|
@@ -2,8 +2,8 @@ engine = "vllm"
|
|
|
name = "DeepSeek V4 Flash (vLLM)"
|
|
name = "DeepSeek V4 Flash (vLLM)"
|
|
|
docker_image = "aidendle94/sparkrun-vllm-ds4-gb10:production-v2"
|
|
docker_image = "aidendle94/sparkrun-vllm-ds4-gb10:production-v2"
|
|
|
model = "deepseek-ai/DeepSeek-V4-Flash"
|
|
model = "deepseek-ai/DeepSeek-V4-Flash"
|
|
|
-required_nodes = 4
|
|
|
|
|
-tensor_parallel_size = 4
|
|
|
|
|
|
|
+required_nodes = 2
|
|
|
|
|
+tensor_parallel_size = 2
|
|
|
pipeline_parallel_size = 1
|
|
pipeline_parallel_size = 1
|
|
|
distributed_backend = "mp"
|
|
distributed_backend = "mp"
|
|
|
|
|
|
