Datasets:

auto-cap
/

moe-cap-results

Modalities:

Formats:

Size:

Libraries:

License:

Dataset card Data Studio Files Files and versions

Dataset Viewer

Auto-converted to Parquet Duplicate

Split (1)

train · 9 rows

prefill_smbu float64	prefill_smfu float64	decoding_smbu float64	decoding_smfu float64	kv_size float64	decoding_throughput float64	prefill_tp float64	ttft float64	tpot float64	exact_match float64	correct int64	total int64	cost float64	model_name string	method string	precision string	e2e_s float64	batch_size int64	gpu_type string	dataset string	model_type string
0.163255	0.002897	0.288016	0.027116	2,828.860067	1,402.86957	1,478.447055	0.067242	0.048451	0.811221	1,070	1,319	15,342.2	Qwen/Qwen3-30B-A3B	sglang	bfloat16	650.12	1	5xRTX A5000	gsm8k	thinking
0.616187	0.040905	0.671695	0.027068	967.951457	599.353998	2,176.333107	0.057978	0.02983	0.457165	603	1,319	7,158.92	Qwen/Qwen1.5-MoE-A2.7B-Chat	sglang	bfloat16	295.91	null	1xRTX A6000	gsm8k	instruct
0.140926	0.001437	0.320366	0.005892	4,033.749704	1,136.774328	1,102.054927	0.092873	0.024498	0.6884	908	1,319	104,052.07	Qwen/Qwen3-235B-A22B-Thinking-2507-FP8	sglang	fp8	728.81	null	2xH200	gsm8k	thinking
0.178495	0.001752	0.338158	0.006024	3,627.284703	1,206.321146	1,344.079539	0.081478	0.022422	0.702805	927	1,319	195,252.11	Qwen/Qwen3-235B-A22B-Thinking-2507	sglang	bfloat16	654.28	null	4xH200	gsm8k	thinking
0.050089	0.000375	0.383759	0.006176	9,510.041884	1,694.303376	575.367769	0.213825	0.026344	0.711903	939	1,319	344,657.14	Qwen/Qwen3-235B-A22B	sglang	bfloat16	1,099.04	null	8xH100	gsm8k	thinking
0.111854	0.001812	0.260047	0.003109	441.661049	638.027283	1,278.344026	0.102821	0.016742	0.53298	703	1,319	21,600.27	Qwen/Qwen3-30B-A3B-Instruct-2507	sglang	bfloat16	333.04	null	4xRTX A6000	gsm8k	instruct
0.08958	0.001176	0.348493	0.020525	2,350.775083	1,701.41349	830.884236	0.12335	0.038886	0.692949	914	1,319	21,600.54	Qwen/Qwen3-30B-A3B-Thinking-2507	sglang	bfloat16	677.63	null	4xRTX A6000	gsm8k	thinking
0.090023	0.001135	0.37876	0.020892	2,865.311679	1,417.494248	803.391382	0.125442	0.033938	0.806672	1,064	1,319	21,600.69	Qwen/Qwen3-30B-A3B	sglang	bfloat16	864.98	null	4xRTX A6000	gsm8k	thinking
0.069852	0.001117	0.314062	0.024781	1,794.97231	1,806.091576	800.771115	0.125366	0.014821	0.809704	1,068	1,319	54,380.91	Qwen/Qwen3-30B-A3B	sglang	bfloat16	575.59	null	2xA100	gsm8k	thinking

README.md exists but content is empty.

Downloads last month: 116

Size of downloaded dataset files:

Size of the auto-converted Parquet files:

Number of rows: