Use no warmup

Author: metaskills

Dockerfile

@@ -1,14 +1,3 @@
-# Temporary server warmup option workaround.
-# https://github.com/Mozilla-Ocho/llamafile/issues/485
-FROM node:20 as llamafile-builder
-RUN mkdir /opt/local && \
-    cd /tmp && \
-    git clone https://github.com/Mozilla-Ocho/llamafile.git && \
-    cd llamafile && \
-    git checkout 21a30bed && \
-    make && \
-    make install PREFIX=/opt/local
-
 FROM node:20
 
 # Create llamafile files
@@ -24,9 +13,6 @@ RUN apt-get update && \
 RUN wget -O /usr/bin/ape https://cosmo.zip/pub/cosmos/bin/ape-$(uname -m).elf && \
     chmod +x /usr/bin/ape
 
-# Copy llamafile build.
-COPY --from=llamafile-builder /opt/local /usr/local
-
 # Lambda Web Adapter
 COPY --from=public.ecr.aws/awsguru/aws-lambda-adapter:0.8.3 /lambda-adapter /opt/extensions/lambda-adapter
 ENV AWS_LWA_INVOKE_MODE=response_stream

bin/build

@@ -4,7 +4,6 @@ set -e
 source ./variables.sh
 
 mkdir -p $llamafiledir
-mkdir -p ./tmp/bin
 
 if [ ! -f "$llamafiledir/$llamafile" ]; then
   curl -L "$llamafileurl" -o "$llamafiledir/$llamafile"

src/command

@@ -1,21 +1,16 @@
 #!/bin/sh
 set -e
 
-LLAMAFILE_PATH=${LLAMAFILE_PATH:-$(find "/opt/llamafile" -name "*.llamafile" -o -name "*.gguf" -type f)}
+LLAMAFILE_PATH=${LLAMAFILE_PATH:-$(find "/opt/llamafile" -name "*.llamafile" -type f)}
 
 echo "Starting llamafile ${LLAMAFILE_PATH}..."
 
-# Options that might make sense for you:
-# 
-# --path /nonexistent
-# Avoid serving the non-API web interface.
-
-llamafile \
-  --model $LLAMAFILE_PATH \
+$LLAMAFILE_PATH \
   --server \
   --fast \
   --threads "$(nproc)" \
   --gpu DISABLE \
+  --no-warmup \
   --timeout 600 \
   --nobrowser \
   --log-disable \

variables.sh

@@ -1,12 +1,4 @@
-
 llamafile=gemma-2-9b-it.Q2_K.llamafile
 llamafileurl="https://huggingface.co/jartine/gemma-2-9b-it-llamafile/resolve/main/gemma-2-9b-it.Q2_K.llamafile"
-
-# llamafile=Mistral-7B-Instruct-v0.3.Q4_0.llamafile
-# llamafileurl="https://huggingface.co/Mozilla/Mistral-7B-Instruct-v0.3-llamafile/resolve/main/Mistral-7B-Instruct-v0.3.Q4_0.llamafile"
-
-# llamafile=Phi-3-mini-4k-instruct.Q2_K.llamafile
-# llamafileurl="https://huggingface.co/Mozilla/Phi-3-mini-4k-instruct-llamafile/resolve/main/Phi-3-mini-4k-instruct.Q2_K.llamafile"
-
 llamafileimage="llamafile-on-lambda"
 llamafiledir="./tmp/llamafile"