hf-transformers-bot commited on
Commit
6d85e11
·
verified ·
1 Parent(s): ed148df

Upload benchmark results for run 18446640769

Browse files
2025-10-12/18446640769/benchmark_results/Llama-2-7b-hf/Llama-2-7b-hf_benchmark_20251012_193159.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": "2025-10-12T19:28:50.044113",
8
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
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
+ 1.4468614501953125,
42
+ 1.4364541015625,
43
+ 1.438771240234375,
44
+ 1.4411251220703125,
45
+ 1.4238294677734376
46
+ ],
47
+ "mean": 1.4374082763671876,
48
+ "median": 1.438771240234375,
49
+ "std": 0.007620099740717883,
50
+ "min": 1.4238294677734376,
51
+ "max": 1.4468614501953125,
52
+ "p25": 1.4364541015625,
53
+ "p75": 1.4411251220703125,
54
+ "p90": 1.4445669189453125,
55
+ "p95": 1.4457141845703125,
56
+ "p99": 1.4466319970703125,
57
+ "unit": "seconds"
58
+ },
59
+ "time_to_first_token_seconds": {
60
+ "name": "time_to_first_token_seconds",
61
+ "measurements": [
62
+ 0.016535646438598634,
63
+ 0.015433306694030762,
64
+ 0.015429428100585937,
65
+ 0.015272191047668457,
66
+ 0.015235914230346679
67
+ ],
68
+ "mean": 0.015581297302246094,
69
+ "median": 0.015429428100585937,
70
+ "std": 0.00048385615456731924,
71
+ "min": 0.015235914230346679,
72
+ "max": 0.016535646438598634,
73
+ "p25": 0.015272191047668457,
74
+ "p75": 0.015433306694030762,
75
+ "p90": 0.016094710540771485,
76
+ "p95": 0.01631517848968506,
77
+ "p99": 0.01649155284881592,
78
+ "unit": "seconds"
79
+ },
80
+ "tokens_per_second": {
81
+ "name": "tokens_per_second",
82
+ "measurements": [
83
+ 69.11511809683017,
84
+ 69.61586861092547,
85
+ 69.50375237116225,
86
+ 69.39022744696904,
87
+ 70.23312992417445
88
+ ],
89
+ "mean": 69.57161929001226,
90
+ "median": 69.50375237116225,
91
+ "std": 0.3702584449501247,
92
+ "min": 69.11511809683017,
93
+ "max": 70.23312992417445,
94
+ "p25": 69.39022744696904,
95
+ "p75": 69.61586861092547,
96
+ "p90": 69.98622539887486,
97
+ "p95": 70.10967766152466,
98
+ "p99": 70.2084394716445,
99
+ "unit": "tokens/sec"
100
+ },
101
+ "time_per_output_token_seconds": {
102
+ "name": "time_per_output_token_seconds",
103
+ "measurements": [
104
+ 0.014468614501953126,
105
+ 0.014364541015625001,
106
+ 0.01438771240234375,
107
+ 0.014411251220703126,
108
+ 0.014238294677734376
109
+ ],
110
+ "mean": 0.014374082763671877,
111
+ "median": 0.01438771240234375,
112
+ "std": 7.620099740717894e-05,
113
+ "min": 0.014238294677734376,
114
+ "max": 0.014468614501953126,
115
+ "p25": 0.014364541015625001,
116
+ "p75": 0.014411251220703126,
117
+ "p90": 0.014445669189453126,
118
+ "p95": 0.014457141845703125,
119
+ "p99": 0.014466319970703126,
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": "2025-10-12T19:29:07.368213",
132
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
133
+ "hardware_info": {
134
+ "gpu_name": "unknown",
135
+ "gpu_memory_total_mb": 0,
136
+ "cpu_count": 160,
137
+ "memory_total_mb": 1285717,
138
+ "python_version": "3.12.10",
139
+ "torch_version": "2.7.1+rocm6.4.1.git2a215e4a",
140
+ "cuda_version": null
141
+ },
142
+ "config": {
143
+ "name": "compiled",
144
+ "model_id": "meta-llama/Llama-2-7b-hf",
145
+ "variant": "compiled",
146
+ "warmup_iterations": 3,
147
+ "measurement_iterations": 5,
148
+ "num_tokens_to_generate": 100,
149
+ "device": "cuda",
150
+ "torch_dtype": "float16",
151
+ "compile_mode": "max-autotune",
152
+ "compile_options": {},
153
+ "use_cache": true,
154
+ "batch_size": 1,
155
+ "sequence_length": null,
156
+ "attn_implementation": "eager",
157
+ "sdpa_backend": null,
158
+ "custom_params": {}
159
+ }
160
+ },
161
+ "measurements": {
162
+ "latency_seconds": {
163
+ "name": "latency_seconds",
164
+ "measurements": [
165
+ 3.001444091796875,
166
+ 3.090043212890625,
167
+ 3.115807373046875,
168
+ 3.20263916015625,
169
+ 3.162328125
170
+ ],
171
+ "mean": 3.114452392578125,
172
+ "median": 3.115807373046875,
173
+ "std": 0.06846533692513611,
174
+ "min": 3.001444091796875,
175
+ "max": 3.20263916015625,
176
+ "p25": 3.090043212890625,
177
+ "p75": 3.162328125,
178
+ "p90": 3.18651474609375,
179
+ "p95": 3.194576953125,
180
+ "p99": 3.20102671875,
181
+ "unit": "seconds"
182
+ },
183
+ "time_to_first_token_seconds": {
184
+ "name": "time_to_first_token_seconds",
185
+ "measurements": [
186
+ 0.0161453742980957,
187
+ 0.018405851364135743,
188
+ 0.019139116287231445,
189
+ 0.01919559669494629,
190
+ 0.019098278045654296
191
+ ],
192
+ "mean": 0.018396843338012694,
193
+ "median": 0.019098278045654296,
194
+ "std": 0.0011619105548774595,
195
+ "min": 0.0161453742980957,
196
+ "max": 0.01919559669494629,
197
+ "p25": 0.018405851364135743,
198
+ "p75": 0.019139116287231445,
199
+ "p90": 0.01917300453186035,
200
+ "p95": 0.019184300613403323,
201
+ "p99": 0.019193337478637697,
202
+ "unit": "seconds"
203
+ },
204
+ "tokens_per_second": {
205
+ "name": "tokens_per_second",
206
+ "measurements": [
207
+ 33.31729558891533,
208
+ 32.362006972211105,
209
+ 32.094410220941334,
210
+ 31.22424819008371,
211
+ 31.622271961420825
212
+ ],
213
+ "mean": 32.124046586714464,
214
+ "median": 32.094410220941334,
215
+ "std": 0.7131285395583542,
216
+ "min": 31.22424819008371,
217
+ "max": 33.31729558891533,
218
+ "p25": 31.622271961420825,
219
+ "p75": 32.362006972211105,
220
+ "p90": 32.93518014223364,
221
+ "p95": 33.12623786557448,
222
+ "p99": 33.279084044247156,
223
+ "unit": "tokens/sec"
224
+ },
225
+ "time_per_output_token_seconds": {
226
+ "name": "time_per_output_token_seconds",
227
+ "measurements": [
228
+ 0.030014440917968753,
229
+ 0.03090043212890625,
230
+ 0.03115807373046875,
231
+ 0.0320263916015625,
232
+ 0.03162328125
233
+ ],
234
+ "mean": 0.031144523925781247,
235
+ "median": 0.03115807373046875,
236
+ "std": 0.0006846533692513614,
237
+ "min": 0.030014440917968753,
238
+ "max": 0.0320263916015625,
239
+ "p25": 0.03090043212890625,
240
+ "p75": 0.03162328125,
241
+ "p90": 0.0318651474609375,
242
+ "p95": 0.03194576953125,
243
+ "p99": 0.032010267187500004,
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": "2025-10-12T19:29:47.956258",
256
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
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
+ 1.3920032958984374,
290
+ 1.4000902099609376,
291
+ 1.4073203125,
292
+ 1.3969749755859375,
293
+ 1.396781494140625
294
+ ],
295
+ "mean": 1.3986340576171874,
296
+ "median": 1.3969749755859375,
297
+ "std": 0.005054165591457105,
298
+ "min": 1.3920032958984374,
299
+ "max": 1.4073203125,
300
+ "p25": 1.396781494140625,
301
+ "p75": 1.4000902099609376,
302
+ "p90": 1.404428271484375,
303
+ "p95": 1.4058742919921876,
304
+ "p99": 1.4070311083984375,
305
+ "unit": "seconds"
306
+ },
307
+ "time_to_first_token_seconds": {
308
+ "name": "time_to_first_token_seconds",
309
+ "measurements": [
310
+ 0.014984875679016114,
311
+ 0.015182592391967774,
312
+ 0.015188632011413574,
313
+ 0.01516703701019287,
314
+ 0.015216631889343262
315
+ ],
316
+ "mean": 0.015147953796386717,
317
+ "median": 0.015182592391967774,
318
+ "std": 8.310219172568018e-05,
319
+ "min": 0.014984875679016114,
320
+ "max": 0.015216631889343262,
321
+ "p25": 0.01516703701019287,
322
+ "p75": 0.015188632011413574,
323
+ "p90": 0.015205431938171386,
324
+ "p95": 0.015211031913757325,
325
+ "p99": 0.015215511894226074,
326
+ "unit": "seconds"
327
+ },
328
+ "tokens_per_second": {
329
+ "name": "tokens_per_second",
330
+ "measurements": [
331
+ 71.83891036368361,
332
+ 71.4239691760933,
333
+ 71.05702881695598,
334
+ 71.58324361397862,
335
+ 71.5931592876131
336
+ ],
337
+ "mean": 71.49926225166493,
338
+ "median": 71.58324361397862,
339
+ "std": 0.2578642358208226,
340
+ "min": 71.05702881695598,
341
+ "max": 71.83891036368361,
342
+ "p25": 71.4239691760933,
343
+ "p75": 71.5931592876131,
344
+ "p90": 71.74060993325541,
345
+ "p95": 71.78976014846951,
346
+ "p99": 71.82908032064078,
347
+ "unit": "tokens/sec"
348
+ },
349
+ "time_per_output_token_seconds": {
350
+ "name": "time_per_output_token_seconds",
351
+ "measurements": [
352
+ 0.013920032958984373,
353
+ 0.014000902099609375,
354
+ 0.014073203125,
355
+ 0.013969749755859375,
356
+ 0.013967814941406248
357
+ ],
358
+ "mean": 0.013986340576171877,
359
+ "median": 0.013969749755859375,
360
+ "std": 5.054165591457104e-05,
361
+ "min": 0.013920032958984373,
362
+ "max": 0.014073203125,
363
+ "p25": 0.013967814941406248,
364
+ "p75": 0.014000902099609375,
365
+ "p90": 0.01404428271484375,
366
+ "p95": 0.014058742919921874,
367
+ "p99": 0.014070311083984374,
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": "2025-10-12T19:30:05.121045",
380
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
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
+ 1.742448974609375,
414
+ 1.7327041015625,
415
+ 1.7389219970703125,
416
+ 1.7449150390625,
417
+ 1.748162841796875
418
+ ],
419
+ "mean": 1.7414305908203123,
420
+ "median": 1.742448974609375,
421
+ "std": 0.005309356753411705,
422
+ "min": 1.7327041015625,
423
+ "max": 1.748162841796875,
424
+ "p25": 1.7389219970703125,
425
+ "p75": 1.7449150390625,
426
+ "p90": 1.746863720703125,
427
+ "p95": 1.74751328125,
428
+ "p99": 1.7480329296875001,
429
+ "unit": "seconds"
430
+ },
431
+ "time_to_first_token_seconds": {
432
+ "name": "time_to_first_token_seconds",
433
+ "measurements": [
434
+ 0.019032760620117186,
435
+ 0.019501869201660155,
436
+ 0.019556947708129883,
437
+ 0.019588228225708007,
438
+ 0.019564512252807618
439
+ ],
440
+ "mean": 0.019448863601684567,
441
+ "median": 0.019556947708129883,
442
+ "std": 0.00020996693449202642,
443
+ "min": 0.019032760620117186,
444
+ "max": 0.019588228225708007,
445
+ "p25": 0.019501869201660155,
446
+ "p75": 0.019564512252807618,
447
+ "p90": 0.019578741836547852,
448
+ "p95": 0.019583485031127928,
449
+ "p99": 0.019587279586791993,
450
+ "unit": "seconds"
451
+ },
452
+ "tokens_per_second": {
453
+ "name": "tokens_per_second",
454
+ "measurements": [
455
+ 57.39048974011888,
456
+ 57.71325866304756,
457
+ 57.50689229791631,
458
+ 57.30938054939772,
459
+ 57.202909024890104
460
+ ],
461
+ "mean": 57.42458605507411,
462
+ "median": 57.39048974011888,
463
+ "std": 0.1753178628365472,
464
+ "min": 57.202909024890104,
465
+ "max": 57.71325866304756,
466
+ "p25": 57.30938054939772,
467
+ "p75": 57.50689229791631,
468
+ "p90": 57.63071211699506,
469
+ "p95": 57.67198539002131,
470
+ "p99": 57.70500400844231,
471
+ "unit": "tokens/sec"
472
+ },
473
+ "time_per_output_token_seconds": {
474
+ "name": "time_per_output_token_seconds",
475
+ "measurements": [
476
+ 0.01742448974609375,
477
+ 0.017327041015625,
478
+ 0.017389219970703126,
479
+ 0.017449150390624998,
480
+ 0.017481628417968752
481
+ ],
482
+ "mean": 0.017414305908203125,
483
+ "median": 0.01742448974609375,
484
+ "std": 5.3093567534117366e-05,
485
+ "min": 0.017327041015625,
486
+ "max": 0.017481628417968752,
487
+ "p25": 0.017389219970703126,
488
+ "p75": 0.017449150390624998,
489
+ "p90": 0.017468637207031252,
490
+ "p95": 0.017475132812500002,
491
+ "p99": 0.017480329296875003,
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": "2025-10-12T19:30:25.767251",
504
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
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
+ 1.224692626953125,
538
+ 1.230744873046875,
539
+ 1.2226318359375,
540
+ 1.230387939453125,
541
+ 1.2419481201171876
542
+ ],
543
+ "mean": 1.2300810791015624,
544
+ "median": 1.230387939453125,
545
+ "std": 0.006721429735064641,
546
+ "min": 1.2226318359375,
547
+ "max": 1.2419481201171876,
548
+ "p25": 1.224692626953125,
549
+ "p75": 1.230744873046875,
550
+ "p90": 1.2374668212890625,
551
+ "p95": 1.239707470703125,
552
+ "p99": 1.241499990234375,
553
+ "unit": "seconds"
554
+ },
555
+ "time_to_first_token_seconds": {
556
+ "name": "time_to_first_token_seconds",
557
+ "measurements": [
558
+ 0.013744298934936523,
559
+ 0.013745698928833008,
560
+ 0.013688020706176758,
561
+ 0.013625461578369141,
562
+ 0.014534845352172851
563
+ ],
564
+ "mean": 0.013867665100097657,
565
+ "median": 0.013744298934936523,
566
+ "std": 0.0003364993612485897,
567
+ "min": 0.013625461578369141,
568
+ "max": 0.014534845352172851,
569
+ "p25": 0.013688020706176758,
570
+ "p75": 0.013745698928833008,
571
+ "p90": 0.014219186782836913,
572
+ "p95": 0.014377016067504883,
573
+ "p99": 0.014503279495239257,
574
+ "unit": "seconds"
575
+ },
576
+ "tokens_per_second": {
577
+ "name": "tokens_per_second",
578
+ "measurements": [
579
+ 81.65314120391736,
580
+ 81.25160802208869,
581
+ 81.7907705824797,
582
+ 81.27517898496906,
583
+ 80.51866127110382
584
+ ],
585
+ "mean": 81.29787201291172,
586
+ "median": 81.27517898496906,
587
+ "std": 0.4424906571833688,
588
+ "min": 80.51866127110382,
589
+ "max": 81.7907705824797,
590
+ "p25": 81.25160802208869,
591
+ "p75": 81.65314120391736,
592
+ "p90": 81.73571883105475,
593
+ "p95": 81.76324470676722,
594
+ "p99": 81.7852654073372,
595
+ "unit": "tokens/sec"
596
+ },
597
+ "time_per_output_token_seconds": {
598
+ "name": "time_per_output_token_seconds",
599
+ "measurements": [
600
+ 0.012246926269531251,
601
+ 0.01230744873046875,
602
+ 0.012226318359374998,
603
+ 0.01230387939453125,
604
+ 0.012419481201171876
605
+ ],
606
+ "mean": 0.012300810791015627,
607
+ "median": 0.01230387939453125,
608
+ "std": 6.721429735064638e-05,
609
+ "min": 0.012226318359374998,
610
+ "max": 0.012419481201171876,
611
+ "p25": 0.012246926269531251,
612
+ "p75": 0.01230744873046875,
613
+ "p90": 0.012374668212890625,
614
+ "p95": 0.01239707470703125,
615
+ "p99": 0.012414999902343751,
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": "2025-10-12T19:30:41.370619",
628
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
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
+ 1.3842001953125,
662
+ 1.38832568359375,
663
+ 1.381914794921875,
664
+ 1.3817178955078124,
665
+ 1.3813216552734375
666
+ ],
667
+ "mean": 1.3834960449218747,
668
+ "median": 1.381914794921875,
669
+ "std": 0.0026157742531782643,
670
+ "min": 1.3813216552734375,
671
+ "max": 1.38832568359375,
672
+ "p25": 1.3817178955078124,
673
+ "p75": 1.3842001953125,
674
+ "p90": 1.38667548828125,
675
+ "p95": 1.3875005859375,
676
+ "p99": 1.3881606640625002,
677
+ "unit": "seconds"
678
+ },
679
+ "time_to_first_token_seconds": {
680
+ "name": "time_to_first_token_seconds",
681
+ "measurements": [
682
+ 0.015022676467895507,
683
+ 0.015092912673950196,
684
+ 0.015317952156066895,
685
+ 0.015087233543395997,
686
+ 0.015217310905456543
687
+ ],
688
+ "mean": 0.015147617149353027,
689
+ "median": 0.015092912673950196,
690
+ "std": 0.00010594353923913614,
691
+ "min": 0.015022676467895507,
692
+ "max": 0.015317952156066895,
693
+ "p25": 0.015087233543395997,
694
+ "p75": 0.015217310905456543,
695
+ "p90": 0.015277695655822755,
696
+ "p95": 0.015297823905944825,
697
+ "p99": 0.01531392650604248,
698
+ "unit": "seconds"
699
+ },
700
+ "tokens_per_second": {
701
+ "name": "tokens_per_second",
702
+ "measurements": [
703
+ 72.24388519712915,
704
+ 72.02920840673711,
705
+ 72.36336159614919,
706
+ 72.37367361681868,
707
+ 72.39443443041125
708
+ ],
709
+ "mean": 72.28091264944908,
710
+ "median": 72.36336159614919,
711
+ "std": 0.13639331263418658,
712
+ "min": 72.02920840673711,
713
+ "max": 72.39443443041125,
714
+ "p25": 72.24388519712915,
715
+ "p75": 72.37367361681868,
716
+ "p90": 72.38613010497423,
717
+ "p95": 72.39028226769274,
718
+ "p99": 72.39360399786754,
719
+ "unit": "tokens/sec"
720
+ },
721
+ "time_per_output_token_seconds": {
722
+ "name": "time_per_output_token_seconds",
723
+ "measurements": [
724
+ 0.013842001953125,
725
+ 0.013883256835937502,
726
+ 0.01381914794921875,
727
+ 0.013817178955078124,
728
+ 0.013813216552734375
729
+ ],
730
+ "mean": 0.01383496044921875,
731
+ "median": 0.01381914794921875,
732
+ "std": 2.615774253178301e-05,
733
+ "min": 0.013813216552734375,
734
+ "max": 0.013883256835937502,
735
+ "p25": 0.013817178955078124,
736
+ "p75": 0.013842001953125,
737
+ "p90": 0.0138667548828125,
738
+ "p95": 0.013875005859375001,
739
+ "p99": 0.013881606640625002,
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": "2025-10-12T19:30:58.237821",
752
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
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
+ 1.6537147216796875,
786
+ 1.40653173828125,
787
+ 1.32960791015625,
788
+ 1.352843505859375,
789
+ 1.3442476806640624
790
+ ],
791
+ "mean": 1.417389111328125,
792
+ "median": 1.352843505859375,
793
+ "std": 0.12098679266895881,
794
+ "min": 1.32960791015625,
795
+ "max": 1.6537147216796875,
796
+ "p25": 1.3442476806640624,
797
+ "p75": 1.40653173828125,
798
+ "p90": 1.5548415283203125,
799
+ "p95": 1.604278125,
800
+ "p99": 1.64382740234375,
801
+ "unit": "seconds"
802
+ },
803
+ "time_to_first_token_seconds": {
804
+ "name": "time_to_first_token_seconds",
805
+ "measurements": [
806
+ 0.017392311096191406,
807
+ 0.017001476287841797,
808
+ 0.014034692764282226,
809
+ 0.014177010536193848,
810
+ 0.01425404930114746
811
+ ],
812
+ "mean": 0.015371907997131345,
813
+ "median": 0.01425404930114746,
814
+ "std": 0.0014968670002432332,
815
+ "min": 0.014034692764282226,
816
+ "max": 0.017392311096191406,
817
+ "p25": 0.014177010536193848,
818
+ "p75": 0.017001476287841797,
819
+ "p90": 0.017235977172851562,
820
+ "p95": 0.017314144134521484,
821
+ "p99": 0.01737667770385742,
822
+ "unit": "seconds"
823
+ },
824
+ "tokens_per_second": {
825
+ "name": "tokens_per_second",
826
+ "measurements": [
827
+ 60.46992186078469,
828
+ 71.09686705128868,
829
+ 75.21014220519223,
830
+ 73.91837974376526,
831
+ 74.39105265972987
832
+ ],
833
+ "mean": 71.01727270415215,
834
+ "median": 73.91837974376526,
835
+ "std": 5.452190623407865,
836
+ "min": 60.46992186078469,
837
+ "max": 75.21014220519223,
838
+ "p25": 71.09686705128868,
839
+ "p75": 74.39105265972987,
840
+ "p90": 74.88250638700728,
841
+ "p95": 75.04632429609975,
842
+ "p99": 75.17737862337373,
843
+ "unit": "tokens/sec"
844
+ },
845
+ "time_per_output_token_seconds": {
846
+ "name": "time_per_output_token_seconds",
847
+ "measurements": [
848
+ 0.016537147216796874,
849
+ 0.014065317382812501,
850
+ 0.013296079101562499,
851
+ 0.013528435058593751,
852
+ 0.013442476806640624
853
+ ],
854
+ "mean": 0.01417389111328125,
855
+ "median": 0.013528435058593751,
856
+ "std": 0.001209867926689588,
857
+ "min": 0.013296079101562499,
858
+ "max": 0.016537147216796874,
859
+ "p25": 0.013442476806640624,
860
+ "p75": 0.014065317382812501,
861
+ "p90": 0.015548415283203125,
862
+ "p95": 0.01604278125,
863
+ "p99": 0.016438274023437498,
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": "2025-10-12T19:31:16.557112",
876
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
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
+ 1.9175531005859374,
910
+ 1.9129412841796876,
911
+ 1.9080184326171874,
912
+ 1.9143787841796875,
913
+ 1.91680126953125
914
+ ],
915
+ "mean": 1.91393857421875,
916
+ "median": 1.9143787841796875,
917
+ "std": 0.0033910551197400595,
918
+ "min": 1.9080184326171874,
919
+ "max": 1.9175531005859374,
920
+ "p25": 1.9129412841796876,
921
+ "p75": 1.91680126953125,
922
+ "p90": 1.9172523681640625,
923
+ "p95": 1.917402734375,
924
+ "p99": 1.91752302734375,
925
+ "unit": "seconds"
926
+ },
927
+ "time_to_first_token_seconds": {
928
+ "name": "time_to_first_token_seconds",
929
+ "measurements": [
930
+ 0.020007579803466798,
931
+ 0.020554569244384764,
932
+ 0.019915382385253907,
933
+ 0.020048460006713866,
934
+ 0.02022429656982422
935
+ ],
936
+ "mean": 0.02015005760192871,
937
+ "median": 0.020048460006713866,
938
+ "std": 0.00022575896215479152,
939
+ "min": 0.019915382385253907,
940
+ "max": 0.020554569244384764,
941
+ "p25": 0.020007579803466798,
942
+ "p75": 0.02022429656982422,
943
+ "p90": 0.020422460174560546,
944
+ "p95": 0.020488514709472654,
945
+ "p99": 0.02054135833740234,
946
+ "unit": "seconds"
947
+ },
948
+ "tokens_per_second": {
949
+ "name": "tokens_per_second",
950
+ "measurements": [
951
+ 52.14979442782757,
952
+ 52.275519811828545,
953
+ 52.410395146357246,
954
+ 52.23626631594231,
955
+ 52.17024925304583
956
+ ],
957
+ "mean": 52.248444991000305,
958
+ "median": 52.23626631594231,
959
+ "std": 0.0926884306887637,
960
+ "min": 52.14979442782757,
961
+ "max": 52.410395146357246,
962
+ "p25": 52.17024925304583,
963
+ "p75": 52.275519811828545,
964
+ "p90": 52.35644501254576,
965
+ "p95": 52.383420079451504,
966
+ "p99": 52.4050001329761,
967
+ "unit": "tokens/sec"
968
+ },
969
+ "time_per_output_token_seconds": {
970
+ "name": "time_per_output_token_seconds",
971
+ "measurements": [
972
+ 0.019175531005859375,
973
+ 0.019129412841796877,
974
+ 0.019080184326171874,
975
+ 0.019143787841796874,
976
+ 0.0191680126953125
977
+ ],
978
+ "mean": 0.0191393857421875,
979
+ "median": 0.019143787841796874,
980
+ "std": 3.391055119740056e-05,
981
+ "min": 0.019080184326171874,
982
+ "max": 0.019175531005859375,
983
+ "p25": 0.019129412841796877,
984
+ "p75": 0.0191680126953125,
985
+ "p90": 0.019172523681640626,
986
+ "p95": 0.01917402734375,
987
+ "p99": 0.0191752302734375,
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": "2025-10-12T19:31:44.634700",
1000
+ "commit_id": "3927ffed31e3c0d2929bf98bd05b7c61fcc48b62",
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
+ 1.62141650390625,
1034
+ 1.6196143798828124,
1035
+ 1.6197181396484375,
1036
+ 1.6232958984375,
1037
+ 1.6195997314453126
1038
+ ],
1039
+ "mean": 1.6207289306640624,
1040
+ "median": 1.6197181396484375,
1041
+ "std": 0.0014560970647672093,
1042
+ "min": 1.6195997314453126,
1043
+ "max": 1.6232958984375,
1044
+ "p25": 1.6196143798828124,
1045
+ "p75": 1.62141650390625,
1046
+ "p90": 1.622544140625,
1047
+ "p95": 1.62292001953125,
1048
+ "p99": 1.62322072265625,
1049
+ "unit": "seconds"
1050
+ },
1051
+ "time_to_first_token_seconds": {
1052
+ "name": "time_to_first_token_seconds",
1053
+ "measurements": [
1054
+ 0.016825319290161133,
1055
+ 0.01713651657104492,
1056
+ 0.016967517852783203,
1057
+ 0.01710011672973633,
1058
+ 0.016902559280395506
1059
+ ],
1060
+ "mean": 0.01698640594482422,
1061
+ "median": 0.016967517852783203,
1062
+ "std": 0.000117302409333325,
1063
+ "min": 0.016825319290161133,
1064
+ "max": 0.01713651657104492,
1065
+ "p25": 0.016902559280395506,
1066
+ "p75": 0.01710011672973633,
1067
+ "p90": 0.017121956634521485,
1068
+ "p95": 0.0171292366027832,
1069
+ "p99": 0.017135060577392575,
1070
+ "unit": "seconds"
1071
+ },
1072
+ "tokens_per_second": {
1073
+ "name": "tokens_per_second",
1074
+ "measurements": [
1075
+ 61.67446782432775,
1076
+ 61.74309220892168,
1077
+ 61.73913692273964,
1078
+ 61.6030633085778,
1079
+ 61.74365064308891
1080
+ ],
1081
+ "mean": 61.70068218153116,
1082
+ "median": 61.73913692273964,
1083
+ "std": 0.05539007693366703,
1084
+ "min": 61.6030633085778,
1085
+ "max": 61.74365064308891,
1086
+ "p25": 61.67446782432775,
1087
+ "p75": 61.74309220892168,
1088
+ "p90": 61.74342726942202,
1089
+ "p95": 61.74353895625546,
1090
+ "p99": 61.743628305722225,
1091
+ "unit": "tokens/sec"
1092
+ },
1093
+ "time_per_output_token_seconds": {
1094
+ "name": "time_per_output_token_seconds",
1095
+ "measurements": [
1096
+ 0.0162141650390625,
1097
+ 0.016196143798828125,
1098
+ 0.016197181396484375,
1099
+ 0.016232958984375,
1100
+ 0.016195997314453127
1101
+ ],
1102
+ "mean": 0.016207289306640626,
1103
+ "median": 0.016197181396484375,
1104
+ "std": 1.4560970647671435e-05,
1105
+ "min": 0.016195997314453127,
1106
+ "max": 0.016232958984375,
1107
+ "p25": 0.016196143798828125,
1108
+ "p75": 0.0162141650390625,
1109
+ "p90": 0.01622544140625,
1110
+ "p95": 0.0162292001953125,
1111
+ "p99": 0.0162322072265625,
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
+ }