Update Dockerfile

Dockerfile

@@ -19,10 +19,10 @@ ENV HF_HOME=/tmp
 # ENV TRANSFORMERS_CACHE=/tmp
 
 # Pre-download K2-Think model to avoid cold start
-RUN python -c "from transformers import AutoTokenizer, AutoModelForCausalLM; \
-    model_id='LLM360/K2-Think'; \
-    AutoTokenizer.from_pretrained(model_id, cache_dir='/tmp'); \
-    AutoModelForCausalLM.from_pretrained(model_id, cache_dir='/tmp')"
+# RUN python -c "from transformers import AutoTokenizer, AutoModelForCausalLM; \
+#     model_id='LLM360/K2-Think'; \
+#     AutoTokenizer.from_pretrained(model_id, cache_dir='/tmp'); \
+#     AutoModelForCausalLM.from_pretrained(model_id, cache_dir='/tmp')"
 
 # Expose port
 EXPOSE 7860