{ "format": "tiny_receipt_vqa_runtime_benchmark_v1", "precision": "fp32", "providers": [ "CPUExecutionProvider" ], "threads": 1, "timing_scope": "model inference from encoder input through EOS; session creation and image/token preprocessing excluded", "warmups_per_vqa": 1, "measured_repeats_per_vqa": 3, "vqa_count": 20, "aggregate": { "e2e_median_of_vqa_medians_ms": 288.50183452595957 }, "results": [ { "id": "00000", "image": "00000.jpg", "question": "The location of the store is Hyocheon St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 37, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 299, 342, 339, 330, 54, 306, 1034, 1023, 326, 54, 300, 336, 1345, 1111, 326, 1349, 54, 1128, 54, 14, 12, 21, 7, 8, 1295, 9, 10, 14, 12, 21, 11, 2 ], "decoder_calls": 36, "selected_family_id": 1, "encoder_ms": { "median": 228.1085120048374, "mean": 228.73557733449465, "min": 224.96837296057492, "max": 233.1298470380716 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 62.75751499924809, "mean": 63.27293966508781, "min": 60.33691199263558, "max": 66.72439200337976 }, "e2e_ms": { "median": 290.924838045612, "mean": 292.0872653485276, "min": 285.4173560044728, "max": 299.91960199549794 } } }, { "id": "00001", "image": "00001.jpg", "question": "What is the first number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 14, 17, 20, 18, 14, 15, 20, 7, 8, 1063, 13, 9, 10, 14, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 218.68814597837627, "mean": 218.8280163294015, "min": 216.8369370047003, "max": 220.95896600512788 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 34.41185498377308, "mean": 34.29271699860692, "min": 32.77926699956879, "max": 35.68702901247889 }, "e2e_ms": { "median": 252.58133298484609, "mean": 253.1810699729249, "min": 251.52513198554516, "max": 255.43674494838342 } } }, { "id": "00002", "image": "00002.jpg", "question": "What is the fourth number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 14, 20, 17, 13, 12, 19, 15, 7, 8, 1063, 16, 9, 10, 13, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 219.080465962179, "mean": 220.12510864684978, "min": 218.29582698410377, "max": 222.9990329942666 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 36.60359600326046, "mean": 36.09063268716758, "min": 34.20854202704504, "max": 37.45976003119722 }, "e2e_ms": { "median": 255.7410699664615, "mean": 256.27760131222504, "min": 252.57335900096223, "max": 260.51837496925145 } } }, { "id": "00003", "image": "00003.jpg", "question": "The location of the store is Seonggokbuk St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 37, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 305, 1369, 337, 336, 54, 311, 1216, 1452, 54, 1303, 1089, 54, 336, 340, 322, 1131, 54, 1128, 54, 13, 12, 13, 7, 8, 1295, 9, 10, 13, 12, 13, 11, 2 ], "decoder_calls": 36, "selected_family_id": 1, "encoder_ms": { "median": 234.17181300465018, "mean": 232.15498200928172, "min": 225.09921196615323, "max": 237.19392105704173 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 62.14503099909052, "mean": 60.96253534390902, "min": 58.41115303337574, "max": 62.33142199926078 }, "e2e_ms": { "median": 292.640573985409, "mean": 293.1756866552557, "min": 287.4870509840548, "max": 299.39943499630317 } } }, { "id": "00004", "image": "00004.jpg", "question": "What is the fourth number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 17, 16, 13, 17, 19, 18, 20, 7, 8, 1063, 16, 9, 10, 17, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 219.73750600591302, "mean": 219.55031366087496, "min": 218.67956197820604, "max": 220.23387299850583 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 33.03238400258124, "mean": 32.85907100265225, "min": 32.23047300707549, "max": 33.314355998300016 }, "e2e_ms": { "median": 252.05038499552757, "mean": 252.4662456708029, "min": 252.02497700229287, "max": 253.32337501458824 } } }, { "id": "00005", "image": "00005.jpg", "question": "What is the third number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 17, 14, 15, 17, 14, 19, 19, 7, 8, 1063, 15, 9, 10, 15, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 221.06853598961607, "mean": 222.28092465472096, "min": 220.07366101024672, "max": 225.70057696430013 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 34.189306024927646, "mean": 34.80713399282346, "min": 32.518005988094956, "max": 37.71408996544778 }, "e2e_ms": { "median": 258.8410570169799, "mean": 257.14559699796763, "min": 252.6491439784877, "max": 259.94658999843523 } } }, { "id": "00006", "image": "00006.jpg", "question": "What is the third number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 17, 16, 19, 19, 20, 17, 14, 7, 8, 1063, 15, 9, 10, 19, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 222.55956701701507, "mean": 222.51495001061508, "min": 220.9344040020369, "max": 224.0508790127933 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 34.30454304907471, "mean": 35.38931436681499, "min": 33.86030503315851, "max": 38.00309501821175 }, "e2e_ms": { "median": 257.96999502927065, "mean": 257.9631620125535, "min": 255.29778801137581, "max": 260.6217029970139 } } }, { "id": "00007", "image": "00007.jpg", "question": "The location of the store is Hwadoan St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 37, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 311, 1035, 1026, 1020, 54, 299, 336, 1463, 54, 300, 344, 1375, 1072, 54, 335, 328, 54, 1128, 54, 19, 19, 17, 7, 8, 1295, 9, 10, 19, 19, 17, 11, 2 ], "decoder_calls": 36, "selected_family_id": 1, "encoder_ms": { "median": 229.6274010441266, "mean": 228.94571267534047, "min": 226.80862696142867, "max": 230.40111002046615 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 62.39598500542343, "mean": 62.76772233347098, "min": 61.7439360357821, "max": 64.16324595920742 }, "e2e_ms": { "median": 292.20226302277297, "mean": 291.7710263476086, "min": 289.26231001969427, "max": 293.8485060003586 } } }, { "id": "00008", "image": "00008.jpg", "question": "The location of the store is Heowanghu St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 38, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 299, 344, 1168, 334, 1380, 54, 317, 326, 1020, 1355, 54, 300, 326, 1113, 322, 1021, 330, 54, 1128, 54, 16, 20, 16, 7, 8, 1295, 9, 10, 16, 20, 16, 11, 2 ], "decoder_calls": 37, "selected_family_id": 1, "encoder_ms": { "median": 230.19460099749267, "mean": 230.10672200083113, "min": 229.20020599849522, "max": 230.92535900650546 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 60.01490697963163, "mean": 60.61753230945518, "min": 59.19383798027411, "max": 62.643851968459785 }, "e2e_ms": { "median": 291.00127099081874, "mean": 290.7840140008678, "min": 289.4475909997709, "max": 291.90318001201376 } } }, { "id": "00009", "image": "00009.jpg", "question": "What is the third number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 12, 12, 18, 16, 16, 17, 17, 7, 8, 1063, 15, 9, 10, 18, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 224.44382100366056, "mean": 225.2649093279615, "min": 224.18938198825344, "max": 227.1615249919705 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 33.209228015039116, "mean": 33.426869000929095, "min": 32.81561698531732, "max": 34.255762002430856 }, "e2e_ms": { "median": 257.70985695999116, "mean": 258.76177366202074, "min": 257.06289801746607, "max": 261.51256600860506 } } }, { "id": "00010", "image": "00010.jpg", "question": "What is the third number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 17, 17, 19, 19, 14, 18, 13, 7, 8, 1063, 15, 9, 10, 19, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 224.5594309642911, "mean": 224.42714798186594, "min": 222.7793749771081, "max": 225.94263800419867 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 34.14588398300111, "mean": 34.801219667618476, "min": 32.99805102869868, "max": 37.259723991155624 }, "e2e_ms": { "median": 258.76506802160293, "mean": 259.28606300537166, "min": 255.8331110049039, "max": 263.2600099896081 } } }, { "id": "00011", "image": "00011.jpg", "question": "The location of the store is Seongsanjungang St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 36, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 312, 1348, 323, 1380, 54, 300, 1313, 1355, 54, 311, 1216, 1452, 54, 311, 54, 341, 288, 54, 17, 15, 13, 7, 8, 1295, 9, 10, 17, 15, 13, 11, 2 ], "decoder_calls": 35, "selected_family_id": 1, "encoder_ms": { "median": 232.3369620135054, "mean": 233.1079383341906, "min": 232.2882090229541, "max": 234.6986439661123 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 59.2429920216091, "mean": 58.67435699716831, "min": 56.36540800333023, "max": 60.41467096656561 }, "e2e_ms": { "median": 292.7600480034016, "mean": 291.8443593468207, "min": 288.7621429981664, "max": 294.010887038894 } } }, { "id": "00012", "image": "00012.jpg", "question": "What is the second number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 14, 20, 12, 15, 15, 14, 17, 7, 8, 1063, 14, 9, 10, 20, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 226.1477159918286, "mean": 229.65850731513152, "min": 224.2999929585494, "max": 238.52781299501657 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 33.429053029976785, "mean": 34.00779399089515, "min": 32.332515984307975, "max": 36.2618129584007 }, "e2e_ms": { "median": 262.48074299655855, "mean": 263.72832199558616, "min": 257.78591400012374, "max": 270.9183089900762 } } }, { "id": "00013", "image": "00013.jpg", "question": "The location of the store is Hagusan St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 34, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 295, 1026, 336, 1065, 54, 1471, 1484, 54, 300, 1095, 323, 342, 332, 54, 1128, 54, 16, 16, 21, 7, 8, 1295, 9, 10, 16, 16, 21, 11, 2 ], "decoder_calls": 33, "selected_family_id": 1, "encoder_ms": { "median": 236.18908500066027, "mean": 235.46872369479388, "min": 230.60361604439095, "max": 239.61347003933042 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 56.88240297604352, "mean": 55.94197332781429, "min": 52.23298800410703, "max": 58.71052900329232 }, "e2e_ms": { "median": 291.9109900249168, "mean": 291.4869276961933, "min": 289.3731670337729, "max": 293.17662602989003 } } }, { "id": "00014", "image": "00014.jpg", "question": "The location of the store is Yongjeonsaeteo St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 35, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 302, 326, 1355, 54, 301, 335, 331, 326, 54, 317, 1125, 1283, 323, 336, 54, 1128, 54, 18, 16, 12, 7, 8, 1295, 9, 10, 18, 16, 12, 11, 2 ], "decoder_calls": 34, "selected_family_id": 1, "encoder_ms": { "median": 232.1202660095878, "mean": 233.38799566651383, "min": 231.80684499675408, "max": 236.23687599319965 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 55.49834400881082, "mean": 56.75215433196475, "min": 54.934261017479, "max": 59.82385796960443 }, "e2e_ms": { "median": 291.79100500186905, "mean": 290.1974580211875, "min": 286.79970605298877, "max": 292.0016630087048 } } }, { "id": "00015", "image": "00015.jpg", "question": "The location of the store is Anpo St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 38, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 299, 1243, 1023, 326, 54, 295, 1026, 1125, 1484, 54, 293, 335, 342, 1249, 1139, 326, 1317, 54, 1128, 54, 16, 19, 15, 7, 8, 1295, 9, 10, 16, 19, 15, 11, 2 ], "decoder_calls": 37, "selected_family_id": 1, "encoder_ms": { "median": 229.83304603258148, "mean": 230.01183301676065, "min": 229.7786429990083, "max": 230.42381001869217 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 63.79281001864001, "mean": 64.01001233219479, "min": 61.98739598039538, "max": 66.249830997549 }, "e2e_ms": { "median": 293.68569899816066, "mean": 294.08051931144047, "min": 292.4686739570461, "max": 296.0871849791147 } } }, { "id": "00016", "image": "00016.jpg", "question": "The location of the store is Jaeunhyang St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 36, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 299, 344, 1035, 1026, 1020, 54, 317, 326, 1355, 54, 302, 326, 1317, 337, 1380, 54, 1128, 54, 19, 15, 20, 7, 8, 1295, 9, 10, 19, 15, 20, 11, 2 ], "decoder_calls": 35, "selected_family_id": 1, "encoder_ms": { "median": 232.301420008298, "mean": 233.35739333803454, "min": 230.61087302630767, "max": 237.15988697949797 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 60.91165903490037, "mean": 61.055687353170164, "min": 57.506710989400744, "max": 64.74869203520939 }, "e2e_ms": { "median": 294.7251980076544, "mean": 294.4723453450327, "min": 293.27324201585725, "max": 295.4185960115865 } } }, { "id": "00017", "image": "00017.jpg", "question": "The location of the store is Namchangjangteo St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 33, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 306, 1034, 1484, 1291, 54, 302, 1065, 1385, 54, 293, 331, 1317, 54, 1128, 54, 14, 16, 21, 7, 8, 1295, 9, 10, 14, 16, 21, 11, 2 ], "decoder_calls": 32, "selected_family_id": 1, "encoder_ms": { "median": 233.2940620253794, "mean": 232.74352065830803, "min": 231.19594698073342, "max": 233.74055296881124 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 55.82460999721661, "mean": 56.32064832995335, "min": 55.750339990481734, "max": 57.38699500216171 }, "e2e_ms": { "median": 289.1032020561397, "mean": 289.12289599732804, "min": 288.6404799646698, "max": 289.6250059711747 } } }, { "id": "00018", "image": "00018.jpg", "question": "The location of the store is Nulha St. [?]. (Fill the blank)", "requested_family": "auto", "selected_family": "address", "token_count": 33, "measurement": { "tokens": [ 1, 4, 1064, 5, 6, 296, 1125, 1095, 54, 311, 342, 1380, 54, 315, 1020, 331, 1065, 54, 1128, 54, 16, 18, 15, 7, 8, 1295, 9, 10, 16, 18, 15, 11, 2 ], "decoder_calls": 32, "selected_family_id": 1, "encoder_ms": { "median": 231.79093899670988, "mean": 232.75611198429638, "min": 230.4946249932982, "max": 235.9827719628811 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 52.24784801248461, "mean": 52.896186690001436, "min": 50.39089702768251, "max": 56.04981502983719 }, "e2e_ms": { "median": 287.9004669957794, "mean": 285.7119679877845, "min": 280.9447349864058, "max": 288.29070198116824 } } }, { "id": "00019", "image": "00019.jpg", "question": "What is the second number of the store's phone number?", "requested_family": "auto", "selected_family": "phone", "token_count": 21, "measurement": { "tokens": [ 1, 4, 1038, 5, 6, 12, 12, 16, 12, 20, 20, 14, 7, 8, 1063, 14, 9, 10, 12, 11, 2 ], "decoder_calls": 20, "selected_family_id": 0, "encoder_ms": { "median": 227.41803695680574, "mean": 228.0839123025847, "min": 225.56102694943547, "max": 231.27267300151289 }, "cross_kv_ms": { "median": 0.0, "mean": 0.0, "min": 0.0, "max": 0.0 }, "decoder_ms": { "median": 34.86111399251968, "mean": 34.86383269773796, "min": 32.26642304798588, "max": 37.46396105270833 }, "e2e_ms": { "median": 260.49909502035007, "mean": 263.01347836852074, "min": 259.74823004798964, "max": 268.7931100372225 } } } ], "artifacts": { "encoder": { "filename": "encoder_model.onnx", "bytes": 61811550, "sha256": "0e2206c54756b15d697d2bbd8d22bb68d0d2475d0323965ca03a14fb186e6eb0" }, "decoder": { "filename": "decoder_model.onnx", "bytes": 28616273, "sha256": "b918a9de531bad205990acff28a2a7e2244b14229c312ed6b2f75d0540d6a6a2" } } }