e26fe1caf1
### What this PR does / why we need it?
1. update expected accuracy for DeepSeek-V2-Lite
2. add batch size
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Accuracy CI passed
- vLLM version: v0.10.2
- vLLM main:
838d7116ba
Signed-off-by: hfadzxy <starmoon_zhang@163.com>
16 lines
355 B
YAML
16 lines
355 B
YAML
model_name: "deepseek-ai/DeepSeek-V2-Lite"
|
|
tasks:
|
|
- name: "gsm8k"
|
|
metrics:
|
|
- name: "exact_match,strict-match"
|
|
value: 0.385
|
|
- name: "exact_match,flexible-extract"
|
|
value: 0.385
|
|
tensor_parallel_size: 2
|
|
batch_size: 32
|
|
gpu_memory_utilization: 0.7
|
|
apply_chat_template: False
|
|
fewshot_as_multiturn: False
|
|
trust_remote_code: True
|
|
enforce_eager: True
|