| .. |
|
disagg_config_cache_aware_balance.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_cache_reuse.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp1_gentp1_deepseek_v3_lite_one_mtp_attention_dp_overlap.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp1_gentp1_deepseek_v3_lite_one_mtp.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp1_gentp1_deepseek_v3_lite.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp2_gentp1.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp2_gentp2_deepseek_v3_lite_attention_dp_one_mtp.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp2_gentp2_deepseek_v3_lite_attention_dp_one.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp2_gentp2_deepseek_v3_lite_attention_dp_overlap_cuda_graph.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp2_gentp2_deepseek_v3_lite_attention_dp_overlap.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp2_gentp2_deepseek_v3_lite_attention_dp.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp2_gentp2_deepseek_v3_lite_overlap_cuda_graph.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_ctxtp2_gentp2_deepseek_v3_lite.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_cuda_graph_padding.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_gen_only.yaml
|
feat: Add option to run disaggregated serving without ctx servers,… (#3243)
|
2025-04-07 21:56:03 -04:00 |
|
disagg_config_load_balance.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_mixed.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
|
disagg_config_overlap.yaml
|
Breaking change: perf: Enable scheduling overlap by default (#4174)
|
2025-05-15 14:27:36 +08:00 |
