Skip to content
End-to-End Tests

e2e Test - kimi aiperf w/ live assistant #2594

Sign in to view logs

e2e Test - kimi aiperf w/ live assistant

e2e Test - kimi aiperf w/ live assistant #2594

Run time

Learn about OS pricing on GitHub Actions
Job Run time
get-jobs
10s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 4, 1... / kimik2.5_tp4_conc8_offloadnone fp4 b200-dgxc vllm | tp=4 ep=1 dpa=false | disagg-false spec-none conc-8
47m 38s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 8, 1... / kimik2.5_tp8_conc2_offloadnone fp4 b200-dgxc vllm | tp=8 ep=1 dpa=false | disagg-false spec-none conc-2
49m 7s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 4, 1... / kimik2.5_tp4_conc12_offloadnone fp4 b200-dgxc vllm | tp=4 ep=1 dpa=false | disagg-false spec-none conc-12
45m 56s
multi-node agentic /
0s
eval /
0s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 8, 1... / kimik2.5_tp8_conc16_offloadnone fp4 b200-dgxc vllm | tp=8 ep=1 dpa=false | disagg-false spec-none conc-16
46m 35s
multi-node eval /
0s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 4, 1... / kimik2.5_tp4_conc14_offloadnone fp4 b200-dgxc vllm | tp=4 ep=1 dpa=false | disagg-false spec-none conc-14
48m 6s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 4, 1... / kimik2.5_tp4_conc20_offloadnone fp4 b200-dgxc vllm | tp=4 ep=1 dpa=false | disagg-false spec-none conc-20
45m 38s
multi-node /
0s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 8, 1... / kimik2.5_tp8_conc1_offloadnone fp4 b200-dgxc vllm | tp=8 ep=1 dpa=false | disagg-false spec-none conc-1
46m 6s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 8, 1... / kimik2.5_tp8_conc8_offloadnone fp4 b200-dgxc vllm | tp=8 ep=1 dpa=false | disagg-false spec-none conc-8
46m 57s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 8, 1... / kimik2.5_tp8_conc24_offloadnone fp4 b200-dgxc vllm | tp=8 ep=1 dpa=false | disagg-false spec-none conc-24
49m 21s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 4, 1... / kimik2.5_tp4_conc16_offloadnone fp4 b200-dgxc vllm | tp=4 ep=1 dpa=false | disagg-false spec-none conc-16
46m 19s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 4, 1... / kimik2.5_tp4_conc18_offloadnone fp4 b200-dgxc vllm | tp=4 ep=1 dpa=false | disagg-false spec-none conc-18
44m 12s
agentic / (vllm/vllm-openai:v0.20.2, nvidia/Kimi-K2.5-NVFP4, kimik2.5, fp4, vllm, b200-dgxc, 8, 1... / kimik2.5_tp8_conc4_offloadnone fp4 b200-dgxc vllm | tp=8 ep=1 dpa=false | disagg-false spec-none conc-4
41m 1s
single-node /
0s
collect-evals
0s
collect-agentic-results
20s
collect-results / collect-results
16s
calc-success-rate
20s
9h 18m 2s