p99": 82.7356707703349,
595
+ "unit": "tokens/sec"
596
+ },
597
+ "time_per_output_token_seconds": {
598
+ "name": "time_per_output_token_seconds",
599
+ "measurements": [
600
+ 0.012086533203125001,
601
+ 0.012092880859374999,
602
+ 0.012239691162109373,
603
+ 0.01209033203125,
604
+ 0.012171166992187501
605
+ ],
606
+ "mean": 0.012136120849609374,
607
+ "median": 0.012092880859374999,
608
+ "std": 6.063050233509487e-05,
609
+ "min": 0.012086533203125001,
610
+ "max": 0.012239691162109373,
611
+ "p25": 0.01209033203125,
612
+ "p75": 0.012171166992187501,
613
+ "p90": 0.012212281494140624,
614
+ "p95": 0.012225986328124999,
615
+ "p99": 0.012236950195312499,
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": "2025-10-07T16:46:30.498651",
628
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 1.177861083984375,
662
+ 1.1721170654296875,
663
+ 1.1841051025390625,
664
+ 1.177828125,
665
+ 1.1835775146484375
666
+ ],
667
+ "mean": 1.1790977783203125,
668
+ "median": 1.177861083984375,
669
+ "std": 0.00440484093462776,
670
+ "min": 1.1721170654296875,
671
+ "max": 1.1841051025390625,
672
+ "p25": 1.177828125,
673
+ "p75": 1.1835775146484375,
674
+ "p90": 1.1838940673828124,
675
+ "p95": 1.1839995849609375,
676
+ "p99": 1.1840839990234375,
677
+ "unit": "seconds"
678
+ },
679
+ "time_to_first_token_seconds": {
680
+ "name": "time_to_first_token_seconds",
681
+ "measurements": [
682
+ 0.013169601440429687,
683
+ 0.013058803558349609,
684
+ 0.013212200164794921,
685
+ 0.013248080253601075,
686
+ 0.013171961784362793
687
+ ],
688
+ "mean": 0.013172129440307618,
689
+ "median": 0.013171961784362793,
690
+ "std": 6.359774117739436e-05,
691
+ "min": 0.013058803558349609,
692
+ "max": 0.013248080253601075,
693
+ "p25": 0.013169601440429687,
694
+ "p75": 0.013212200164794921,
695
+ "p90": 0.013233728218078613,
696
+ "p95": 0.013240904235839844,
697
+ "p99": 0.01324664505004883,
698
+ "unit": "seconds"
699
+ },
700
+ "tokens_per_second": {
701
+ "name": "tokens_per_second",
702
+ "measurements": [
703
+ 84.8996552816975,
704
+ 85.31571030692305,
705
+ 84.45196274010743,
706
+ 84.9020310157732,
707
+ 84.48960778855567
708
+ ],
709
+ "mean": 84.81179342661137,
710
+ "median": 84.8996552816975,
711
+ "std": 0.3171991357678093,
712
+ "min": 84.45196274010743,
713
+ "max": 85.31571030692305,
714
+ "p25": 84.48960778855567,
715
+ "p75": 84.9020310157732,
716
+ "p90": 85.15023859046312,
717
+ "p95": 85.23297444869308,
718
+ "p99": 85.29916313527706,
719
+ "unit": "tokens/sec"
720
+ },
721
+ "time_per_output_token_seconds": {
722
+ "name": "time_per_output_token_seconds",
723
+ "measurements": [
724
+ 0.01177861083984375,
725
+ 0.011721170654296875,
726
+ 0.011841051025390625,
727
+ 0.01177828125,
728
+ 0.011835775146484375
729
+ ],
730
+ "mean": 0.011790977783203125,
731
+ "median": 0.01177861083984375,
732
+ "std": 4.404840934627765e-05,
733
+ "min": 0.011721170654296875,
734
+ "max": 0.011841051025390625,
735
+ "p25": 0.01177828125,
736
+ "p75": 0.011835775146484375,
737
+ "p90": 0.011838940673828124,
738
+ "p95": 0.011839995849609375,
739
+ "p99": 0.011840839990234376,
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": "2025-10-07T16:46:45.783950",
752
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 1.4535340576171876,
786
+ 1.4410535888671876,
787
+ 1.3208526611328124,
788
+ 1.430260498046875,
789
+ 1.46521826171875
790
+ ],
791
+ "mean": 1.4221838134765625,
792
+ "median": 1.4410535888671876,
793
+ "std": 0.05200789408433984,
794
+ "min": 1.3208526611328124,
795
+ "max": 1.46521826171875,
796
+ "p25": 1.430260498046875,
797
+ "p75": 1.4535340576171876,
798
+ "p90": 1.460544580078125,
799
+ "p95": 1.4628814208984375,
800
+ "p99": 1.4647508935546873,
801
+ "unit": "seconds"
802
+ },
803
+ "time_to_first_token_seconds": {
804
+ "name": "time_to_first_token_seconds",
805
+ "measurements": [
806
+ 0.016759977340698243,
807
+ 0.015211125373840333,
808
+ 0.014032106399536132,
809
+ 0.013319639205932618,
810
+ 0.015094846725463866
811
+ ],
812
+ "mean": 0.01488353900909424,
813
+ "median": 0.015094846725463866,
814
+ "std": 0.001169932021326505,
815
+ "min": 0.013319639205932618,
816
+ "max": 0.016759977340698243,
817
+ "p25": 0.014032106399536132,
818
+ "p75": 0.015211125373840333,
819
+ "p90": 0.016140436553955077,
820
+ "p95": 0.01645020694732666,
821
+ "p99": 0.016698023262023926,
822
+ "unit": "seconds"
823
+ },
824
+ "tokens_per_second": {
825
+ "name": "tokens_per_second",
826
+ "measurements": [
827
+ 68.79783757109368,
828
+ 69.39367194429599,
829
+ 75.70867133222586,
830
+ 69.91733333651968,
831
+ 68.249217616696
832
+ ],
833
+ "mean": 70.41334636016624,
834
+ "median": 69.39367194429599,
835
+ "std": 2.7062724644651115,
836
+ "min": 68.249217616696,
837
+ "max": 75.70867133222586,
838
+ "p25": 68.79783757109368,
839
+ "p75": 69.91733333651968,
840
+ "p90": 73.39213613394338,
841
+ "p95": 74.55040373308462,
842
+ "p99": 75.47701781239762,
843
+ "unit": "tokens/sec"
844
+ },
845
+ "time_per_output_token_seconds": {
846
+ "name": "time_per_output_token_seconds",
847
+ "measurements": [
848
+ 0.014535340576171875,
849
+ 0.014410535888671876,
850
+ 0.013208526611328123,
851
+ 0.01430260498046875,
852
+ 0.014652182617187499
853
+ ],
854
+ "mean": 0.014221838134765625,
855
+ "median": 0.014410535888671876,
856
+ "std": 0.0005200789408433985,
857
+ "min": 0.013208526611328123,
858
+ "max": 0.014652182617187499,
859
+ "p25": 0.01430260498046875,
860
+ "p75": 0.014535340576171875,
861
+ "p90": 0.014605445800781248,
862
+ "p95": 0.014628814208984374,
863
+ "p99": 0.014647508935546874,
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": "2025-10-07T16:47:02.008820",
876
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 1.49470458984375,
910
+ 1.62195849609375,
911
+ 1.6966259765625,
912
+ 1.7188548583984375,
913
+ 1.7148966064453126
914
+ ],
915
+ "mean": 1.6494081054687502,
916
+ "median": 1.6966259765625,
917
+ "std": 0.08488554812785858,
918
+ "min": 1.49470458984375,
919
+ "max": 1.7188548583984375,
920
+ "p25": 1.62195849609375,
921
+ "p75": 1.7148966064453126,
922
+ "p90": 1.7172715576171875,
923
+ "p95": 1.7180632080078126,
924
+ "p99": 1.7186965283203126,
925
+ "unit": "seconds"
926
+ },
927
+ "time_to_first_token_seconds": {
928
+ "name": "time_to_first_token_seconds",
929
+ "measurements": [
930
+ 0.017068172454833986,
931
+ 0.01601987075805664,
932
+ 0.018238950729370117,
933
+ 0.018245630264282227,
934
+ 0.018222551345825195
935
+ ],
936
+ "mean": 0.017559035110473634,
937
+ "median": 0.018222551345825195,
938
+ "std": 0.0008926281419584272,
939
+ "min": 0.01601987075805664,
940
+ "max": 0.018245630264282227,
941
+ "p25": 0.017068172454833986,
942
+ "p75": 0.018238950729370117,
943
+ "p90": 0.018242958450317384,
944
+ "p95": 0.018244294357299806,
945
+ "p99": 0.018245363082885744,
946
+ "unit": "seconds"
947
+ },
948
+ "tokens_per_second": {
949
+ "name": "tokens_per_second",
950
+ "measurements": [
951
+ 66.90285202807436,
952
+ 61.65385873981078,
953
+ 58.940509800874324,
954
+ 58.17826881158315,
955
+ 58.31255343567499
956
+ ],
957
+ "mean": 60.797608563203525,
958
+ "median": 58.940509800874324,
959
+ "std": 3.3012851335383684,
960
+ "min": 58.17826881158315,
961
+ "max": 66.90285202807436,
962
+ "p25": 58.31255343567499,
963
+ "p75": 61.65385873981078,
964
+ "p90": 64.80325471276893,
965
+ "p95": 65.85305337042165,
966
+ "p99": 66.69289229654382,
967
+ "unit": "tokens/sec"
968
+ },
969
+ "time_per_output_token_seconds": {
970
+ "name": "time_per_output_token_seconds",
971
+ "measurements": [
972
+ 0.0149470458984375,
973
+ 0.016219584960937502,
974
+ 0.016966259765625002,
975
+ 0.017188548583984376,
976
+ 0.017148966064453125
977
+ ],
978
+ "mean": 0.016494081054687502,
979
+ "median": 0.016966259765625002,
980
+ "std": 0.000848855481278586,
981
+ "min": 0.0149470458984375,
982
+ "max": 0.017188548583984376,
983
+ "p25": 0.016219584960937502,
984
+ "p75": 0.017148966064453125,
985
+ "p90": 0.017172715576171876,
986
+ "p95": 0.017180632080078126,
987
+ "p99": 0.017186965283203126,
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": "2025-10-07T16:47:26.169132",
1000
+ "commit_id": "50090c3fc82e1e0a06b4da366ea2fb6055d529e9",
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
+ 1.3039649658203125,
1034
+ 1.37692822265625,
1035
+ 1.4580511474609374,
1036
+ 1.451307373046875,
1037
+ 1.4609283447265624
1038
+ ],
1039
+ "mean": 1.4102360107421876,
1040
+ "median": 1.451307373046875,
1041
+ "std": 0.061556154138711364,
1042
+ "min": 1.3039649658203125,
1043
+ "max": 1.4609283447265624,
1044
+ "p25": 1.37692822265625,
1045
+ "p75": 1.4580511474609374,
1046
+ "p90": 1.4597774658203124,
1047
+ "p95": 1.4603529052734374,
1048
+ "p99": 1.4608132568359373,
1049
+ "unit": "seconds"
1050
+ },
1051
+ "time_to_first_token_seconds": {
1052
+ "name": "time_to_first_token_seconds",
1053
+ "measurements": [
1054
+ 0.016150306701660157,
1055
+ 0.013551834106445313,
1056
+ 0.01530952262878418,
1057
+ 0.015415559768676758,
1058
+ 0.015149684906005859
1059
+ ],
1060
+ "mean": 0.015115381622314452,
1061
+ "median": 0.01530952262878418,
1062
+ "std": 0.000853782773245967,
1063
+ "min": 0.013551834106445313,
1064
+ "max": 0.016150306701660157,
1065
+ "p25": 0.015149684906005859,
1066
+ "p75": 0.015415559768676758,
1067
+ "p90": 0.015856407928466796,
1068
+ "p95": 0.016003357315063478,
1069
+ "p99": 0.01612091682434082,
1070
+ "unit": "seconds"
1071
+ },
1072
+ "tokens_per_second": {
1073
+ "name": "tokens_per_second",
1074
+ "measurements": [
1075
+ 76.68917694969734,
1076
+ 72.62542691374915,
1077
+ 68.58469963426238,
1078
+ 68.90339142290718,
1079
+ 68.44962681500762
1080
+ ],
1081
+ "mean": 71.05046434712474,
1082
+ "median": 68.90339142290718,
1083
+ "std": 3.2165155224327355,
1084
+ "min": 68.44962681500762,
1085
+ "max": 76.68917694969734,
1086
+ "p25": 68.58469963426238,
1087
+ "p75": 72.62542691374915,
1088
+ "p90": 75.06367693531807,
1089
+ "p95": 75.8764269425077,
1090
+ "p99": 76.52662694825942,
1091
+ "unit": "tokens/sec"
1092
+ },
1093
+ "time_per_output_token_seconds": {
1094
+ "name": "time_per_output_token_seconds",
1095
+ "measurements": [
1096
+ 0.013039649658203124,
1097
+ 0.0137692822265625,
1098
+ 0.014580511474609375,
1099
+ 0.014513073730468751,
1100
+ 0.014609283447265625
1101
+ ],
1102
+ "mean": 0.014102360107421875,
1103
+ "median": 0.014513073730468751,
1104
+ "std": 0.0006155615413871141,
1105
+ "min": 0.013039649658203124,
1106
+ "max": 0.014609283447265625,
1107
+ "p25": 0.0137692822265625,
1108
+ "p75": 0.014580511474609375,
1109
+ "p90": 0.014597774658203125,
1110
+ "p95": 0.014603529052734375,
1111
+ "p99": 0.014608132568359375,
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
+ }