fix(server): Update embeddings=False by default. Embeddings should be enabled by default for embedding models.

abetlen · abetlen · commit bf5e0bb4b151 · 2024-07-01T21:29:13.000-04:00
diff --git a/llama_cpp/server/settings.py b/llama_cpp/server/settings.py
@@ -96,7 +96,7 @@ class ModelSettings(BaseSettings):
         default=True, description="if true, use experimental mul_mat_q kernels"
     )
     logits_all: bool = Field(default=True, description="Whether to return logits.")
-    embedding: bool = Field(default=True, description="Whether to use embeddings.")
+    embedding: bool = Field(default=False, description="Whether to use embeddings.")
     offload_kqv: bool = Field(
         default=True, description="Whether to offload kqv to the GPU."
     )

Original file line number	Diff line number	Diff line change
`@@ -96,7 +96,7 @@ class ModelSettings(BaseSettings):`
`96`	`96`	`default=True, description="if true, use experimental mul_mat_q kernels"`
`97`	`97`	`)`
`98`	`98`	`logits_all: bool = Field(default=True, description="Whether to return logits.")`
`99`		`- embedding: bool = Field(default=True, description="Whether to use embeddings.")`
	`99`	`+ embedding: bool = Field(default=False, description="Whether to use embeddings.")`
`100`	`100`	`offload_kqv: bool = Field(`
`101`	`101`	`default=True, description="Whether to offload kqv to the GPU."`
`102`	`102`	`)`