Skip to content

Commit a738dbb

Browse files
authored
Update test case parameter to have the throughput above 8.0 (#19994)
Signed-off-by: Qiliang Cui <derrhein@gmail.com>
1 parent 33d5e29 commit a738dbb

File tree

1 file changed

+2
-2
lines changed

1 file changed

+2
-2
lines changed

.buildkite/scripts/tpu/config_v6e_1.env

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -4,8 +4,8 @@ CONTAINER_NAME=vllm-tpu
44

55
# vllm config
66
MODEL=meta-llama/Llama-3.1-8B-Instruct
7-
MAX_NUM_SEQS=512
8-
MAX_NUM_BATCHED_TOKENS=512
7+
MAX_NUM_SEQS=256
8+
MAX_NUM_BATCHED_TOKENS=1024
99
TENSOR_PARALLEL_SIZE=1
1010
MAX_MODEL_LEN=2048
1111
DOWNLOAD_DIR=/mnt/disks/persist

0 commit comments

Comments
 (0)