A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 15:45 UTC
[ci] Soft fail Entrypoints, Samplers, LoRA, Decoder-only VLM (#8872)
Signed-off-by: kevin <kevin@anyscale.com>
Kevin H. Luu committed
Sep 26, 2024 at 15:02 UTC
d9cfbc891e2e1d62d74c7aae93bde436a29bd574
1 file changed
+4
.buildkite/test-pipeline.yaml
+4
@@ -83,6 +83,7 @@ steps:
83
84
- label: Entrypoints Test # 20min
85
working_dir: "/vllm-workspace/tests"
86
+ soft_fail: true
87
fast_check: true
88
mirror_hardwares: [amd]
89
source_file_dependencies:
@@ -177,6 +178,7 @@ steps:
178
- pytest -v -s prefix_caching
179
180
- label: Samplers Test # 18min
181
+ soft_fail: true
182
source_file_dependencies:
183
- vllm/model_executor/layers
184
- vllm/sampling_metadata.py
@@ -204,6 +206,7 @@ steps:
206
207
- label: LoRA Test %N # 30min each
208
mirror_hardwares: [amd]
209
+ soft_fail: true
210
source_file_dependencies:
211
- vllm/lora
212
- tests/lora
@@ -308,6 +311,7 @@ steps:
311
- pytest -v -s models/decoder_only/language
312
313
- label: Decoder-only Multi-Modal Models Test # 56min
314
+ soft_fail: true
315
#mirror_hardwares: [amd]
316
source_file_dependencies:
317
- vllm/