expanded example.env with new LLM model, increased MAX_ARTICLE_LENGTH, updated context size, and extended .gitignore entries

.gitignore

@@ -60,3 +60,6 @@ htmlcov/
 !.yarn/versions
 backend-rust/owlynews.sqlite3
 backend-rust/target
+/backend-rust/config.toml
+/backend-rust/owlynews.sqlite3-shm
+/backend-rust/owlynews.sqlite3-wal

backend/app/config.py

@@ -12,7 +12,7 @@ LLM_MODEL = os.getenv("LLM_MODEL", "mistral-nemo:12b")
 LLM_TIMEOUT_SECONDS = int(os.getenv("LLM_TIMEOUT_SECONDS", 180))
 OLLAMA_API_TIMEOUT_SECONDS = int(os.getenv("OLLAMA_API_TIMEOUT_SECONDS", 10))
 ARTICLE_FETCH_TIMEOUT = int(os.getenv("ARTICLE_FETCH_TIMEOUT", 30))
-MAX_ARTICLE_LENGTH = int(os.getenv("MAX_ARTICLE_LENGTH", 5000))
+MAX_ARTICLE_LENGTH = int(os.getenv("MAX_ARTICLE_LENGTH", 10_000))
 
 frontend_path = os.path.join(
     os.path.dirname(os.path.dirname(os.path.dirname(__file__))),

backend/app/services.py

@@ -209,7 +209,7 @@ class NewsFetcher:
             "format": "json",
             "options": {
                 "num_gpu": 1,  # Force GPU usage
-                "num_ctx": 8192,  # Context size
+                "num_ctx": 128_000,  # Context size
             }
         }
 

backend/example.env

@@ -14,6 +14,7 @@ SYNC_COOLDOWN_MINUTES=30
 LLM_MODEL=qwen2:7b-instruct-q4_K_M
 LLM_MODEL=phi3:3.8b-mini-128k-instruct-q4_0
 LLM_MODEL=mistral-nemo:12b
+LLM_MODEL=cnjack/mistral-samll-3.1:24b-it-q4_K_S
 
 # Timeout in seconds for LLM requests
 LLM_TIMEOUT_SECONDS=180