Bugfix: wrong signature for quantize function

abetlen · abetlen · commit 51dbcf269357 · 2023-04-04T22:36:59.000-04:00
diff --git a/llama_cpp/llama_cpp.py b/llama_cpp/llama_cpp.py
@@ -125,12 +125,12 @@ def llama_free(ctx: llama_context_p):
 # TODO: not great API - very likely to change
 # Returns 0 on success
 def llama_model_quantize(
-    fname_inp: bytes, fname_out: bytes, itype: c_int, qk: c_int
+    fname_inp: bytes, fname_out: bytes, itype: c_int
 ) -> c_int:
-    return _lib.llama_model_quantize(fname_inp, fname_out, itype, qk)
+    return _lib.llama_model_quantize(fname_inp, fname_out, itype)
 
 
-_lib.llama_model_quantize.argtypes = [c_char_p, c_char_p, c_int, c_int]
+_lib.llama_model_quantize.argtypes = [c_char_p, c_char_p, c_int]
 _lib.llama_model_quantize.restype = c_int
 
 # Returns the KV cache that will contain the context for the
