fix: allow for transformers_version

Dockerfile

@@ -53,6 +53,7 @@ RUN if [ "${VLLM_NIGHTLY}" = "true" ]; then \
 fi
 
 COPY src /src
+RUN chmod +x /src/start.sh
 RUN --mount=type=secret,id=HF_TOKEN,required=false \
     if [ -f /run/secrets/HF_TOKEN ]; then \
     export HF_TOKEN=$(cat /run/secrets/HF_TOKEN); \
@@ -62,4 +63,4 @@ RUN --mount=type=secret,id=HF_TOKEN,required=false \
     fi
 
 # Start the handler
-CMD ["python3", "/src/handler.py"]
+CMD ["/bin/bash", "/src/start.sh"]

builder/requirements.txt

@@ -9,7 +9,7 @@ typing-extensions>=4.8.0
 pydantic
 pydantic-settings
 hf-transfer
-transformers>=5.2.0,<5.3.0
+transformers>= 4.57.0,< 5
 bitsandbytes>=0.45.0
 kernels
 torch-c-dlpack-ext

src/start.sh

@@ -0,0 +1,9 @@
+#!/bin/bash
+set -e
+
+if [ -n "${TRANSFORMERS_VERSION}" ]; then
+    echo "Installing transformers==${TRANSFORMERS_VERSION}"
+    uv pip install --system "transformers==${TRANSFORMERS_VERSION}"
+fi
+
+exec python3 /src/handler.py