From 9a95334503904d1f9952018bf0ccd5286f587ac9 Mon Sep 17 00:00:00 2001
From: lebrunel <124721263+lebrunel@users.noreply.github.com>
Date: Sat, 11 May 2024 17:56:16 +0100
Subject: [PATCH] support model quanitze option

---
 lib/ollama.ex | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/lib/ollama.ex b/lib/ollama.ex
index 6f09a6b..802eb6a 100644
--- a/lib/ollama.ex
+++ b/lib/ollama.ex
@@ -251,10 +251,6 @@ defmodule Ollama do
       required: true,
       doc: "List of messages - used to keep a chat memory.",
     ],
-    template: [
-      type: :string,
-      doc: "Prompt template, overriding the model default.",
-    ],
     format: [
       type: :string,
       doc: "Set the expected format of the response (`json`).",
@@ -419,6 +415,10 @@ defmodule Ollama do
       required: true,
       doc: "Contents of the Modelfile.",
     ],
+    quantize: [
+      type: :string,
+      doc: "Quantize f16 and f32 models when importing them.",
+    ],
     stream: [
       type: {:or, [:boolean, :pid]},
       default: false,