Revert ctypes argtype change

llama_cpp/llama_cpp.py

@@ -175,6 +175,7 @@ llama_progress_callback = ctypes.CFUNCTYPE(None, c_float, c_void_p)
 #     // context pointer passed to the progress callback
 #     void * progress_callback_user_data;
 
+
 #     // Keep the booleans together to avoid misalignment during copy-by-value.
 #     bool low_vram;   // if true, reduce VRAM usage at the cost of performance
 #     bool f16_kv;     // use fp16 for KV cache
@@ -748,7 +749,12 @@ def llama_get_vocab(
     return _lib.llama_get_vocab(ctx, strings, scores, capacity)
 
 
-_lib.llama_get_vocab.argtypes = [llama_context_p, Array[c_char_p], Array[c_float], c_int]
+_lib.llama_get_vocab.argtypes = [
+    llama_context_p,
+    POINTER(c_char_p),
+    POINTER(c_float),
+    c_int,
+]
 _lib.llama_get_vocab.restype = c_int
 
 
@@ -766,6 +772,15 @@ def llama_get_vocab_from_model(
     return _lib.llama_get_vocab_from_model(model, strings, scores, capacity)
 
 
+_lib.llama_get_vocab_from_model.argtypes = [
+    llama_model_p,
+    POINTER(c_char_p),
+    POINTER(c_float),
+    c_int,
+]
+_lib.llama_get_vocab_from_model.restype = c_int
+
+
 # Token logits obtained from the last call to llama_eval()
 # The logits for the last token are stored in the last row
 # Can be mutated in order to change the probabilities of the next token