0xez
diff --git a/‎llama_cpp/llama_cpp.py
Copy file name to clipboardExpand all lines: llama_cpp/llama_cpp.py
+9Lines changed: 9 additions & 0 deletions b/‎llama_cpp/llama_cpp.py
Copy file name to clipboardExpand all lines: llama_cpp/llama_cpp.py
+9Lines changed: 9 additions & 0 deletions
diff --git a/‎vendor/llama.cpp
Copy file name to clipboard b/‎vendor/llama.cpp
Copy file name to clipboard
@@ -230,6 +230,15 @@ def _load_shared_library(lib_base_name: str):
 LLAMA_ROPE_SCALING_YARN = 2
 LLAMA_ROPE_SCALING_MAX_VALUE = LLAMA_ROPE_SCALING_YARN
 
+# enum llama_pooling_type {
+#     LLAMA_POOLING_NONE = 0,
+#     LLAMA_POOLING_MEAN = 1,
+#     LLAMA_POOLING_CLS  = 2,
+# };
+LLAMA_POOLING_NONE = 0
+LLAMA_POOLING_MEAN = 1
+LLAMA_POOLING_CLS = 2
+
 # enum llama_split_mode {
 #     LLAMA_SPLIT_NONE    = 0, // single GPU
 #     LLAMA_SPLIT_LAYER   = 1, // split layers and KV across GPUs