Merge branch 'dream-build' into dream-build-glm

Dockerfile

@@ -18,7 +18,7 @@ RUN apt-get update && apt-get install -y git \
     CUDA_HOME=/usr/local/cuda \
     TORCH_CUDA_ARCH_LIST="10.0" \
     pip install --no-cache-dir --no-build-isolation . && \
-    rm -rf /tmp/lmcache && export CACHE_BUSTER=3
+    rm -rf /tmp/lmcache && export CACHE_BUSTER=4
 
 
 # Make sure we have patch to make MTP work on GLM

lmcache-config-glm-51.yaml

@@ -1,7 +1,7 @@
 chunk_size: 256
 local_cpu: true
 max_local_cpu_size: 256.0
-save_decode_cache: false
+save_decode_cache: true
 enable_lazy_memory_allocator: true
 lazy_memory_initial_ratio: 1.0
 use_gpu_connector_v3: true