|
@@ -66,7 +66,6 @@ TP_SIZE = "2"
|
|
|
DCP_SIZE = "1"
|
|
DCP_SIZE = "1"
|
|
|
MODEL_ID = "deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"
|
|
MODEL_ID = "deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"
|
|
|
DSPARK_MODEL_ID = "deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"
|
|
DSPARK_MODEL_ID = "deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"
|
|
|
-DSPARK_MODEL_REVISION = "e46e16bf6035c6f317eb2ac7458eb0362926d402"
|
|
|
|
|
SERVED_MODEL_NAME = "deepseek-v4-vision"
|
|
SERVED_MODEL_NAME = "deepseek-v4-vision"
|
|
|
GPU_MEMORY_UTILIZATION = "0.80"
|
|
GPU_MEMORY_UTILIZATION = "0.80"
|
|
|
KV_CACHE_MEMORY_BYTES = "12884901888"
|
|
KV_CACHE_MEMORY_BYTES = "12884901888"
|