hf-transformers-bot commited on
Commit
86d60d5
·
verified ·
1 Parent(s): 39ff5bc

Upload benchmark results for run 18199475800

Browse files
2025-10-02/18199475800/benchmark_results/Llama-2-7b-hf/Llama-2-7b-hf_benchmark_20251002_164841.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": "2025-10-02T16:45:28.978282",
8
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 1.504121337890625,
42
+ 1.5865084228515625,
43
+ 1.62937841796875,
44
+ 1.5698193359375,
45
+ 1.57917529296875
46
+ ],
47
+ "mean": 1.5738005615234374,
48
+ "median": 1.57917529296875,
49
+ "std": 0.04037405099302636,
50
+ "min": 1.504121337890625,
51
+ "max": 1.62937841796875,
52
+ "p25": 1.5698193359375,
53
+ "p75": 1.5865084228515625,
54
+ "p90": 1.612230419921875,
55
+ "p95": 1.6208044189453126,
56
+ "p99": 1.6276636181640625,
57
+ "unit": "seconds"
58
+ },
59
+ "time_to_first_token_seconds": {
60
+ "name": "time_to_first_token_seconds",
61
+ "measurements": [
62
+ 0.015955239295959472,
63
+ 0.01636355781555176,
64
+ 0.017672096252441406,
65
+ 0.01673902130126953,
66
+ 0.01667666244506836
67
+ ],
68
+ "mean": 0.016681315422058102,
69
+ "median": 0.01667666244506836,
70
+ "std": 0.000568006110870032,
71
+ "min": 0.015955239295959472,
72
+ "max": 0.017672096252441406,
73
+ "p25": 0.01636355781555176,
74
+ "p75": 0.01673902130126953,
75
+ "p90": 0.017298866271972656,
76
+ "p95": 0.01748548126220703,
77
+ "p99": 0.017634773254394533,
78
+ "unit": "seconds"
79
+ },
80
+ "tokens_per_second": {
81
+ "name": "tokens_per_second",
82
+ "measurements": [
83
+ 66.48399798665159,
84
+ 63.031496435588885,
85
+ 61.373097186756716,
86
+ 63.70159782767598,
87
+ 63.32419234599745
88
+ ],
89
+ "mean": 63.58287635653412,
90
+ "median": 63.32419234599745,
91
+ "std": 1.654367716933363,
92
+ "min": 61.373097186756716,
93
+ "max": 66.48399798665159,
94
+ "p25": 63.031496435588885,
95
+ "p75": 63.70159782767598,
96
+ "p90": 65.37103792306135,
97
+ "p95": 65.92751795485647,
98
+ "p99": 66.37270198029256,
99
+ "unit": "tokens/sec"
100
+ },
101
+ "time_per_output_token_seconds": {
102
+ "name": "time_per_output_token_seconds",
103
+ "measurements": [
104
+ 0.015041213378906251,
105
+ 0.015865084228515625,
106
+ 0.0162937841796875,
107
+ 0.015698193359374998,
108
+ 0.0157917529296875
109
+ ],
110
+ "mean": 0.015738005615234373,
111
+ "median": 0.0157917529296875,
112
+ "std": 0.0004037405099302632,
113
+ "min": 0.015041213378906251,
114
+ "max": 0.0162937841796875,
115
+ "p25": 0.015698193359374998,
116
+ "p75": 0.015865084228515625,
117
+ "p90": 0.01612230419921875,
118
+ "p95": 0.016208044189453125,
119
+ "p99": 0.016276636181640623,
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": "2025-10-02T16:45:46.669294",
132
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 3.182198486328125,
166
+ 3.213540771484375,
167
+ 3.303626220703125,
168
+ 3.268472900390625,
169
+ 3.290274658203125
170
+ ],
171
+ "mean": 3.251622607421875,
172
+ "median": 3.268472900390625,
173
+ "std": 0.04637288866362582,
174
+ "min": 3.182198486328125,
175
+ "max": 3.303626220703125,
176
+ "p25": 3.213540771484375,
177
+ "p75": 3.290274658203125,
178
+ "p90": 3.298285595703125,
179
+ "p95": 3.3009559082031252,
180
+ "p99": 3.303092158203125,
181
+ "unit": "seconds"
182
+ },
183
+ "time_to_first_token_seconds": {
184
+ "name": "time_to_first_token_seconds",
185
+ "measurements": [
186
+ 0.017514572143554688,
187
+ 0.017442811965942384,
188
+ 0.021741796493530273,
189
+ 0.02116944122314453,
190
+ 0.018107521057128907
191
+ ],
192
+ "mean": 0.019195228576660155,
193
+ "median": 0.018107521057128907,
194
+ "std": 0.0018687536702558863,
195
+ "min": 0.017442811965942384,
196
+ "max": 0.021741796493530273,
197
+ "p25": 0.017514572143554688,
198
+ "p75": 0.02116944122314453,
199
+ "p90": 0.021512854385375976,
200
+ "p95": 0.021627325439453126,
201
+ "p99": 0.021718902282714843,
202
+ "unit": "seconds"
203
+ },
204
+ "tokens_per_second": {
205
+ "name": "tokens_per_second",
206
+ "measurements": [
207
+ 31.42481540030773,
208
+ 31.118323093130925,
209
+ 30.269768224177785,
210
+ 30.595327863372738,
211
+ 30.392599520737793
212
+ ],
213
+ "mean": 30.760166820345397,
214
+ "median": 30.595327863372738,
215
+ "std": 0.441090021465322,
216
+ "min": 30.269768224177785,
217
+ "max": 31.42481540030773,
218
+ "p25": 30.392599520737793,
219
+ "p75": 31.118323093130925,
220
+ "p90": 31.30221847743701,
221
+ "p95": 31.36351693887237,
222
+ "p99": 31.41255570802066,
223
+ "unit": "tokens/sec"
224
+ },
225
+ "time_per_output_token_seconds": {
226
+ "name": "time_per_output_token_seconds",
227
+ "measurements": [
228
+ 0.03182198486328125,
229
+ 0.03213540771484375,
230
+ 0.03303626220703125,
231
+ 0.03268472900390625,
232
+ 0.03290274658203125
233
+ ],
234
+ "mean": 0.032516226074218745,
235
+ "median": 0.03268472900390625,
236
+ "std": 0.0004637288866362585,
237
+ "min": 0.03182198486328125,
238
+ "max": 0.03303626220703125,
239
+ "p25": 0.03213540771484375,
240
+ "p75": 0.03290274658203125,
241
+ "p90": 0.03298285595703125,
242
+ "p95": 0.03300955908203125,
243
+ "p99": 0.03303092158203125,
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": "2025-10-02T16:46:29.754274",
256
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 1.2930740966796874,
290
+ 1.3008634033203126,
291
+ 1.1206807861328125,
292
+ 1.05006494140625,
293
+ 1.1517652587890626
294
+ ],
295
+ "mean": 1.1832896972656248,
296
+ "median": 1.1517652587890626,
297
+ "std": 0.09852785622191274,
298
+ "min": 1.05006494140625,
299
+ "max": 1.3008634033203126,
300
+ "p25": 1.1206807861328125,
301
+ "p75": 1.2930740966796874,
302
+ "p90": 1.2977476806640624,
303
+ "p95": 1.2993055419921875,
304
+ "p99": 1.3005518310546875,
305
+ "unit": "seconds"
306
+ },
307
+ "time_to_first_token_seconds": {
308
+ "name": "time_to_first_token_seconds",
309
+ "measurements": [
310
+ 0.013690303802490235,
311
+ 0.014367932319641114,
312
+ 0.01406773567199707,
313
+ 0.01174488639831543,
314
+ 0.011559569358825683
315
+ ],
316
+ "mean": 0.013086085510253906,
317
+ "median": 0.013690303802490235,
318
+ "std": 0.0011917142816879657,
319
+ "min": 0.011559569358825683,
320
+ "max": 0.014367932319641114,
321
+ "p25": 0.01174488639831543,
322
+ "p75": 0.01406773567199707,
323
+ "p90": 0.014247853660583497,
324
+ "p95": 0.014307892990112306,
325
+ "p99": 0.014355924453735352,
326
+ "unit": "seconds"
327
+ },
328
+ "tokens_per_second": {
329
+ "name": "tokens_per_second",
330
+ "measurements": [
331
+ 77.33508872908108,
332
+ 76.8720218777474,
333
+ 89.23147540083635,
334
+ 95.23220522540227,
335
+ 86.82324739082469
336
+ ],
337
+ "mean": 85.09880772477835,
338
+ "median": 86.82324739082469,
339
+ "std": 7.080871658281613,
340
+ "min": 76.8720218777474,
341
+ "max": 95.23220522540227,
342
+ "p25": 77.33508872908108,
343
+ "p75": 89.23147540083635,
344
+ "p90": 92.8319132955759,
345
+ "p95": 94.03205926048908,
346
+ "p99": 94.99217603241964,
347
+ "unit": "tokens/sec"
348
+ },
349
+ "time_per_output_token_seconds": {
350
+ "name": "time_per_output_token_seconds",
351
+ "measurements": [
352
+ 0.012930740966796875,
353
+ 0.013008634033203126,
354
+ 0.011206807861328126,
355
+ 0.0105006494140625,
356
+ 0.011517652587890625
357
+ ],
358
+ "mean": 0.011832896972656251,
359
+ "median": 0.011517652587890625,
360
+ "std": 0.0009852785622191272,
361
+ "min": 0.0105006494140625,
362
+ "max": 0.013008634033203126,
363
+ "p25": 0.011206807861328126,
364
+ "p75": 0.012930740966796875,
365
+ "p90": 0.012977476806640626,
366
+ "p95": 0.012993055419921875,
367
+ "p99": 0.013005518310546875,
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": "2025-10-02T16:46:45.703030",
380
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 1.6619581298828126,
414
+ 1.58275244140625,
415
+ 1.669803466796875,
416
+ 1.7710206298828124,
417
+ 1.7816763916015625
418
+ ],
419
+ "mean": 1.6934422119140624,
420
+ "median": 1.669803466796875,
421
+ "std": 0.07430470292895079,
422
+ "min": 1.58275244140625,
423
+ "max": 1.7816763916015625,
424
+ "p25": 1.6619581298828126,
425
+ "p75": 1.7710206298828124,
426
+ "p90": 1.7774140869140624,
427
+ "p95": 1.7795452392578126,
428
+ "p99": 1.7812501611328126,
429
+ "unit": "seconds"
430
+ },
431
+ "time_to_first_token_seconds": {
432
+ "name": "time_to_first_token_seconds",
433
+ "measurements": [
434
+ 0.018489362716674805,
435
+ 0.019263227462768553,
436
+ 0.015909635543823242,
437
+ 0.020379375457763672,
438
+ 0.019595308303833006
439
+ ],
440
+ "mean": 0.018727381896972657,
441
+ "median": 0.019263227462768553,
442
+ "std": 0.0015340050250165485,
443
+ "min": 0.015909635543823242,
444
+ "max": 0.020379375457763672,
445
+ "p25": 0.018489362716674805,
446
+ "p75": 0.019595308303833006,
447
+ "p90": 0.020065748596191407,
448
+ "p95": 0.020222562026977538,
449
+ "p99": 0.020348012771606444,
450
+ "unit": "seconds"
451
+ },
452
+ "tokens_per_second": {
453
+ "name": "tokens_per_second",
454
+ "measurements": [
455
+ 60.16998755982569,
456
+ 63.18107455335947,
457
+ 59.88728732958404,
458
+ 56.46461611608497,
459
+ 56.12691534297608
460
+ ],
461
+ "mean": 59.16597618036606,
462
+ "median": 59.88728732958404,
463
+ "std": 2.614673231558206,
464
+ "min": 56.12691534297608,
465
+ "max": 63.18107455335947,
466
+ "p25": 56.46461611608497,
467
+ "p75": 60.16998755982569,
468
+ "p90": 61.97663975594596,
469
+ "p95": 62.57885715465272,
470
+ "p99": 63.06063107361812,
471
+ "unit": "tokens/sec"
472
+ },
473
+ "time_per_output_token_seconds": {
474
+ "name": "time_per_output_token_seconds",
475
+ "measurements": [
476
+ 0.016619581298828127,
477
+ 0.0158275244140625,
478
+ 0.01669803466796875,
479
+ 0.017710206298828125,
480
+ 0.017816763916015624
481
+ ],
482
+ "mean": 0.016934422119140626,
483
+ "median": 0.01669803466796875,
484
+ "std": 0.0007430470292895078,
485
+ "min": 0.0158275244140625,
486
+ "max": 0.017816763916015624,
487
+ "p25": 0.016619581298828127,
488
+ "p75": 0.017710206298828125,
489
+ "p90": 0.017774140869140625,
490
+ "p95": 0.017795452392578123,
491
+ "p99": 0.017812501611328125,
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": "2025-10-02T16:47:05.620395",
504
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 1.4068468017578124,
538
+ 1.344148681640625,
539
+ 1.3414461669921875,
540
+ 1.336919677734375,
541
+ 1.3870826416015625
542
+ ],
543
+ "mean": 1.3632887939453124,
544
+ "median": 1.344148681640625,
545
+ "std": 0.028292125306798685,
546
+ "min": 1.336919677734375,
547
+ "max": 1.4068468017578124,
548
+ "p25": 1.3414461669921875,
549
+ "p75": 1.3870826416015625,
550
+ "p90": 1.3989411376953125,
551
+ "p95": 1.4028939697265623,
552
+ "p99": 1.4060562353515624,
553
+ "unit": "seconds"
554
+ },
555
+ "time_to_first_token_seconds": {
556
+ "name": "time_to_first_token_seconds",
557
+ "measurements": [
558
+ 0.014988484382629395,
559
+ 0.015139925003051757,
560
+ 0.014728490829467773,
561
+ 0.01472977066040039,
562
+ 0.01467092990875244
563
+ ],
564
+ "mean": 0.014851520156860351,
565
+ "median": 0.01472977066040039,
566
+ "std": 0.00018138828377037096,
567
+ "min": 0.01467092990875244,
568
+ "max": 0.015139925003051757,
569
+ "p25": 0.014728490829467773,
570
+ "p75": 0.014988484382629395,
571
+ "p90": 0.015079348754882812,
572
+ "p95": 0.015109636878967285,
573
+ "p99": 0.015133867378234862,
574
+ "unit": "seconds"
575
+ },
576
+ "tokens_per_second": {
577
+ "name": "tokens_per_second",
578
+ "measurements": [
579
+ 71.08094490107455,
580
+ 74.39653169762678,
581
+ 74.5464130135178,
582
+ 74.7988092818456,
583
+ 72.09375779119934
584
+ ],
585
+ "mean": 73.38329133705281,
586
+ "median": 74.39653169762678,
587
+ "std": 1.5064458165982932,
588
+ "min": 71.08094490107455,
589
+ "max": 74.7988092818456,
590
+ "p25": 72.09375779119934,
591
+ "p75": 74.5464130135178,
592
+ "p90": 74.69785077451449,
593
+ "p95": 74.74833002818005,
594
+ "p99": 74.7887134311125,
595
+ "unit": "tokens/sec"
596
+ },
597
+ "time_per_output_token_seconds": {
598
+ "name": "time_per_output_token_seconds",
599
+ "measurements": [
600
+ 0.014068468017578124,
601
+ 0.013441486816406249,
602
+ 0.013414461669921876,
603
+ 0.01336919677734375,
604
+ 0.013870826416015624
605
+ ],
606
+ "mean": 0.013632887939453126,
607
+ "median": 0.013441486816406249,
608
+ "std": 0.0002829212530679867,
609
+ "min": 0.01336919677734375,
610
+ "max": 0.014068468017578124,
611
+ "p25": 0.013414461669921876,
612
+ "p75": 0.013870826416015624,
613
+ "p90": 0.013989411376953123,
614
+ "p95": 0.014028939697265624,
615
+ "p99": 0.014060562353515623,
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": "2025-10-02T16:47:22.866524",
628
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 1.4011036376953125,
662
+ 1.41193798828125,
663
+ 1.4061102294921874,
664
+ 1.3923253173828125,
665
+ 1.419921142578125
666
+ ],
667
+ "mean": 1.4062796630859375,
668
+ "median": 1.4061102294921874,
669
+ "std": 0.009377103149065705,
670
+ "min": 1.3923253173828125,
671
+ "max": 1.419921142578125,
672
+ "p25": 1.4011036376953125,
673
+ "p75": 1.41193798828125,
674
+ "p90": 1.416727880859375,
675
+ "p95": 1.41832451171875,
676
+ "p99": 1.41960181640625,
677
+ "unit": "seconds"
678
+ },
679
+ "time_to_first_token_seconds": {
680
+ "name": "time_to_first_token_seconds",
681
+ "measurements": [
682
+ 0.015075284957885743,
683
+ 0.015191003799438477,
684
+ 0.017890287399291993,
685
+ 0.015193683624267578,
686
+ 0.015172002792358398
687
+ ],
688
+ "mean": 0.015704452514648436,
689
+ "median": 0.015191003799438477,
690
+ "std": 0.001093777226559946,
691
+ "min": 0.015075284957885743,
692
+ "max": 0.017890287399291993,
693
+ "p25": 0.015172002792358398,
694
+ "p75": 0.015193683624267578,
695
+ "p90": 0.016811645889282228,
696
+ "p95": 0.01735096664428711,
697
+ "p99": 0.017782423248291015,
698
+ "unit": "seconds"
699
+ },
700
+ "tokens_per_second": {
701
+ "name": "tokens_per_second",
702
+ "measurements": [
703
+ 71.372307736272,
704
+ 70.82464019664904,
705
+ 71.11817971491091,
706
+ 71.82229522908655,
707
+ 70.42644623097294
708
+ ],
709
+ "mean": 71.11277382157829,
710
+ "median": 71.11817971491091,
711
+ "std": 0.4743151382177212,
712
+ "min": 70.42644623097294,
713
+ "max": 71.82229522908655,
714
+ "p25": 70.82464019664904,
715
+ "p75": 71.372307736272,
716
+ "p90": 71.64230023196073,
717
+ "p95": 71.73229773052364,
718
+ "p99": 71.80429572937396,
719
+ "unit": "tokens/sec"
720
+ },
721
+ "time_per_output_token_seconds": {
722
+ "name": "time_per_output_token_seconds",
723
+ "measurements": [
724
+ 0.014011036376953126,
725
+ 0.0141193798828125,
726
+ 0.014061102294921874,
727
+ 0.013923253173828126,
728
+ 0.01419921142578125
729
+ ],
730
+ "mean": 0.014062796630859376,
731
+ "median": 0.014061102294921874,
732
+ "std": 9.37710314906568e-05,
733
+ "min": 0.013923253173828126,
734
+ "max": 0.01419921142578125,
735
+ "p25": 0.014011036376953126,
736
+ "p75": 0.0141193798828125,
737
+ "p90": 0.01416727880859375,
738
+ "p95": 0.0141832451171875,
739
+ "p99": 0.014196018164062501,
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": "2025-10-02T16:47:40.330426",
752
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 1.6235733642578125,
786
+ 1.6032005615234375,
787
+ 1.6021641845703125,
788
+ 1.6020162353515626,
789
+ 1.6098470458984375
790
+ ],
791
+ "mean": 1.6081602783203124,
792
+ "median": 1.6032005615234375,
793
+ "std": 0.008230547951931426,
794
+ "min": 1.6020162353515626,
795
+ "max": 1.6235733642578125,
796
+ "p25": 1.6021641845703125,
797
+ "p75": 1.6098470458984375,
798
+ "p90": 1.6180828369140625,
799
+ "p95": 1.6208281005859375,
800
+ "p99": 1.6230243115234375,
801
+ "unit": "seconds"
802
+ },
803
+ "time_to_first_token_seconds": {
804
+ "name": "time_to_first_token_seconds",
805
+ "measurements": [
806
+ 0.016880382537841798,
807
+ 0.01688018226623535,
808
+ 0.01649374771118164,
809
+ 0.016784343719482422,
810
+ 0.016762063980102538
811
+ ],
812
+ "mean": 0.01676014404296875,
813
+ "median": 0.016784343719482422,
814
+ "std": 0.00014171999632086487,
815
+ "min": 0.01649374771118164,
816
+ "max": 0.016880382537841798,
817
+ "p25": 0.016762063980102538,
818
+ "p75": 0.01688018226623535,
819
+ "p90": 0.016880302429199218,
820
+ "p95": 0.016880342483520508,
821
+ "p99": 0.01688037452697754,
822
+ "unit": "seconds"
823
+ },
824
+ "tokens_per_second": {
825
+ "name": "tokens_per_second",
826
+ "measurements": [
827
+ 61.59253545386488,
828
+ 62.375227653971905,
829
+ 62.41557573378111,
830
+ 62.421339929838474,
831
+ 62.11770258222956
832
+ ],
833
+ "mean": 62.184476270737186,
834
+ "median": 62.375227653971905,
835
+ "std": 0.3164648799422655,
836
+ "min": 61.59253545386488,
837
+ "max": 62.421339929838474,
838
+ "p25": 62.11770258222956,
839
+ "p75": 62.41557573378111,
840
+ "p90": 62.41903425141553,
841
+ "p95": 62.420187090627,
842
+ "p99": 62.42110936199618,
843
+ "unit": "tokens/sec"
844
+ },
845
+ "time_per_output_token_seconds": {
846
+ "name": "time_per_output_token_seconds",
847
+ "measurements": [
848
+ 0.016235733642578125,
849
+ 0.016032005615234376,
850
+ 0.016021641845703125,
851
+ 0.016020162353515624,
852
+ 0.016098470458984375
853
+ ],
854
+ "mean": 0.016081602783203124,
855
+ "median": 0.016032005615234376,
856
+ "std": 8.230547951931426e-05,
857
+ "min": 0.016020162353515624,
858
+ "max": 0.016235733642578125,
859
+ "p25": 0.016021641845703125,
860
+ "p75": 0.016098470458984375,
861
+ "p90": 0.016180828369140626,
862
+ "p95": 0.016208281005859373,
863
+ "p99": 0.016230243115234376,
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": "2025-10-02T16:47:59.552741",
876
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 1.926750244140625,
910
+ 1.9081624755859374,
911
+ 1.9543255615234374,
912
+ 1.921574951171875,
913
+ 1.941179931640625
914
+ ],
915
+ "mean": 1.9303986328125,
916
+ "median": 1.926750244140625,
917
+ "std": 0.015964642857014946,
918
+ "min": 1.9081624755859374,
919
+ "max": 1.9543255615234374,
920
+ "p25": 1.921574951171875,
921
+ "p75": 1.941179931640625,
922
+ "p90": 1.9490673095703124,
923
+ "p95": 1.9516964355468749,
924
+ "p99": 1.953799736328125,
925
+ "unit": "seconds"
926
+ },
927
+ "time_to_first_token_seconds": {
928
+ "name": "time_to_first_token_seconds",
929
+ "measurements": [
930
+ 0.01997422218322754,
931
+ 0.02052933692932129,
932
+ 0.020223737716674805,
933
+ 0.020222537994384764,
934
+ 0.020372894287109374
935
+ ],
936
+ "mean": 0.020264545822143553,
937
+ "median": 0.020223737716674805,
938
+ "std": 0.00018415835288102274,
939
+ "min": 0.01997422218322754,
940
+ "max": 0.02052933692932129,
941
+ "p25": 0.020222537994384764,
942
+ "p75": 0.020372894287109374,
943
+ "p90": 0.020466759872436523,
944
+ "p95": 0.020498048400878904,
945
+ "p99": 0.020523079223632813,
946
+ "unit": "seconds"
947
+ },
948
+ "tokens_per_second": {
949
+ "name": "tokens_per_second",
950
+ "measurements": [
951
+ 51.900862763145675,
952
+ 52.40643880144069,
953
+ 51.1685473335609,
954
+ 52.04064506514038,
955
+ 51.515059665531936
956
+ ],
957
+ "mean": 51.80631072576391,
958
+ "median": 51.900862763145675,
959
+ "std": 0.4279195490089476,
960
+ "min": 51.1685473335609,
961
+ "max": 52.40643880144069,
962
+ "p25": 51.515059665531936,
963
+ "p75": 52.04064506514038,
964
+ "p90": 52.26012130692057,
965
+ "p95": 52.333280054180634,
966
+ "p99": 52.39180705198868,
967
+ "unit": "tokens/sec"
968
+ },
969
+ "time_per_output_token_seconds": {
970
+ "name": "time_per_output_token_seconds",
971
+ "measurements": [
972
+ 0.01926750244140625,
973
+ 0.019081624755859374,
974
+ 0.019543255615234373,
975
+ 0.019215749511718748,
976
+ 0.019411799316406252
977
+ ],
978
+ "mean": 0.019303986328125,
979
+ "median": 0.01926750244140625,
980
+ "std": 0.0001596464285701497,
981
+ "min": 0.019081624755859374,
982
+ "max": 0.019543255615234373,
983
+ "p25": 0.019215749511718748,
984
+ "p75": 0.019411799316406252,
985
+ "p90": 0.019490673095703126,
986
+ "p95": 0.01951696435546875,
987
+ "p99": 0.019537997363281247,
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": "2025-10-02T16:48:27.028688",
1000
+ "commit_id": "41e5abac5cb49983a08ddef3e8645d6efd23c8f3",
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
+ 1.64763818359375,
1034
+ 1.6286192626953124,
1035
+ 1.6526793212890625,
1036
+ 1.472946533203125,
1037
+ 1.4077335205078125
1038
+ ],
1039
+ "mean": 1.5619233642578125,
1040
+ "median": 1.6286192626953124,
1041
+ "std": 0.10170880123078788,
1042
+ "min": 1.4077335205078125,
1043
+ "max": 1.6526793212890625,
1044
+ "p25": 1.472946533203125,
1045
+ "p75": 1.64763818359375,
1046
+ "p90": 1.6506628662109375,
1047
+ "p95": 1.6516710937499999,
1048
+ "p99": 1.65247767578125,
1049
+ "unit": "seconds"
1050
+ },
1051
+ "time_to_first_token_seconds": {
1052
+ "name": "time_to_first_token_seconds",
1053
+ "measurements": [
1054
+ 0.016830020904541016,
1055
+ 0.01698569869995117,
1056
+ 0.01697330093383789,
1057
+ 0.017517572402954103,
1058
+ 0.013857258796691894
1059
+ ],
1060
+ "mean": 0.016432770347595216,
1061
+ "median": 0.01697330093383789,
1062
+ "std": 0.0013088748091887145,
1063
+ "min": 0.013857258796691894,
1064
+ "max": 0.017517572402954103,
1065
+ "p25": 0.016830020904541016,
1066
+ "p75": 0.01698569869995117,
1067
+ "p90": 0.01730482292175293,
1068
+ "p95": 0.017411197662353516,
1069
+ "p99": 0.017496297454833984,
1070
+ "unit": "seconds"
1071
+ },
1072
+ "tokens_per_second": {
1073
+ "name": "tokens_per_second",
1074
+ "measurements": [
1075
+ 60.69293671131411,
1076
+ 61.40170529145236,
1077
+ 60.50780614959329,
1078
+ 67.8911268982291,
1079
+ 71.03617164982116
1080
+ ],
1081
+ "mean": 64.30594934008201,
1082
+ "median": 61.40170529145236,
1083
+ "std": 4.337367309948397,
1084
+ "min": 60.50780614959329,
1085
+ "max": 71.03617164982116,
1086
+ "p25": 60.69293671131411,
1087
+ "p75": 67.8911268982291,
1088
+ "p90": 69.77815374918434,
1089
+ "p95": 70.40716269950275,
1090
+ "p99": 70.91036985975748,
1091
+ "unit": "tokens/sec"
1092
+ },
1093
+ "time_per_output_token_seconds": {
1094
+ "name": "time_per_output_token_seconds",
1095
+ "measurements": [
1096
+ 0.016476381835937498,
1097
+ 0.016286192626953123,
1098
+ 0.016526793212890623,
1099
+ 0.01472946533203125,
1100
+ 0.014077335205078124
1101
+ ],
1102
+ "mean": 0.015619233642578122,
1103
+ "median": 0.016286192626953123,
1104
+ "std": 0.0010170880123078782,
1105
+ "min": 0.014077335205078124,
1106
+ "max": 0.016526793212890623,
1107
+ "p25": 0.01472946533203125,
1108
+ "p75": 0.016476381835937498,
1109
+ "p90": 0.016506628662109373,
1110
+ "p95": 0.016516710937499998,
1111
+ "p99": 0.016524776757812498,
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
+ }