Rename SearchModel to SearchModelConfig DB model, Require Cross-Encoder

src/database/adapters/__init__.py

@@ -31,7 +31,7 @@ from database.models import (
     GithubRepoConfig,
     Conversation,
     ChatModelOptions,
-    SearchModel,
+    SearchModelConfig,
     Subscription,
     UserConversationConfig,
     OpenAIProcessorConversationConfig,
@@ -216,9 +216,9 @@ async def set_user_github_config(user: KhojUser, pat_token: str, repos: list):
 
 
 def get_or_create_search_model():
-    search_model = SearchModel.objects.filter().first()
+    search_model = SearchModelConfig.objects.filter().first()
     if not search_model:
-        search_model = SearchModel.objects.create()
+        search_model = SearchModelConfig.objects.create()
 
     return search_model
 

src/database/admin.py

@@ -8,7 +8,7 @@ from database.models import (
     ChatModelOptions,
     OpenAIProcessorConversationConfig,
     OfflineChatProcessorConversationConfig,
-    SearchModel,
+    SearchModelConfig,
     Subscription,
 )
 
@@ -17,5 +17,5 @@ admin.site.register(KhojUser, UserAdmin)
 admin.site.register(ChatModelOptions)
 admin.site.register(OpenAIProcessorConversationConfig)
 admin.site.register(OfflineChatProcessorConversationConfig)
-admin.site.register(SearchModel)
+admin.site.register(SearchModelConfig)
 admin.site.register(Subscription)

src/database/migrations/0018_searchmodelconfig_delete_searchmodel.py

@@ -0,0 +1,30 @@
+# Generated by Django 4.2.5 on 2023-11-16 01:13
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("database", "0017_searchmodel"),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name="SearchModelConfig",
+            fields=[
+                ("id", models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name="ID")),
+                ("created_at", models.DateTimeField(auto_now_add=True)),
+                ("updated_at", models.DateTimeField(auto_now=True)),
+                ("name", models.CharField(default="default", max_length=200)),
+                ("model_type", models.CharField(choices=[("text", "Text")], default="text", max_length=200)),
+                ("bi_encoder", models.CharField(default="thenlper/gte-small", max_length=200)),
+                ("cross_encoder", models.CharField(default="cross-encoder/ms-marco-MiniLM-L-6-v2", max_length=200)),
+            ],
+            options={
+                "abstract": False,
+            },
+        ),
+        migrations.DeleteModel(
+            name="SearchModel",
+        ),
+    ]

src/database/models/__init__.py

@@ -102,16 +102,14 @@ class LocalPlaintextConfig(BaseModel):
     user = models.ForeignKey(KhojUser, on_delete=models.CASCADE)
 
 
-class SearchModel(BaseModel):
+class SearchModelConfig(BaseModel):
     class ModelType(models.TextChoices):
         TEXT = "text"
 
     name = models.CharField(max_length=200, default="default")
     model_type = models.CharField(max_length=200, choices=ModelType.choices, default=ModelType.TEXT)
     bi_encoder = models.CharField(max_length=200, default="thenlper/gte-small")
-    cross_encoder = models.CharField(
+    cross_encoder = models.CharField(max_length=200, default="cross-encoder/ms-marco-MiniLM-L-6-v2")
-        max_length=200, default="cross-encoder/ms-marco-MiniLM-L-6-v2", null=True, blank=True
-    )
 
 
 class OpenAIProcessorConversationConfig(BaseModel):

src/khoj/migrations/migrate_server_pg.py

@@ -64,7 +64,7 @@ from database.models import (
     OpenAIProcessorConversationConfig,
     OfflineChatProcessorConversationConfig,
     ChatModelOptions,
-    SearchModel,
+    SearchModelConfig,
 )
 
 logger = logging.getLogger(__name__)
@@ -87,12 +87,12 @@ def migrate_server_pg(args):
         if "search-type" in raw_config and raw_config["search-type"]:
             if "asymmetric" in raw_config["search-type"]:
                 # Delete all existing search models
-                SearchModel.objects.filter(model_type=SearchModel.ModelType.TEXT).delete()
+                SearchModelConfig.objects.filter(model_type=SearchModelConfig.ModelType.TEXT).delete()
                 # Create new search model from existing Khoj YAML config
                 asymmetric_search = raw_config["search-type"]["asymmetric"]
-                SearchModel.objects.create(
+                SearchModelConfig.objects.create(
                     name="default",
-                    model_type=SearchModel.ModelType.TEXT,
+                    model_type=SearchModelConfig.ModelType.TEXT,
                     bi_encoder=asymmetric_search.get("encoder"),
                     cross_encoder=asymmetric_search.get("cross-encoder"),
                 )

tests/helpers.py

@@ -7,7 +7,7 @@ from database.models import (
     ChatModelOptions,
     OfflineChatProcessorConversationConfig,
     OpenAIProcessorConversationConfig,
-    SearchModel,
+    SearchModelConfig,
     UserConversationConfig,
     Conversation,
     Subscription,
@@ -74,7 +74,7 @@ class ConversationFactory(factory.django.DjangoModelFactory):
 
 class SearchModelFactory(factory.django.DjangoModelFactory):
     class Meta:
-        model = SearchModel
+        model = SearchModelConfig
 
     name = "default"
     model_type = "text"