fix(Model): remove unsupported default model setting params (janhq#1382)

Signed-off-by: James <[email protected]> Co-authored-by: James <[email protected]>
m13810859213 · Jan 5, 2024 · 764e4c6 · 764e4c6
1 parent e0a7d3c
commit 764e4c6
Show file tree

Hide file tree

Showing 2 changed files with 20 additions and 27 deletions.
diff --git a/extensions/inference-nitro-extension/src/module.ts b/extensions/inference-nitro-extension/src/module.ts
@@ -24,19 +24,19 @@ const NVIDIA_INFO_FILE = path.join(
 );
 
 const DEFALT_SETTINGS = {
-  "notify": true,
-  "run_mode": "cpu",
-  "nvidia_driver": {
-    "exist": false,
-    "version": ""
+  notify: true,
+  run_mode: "cpu",
+  nvidia_driver: {
+    exist: false,
+    version: "",
   },
-  "cuda": {
-    "exist": false,
-    "version": ""
+  cuda: {
+    exist: false,
+    version: "",
   },
-  "gpus": [],
-  "gpu_highest_vram": ""
-}
+  gpus: [],
+  gpu_highest_vram: "",
+};
 
 // The subprocess instance for Nitro
 let subprocess = undefined;
@@ -114,8 +114,7 @@ function updateCudaExistence() {
     if (cudaExists) {
       cudaVersion = "11";
     }
-  }
-  else {
+  } else {
     cudaVersion = "12";
   }
 
@@ -395,8 +394,7 @@ function spawnNitroProcess(nitroResourceProbe: any): Promise<any> {
       } else {
         if (nvidiaInfo["cuda"].version === "12") {
           binaryFolder = path.join(binaryFolder, "win-cuda-12-0");
-        }
-        else {
+        } else {
           binaryFolder = path.join(binaryFolder, "win-cuda-11-4");
         }
         cudaVisibleDevices = nvidiaInfo["gpu_highest_vram"];
@@ -416,8 +414,7 @@ function spawnNitroProcess(nitroResourceProbe: any): Promise<any> {
       } else {
         if (nvidiaInfo["cuda"].version === "12") {
           binaryFolder = path.join(binaryFolder, "linux-cuda-12-0");
-        }
-        else {
+        } else {
           binaryFolder = path.join(binaryFolder, "linux-cuda-11-4");
         }
         cudaVisibleDevices = nvidiaInfo["gpu_highest_vram"];

diff --git a/models/config/default-model.json b/models/config/default-model.json
@@ -11,18 +11,14 @@
     "ctx_len": 4096,
     "ngl": 0,
     "embedding": false,
-    "n_parallel": 0,
-    "cpu_threads": 0,
-    "prompt_template": ""
+    "prompt_template": "{system_message}\n### Instruction: {prompt}\n### Response:"
   },
   "parameters": {
-    "temperature": 0,
-    "token_limit": 0,
-    "top_k": 0,
-    "top_p": 0,
-    "stream": false,
-    "max_tokens": 4096,
-    "stop": [],
+    "temperature": 0.7,
+    "top_p": 0.95,
+    "stream": true,
+    "max_tokens": 2048,
+    "stop": ["<endofstring>"],
     "frequency_penalty": 0,
     "presence_penalty": 0
   },