File tree Expand file tree Collapse file tree 3 files changed +48
-2
lines changed
vllm-benchmarks/benchmarks/cuda Expand file tree Collapse file tree 3 files changed +48
-2
lines changed Original file line number Diff line number Diff line change @@ -2,8 +2,8 @@ name: vLLM Benchmark
2
2
3
3
on :
4
4
schedule :
5
- # Run every 6 hours
6
- - cron : ' 0 */6 * * *'
5
+ # Run every 12 hours
6
+ - cron : ' 0 */12 * * *'
7
7
workflow_dispatch :
8
8
inputs :
9
9
vllm_branch :
Original file line number Diff line number Diff line change 105
105
"num_iters" : 15 ,
106
106
"max_model_len" : 8192
107
107
}
108
+ },
109
+ {
110
+ "test_name" : " latency_gemma_3_27b_it_tp8" ,
111
+ "parameters" : {
112
+ "model" : " google/gemma-3-27b-it" ,
113
+ "tensor_parallel_size" : 8 ,
114
+ "load_format" : " dummy" ,
115
+ "num_iters_warmup" : 5 ,
116
+ "num_iters" : 15 ,
117
+ "max_model_len" : 8192
118
+ }
119
+ },
120
+ {
121
+ "test_name" : " latency_qwen3_30b_a3b_tp8" ,
122
+ "parameters" : {
123
+ "model" : " Qwen/Qwen3-30B-A3B" ,
124
+ "tensor_parallel_size" : 8 ,
125
+ "load_format" : " dummy" ,
126
+ "num_iters_warmup" : 5 ,
127
+ "num_iters" : 15 ,
128
+ "max_model_len" : 8192
129
+ }
108
130
}
109
131
]
Original file line number Diff line number Diff line change 115
115
"backend" : " vllm" ,
116
116
"max_model_len" : 8192
117
117
}
118
+ },
119
+ {
120
+ "test_name" : " throughput_gemma_3_27b_it_tp8" ,
121
+ "parameters" : {
122
+ "model" : " google/gemma-3-27b-it" ,
123
+ "tensor_parallel_size" : 8 ,
124
+ "load_format" : " dummy" ,
125
+ "dataset" : " ./ShareGPT_V3_unfiltered_cleaned_split.json" ,
126
+ "num_prompts" : 200 ,
127
+ "backend" : " vllm" ,
128
+ "max_model_len" : 8192
129
+ }
130
+ },
131
+ {
132
+ "test_name" : " throughput_qwen3_30b_a3b_tp8" ,
133
+ "parameters" : {
134
+ "model" : " Qwen/Qwen3-30B-A3B" ,
135
+ "tensor_parallel_size" : 8 ,
136
+ "load_format" : " dummy" ,
137
+ "dataset" : " ./ShareGPT_V3_unfiltered_cleaned_split.json" ,
138
+ "num_prompts" : 200 ,
139
+ "backend" : " vllm" ,
140
+ "max_model_len" : 8192
141
+ }
118
142
}
119
143
]
You can’t perform that action at this time.
0 commit comments