Fix up env

default.env

@@ -37,3 +37,9 @@ CACHE = 1
 # Set number of threads to use, currently, a standard thread will utilize 1 whole core
 # I usually will set this between all cores I physcally have OR 2 cores less to allow for other processes.
 N_THREADS = 4
+
+# Always use MMAP unless you know what you are doing
+USE_MMAP=1
+
+# Only use MLOCK if you know what it does!
+USE_MLOCK=0

docker-compose.yml

@@ -11,6 +11,7 @@ services:
       - ${DATA_DIR}:/usr/src/app/models
     environment:
       - HOST=llama-python-server
+      - MODEL=./models/ggml-vic7b-q4_0.bin
   llama-python-djs-bot:
     container_name: llama-python-djs-bot
     restart: unless-stopped

server/Dockerfile

@@ -8,4 +8,4 @@ WORKDIR /usr/src/app
 
 RUN pip install --no-cache-dir llama-cpp-python[server]
 
-CMD python3 -m llama_cpp.server --model /usr/src/app/models/gpt4-x-alpaca-13b-native-4bit-128g.bin
+CMD python3 -m llama_cpp.server