Remove flash attention property shim

Author: abetlen

llama_cpp/llama.py

@@ -341,7 +341,11 @@ def __init__(
         self._logits_all = logits_all if draft_model is None else True
         self.context_params.embeddings = embedding  # TODO: Rename to embeddings
         self.context_params.offload_kqv = offload_kqv
-        self.context_params.flash_attn = flash_attn
+        self.context_params.flash_attn_type = (
+            llama_cpp.LLAMA_FLASH_ATTN_TYPE_ENABLED
+            if flash_attn
+            else llama_cpp.LLAMA_FLASH_ATTN_TYPE_DISABLED
+        )
 
         if op_offload is not None:
             self.context_params.op_offload = op_offload
@@ -2103,7 +2107,10 @@ def __getstate__(self):
             logits_all=self._logits_all,
             embedding=self.context_params.embeddings,
             offload_kqv=self.context_params.offload_kqv,
-            flash_attn=self.context_params.flash_attn,
+            flash_attn=(
+                self.context_params.flash_attn_type
+                == llama_cpp.LLAMA_FLASH_ATTN_TYPE_ENABLED
+            ),
             op_offload=self.context_params.op_offload,
             swa_full=self.context_params.swa_full,
             # Sampling Params

llama_cpp/llama_chat_format.py

@@ -2757,7 +2757,10 @@ def _init_mtmd_context(self, llama_model: llama.Llama):
             ctx_params.n_threads = llama_model.n_threads
             ctx_params.flash_attn_type = (
                 llama_cpp.LLAMA_FLASH_ATTN_TYPE_ENABLED
-                if llama_model.context_params.flash_attn
+                if (
+                    llama_model.context_params.flash_attn_type
+                    == llama_cpp.LLAMA_FLASH_ATTN_TYPE_ENABLED
+                )
                 else llama_cpp.LLAMA_FLASH_ATTN_TYPE_DISABLED
             )
 

llama_cpp/llama_cpp.py

@@ -993,18 +993,6 @@ class llama_context_params(ctypes.Structure):
         ("n_samplers", ctypes.c_size_t),
     ]
 
-    @property
-    def flash_attn(self) -> bool:
-        return self.flash_attn_type == LLAMA_FLASH_ATTN_TYPE_ENABLED
-
-    @flash_attn.setter
-    def flash_attn(self, enabled: bool):
-        self.flash_attn_type = (
-            LLAMA_FLASH_ATTN_TYPE_ENABLED
-            if enabled
-            else LLAMA_FLASH_ATTN_TYPE_DISABLED
-        )
-
 
 # // Signature for logging events
 # // Note that text includes the new line character at the end for most events.

tests/test_llama.py

@@ -83,7 +83,7 @@ def test_real_model(llama_cpp_model_path):
     cparams.n_threads = multiprocessing.cpu_count()
     cparams.n_threads_batch = multiprocessing.cpu_count()
     cparams.logits_all = False
-    cparams.flash_attn = True
+    cparams.flash_attn_type = llama_cpp.LLAMA_FLASH_ATTN_TYPE_ENABLED
 
     context = internals.LlamaContext(model=model, params=cparams)
     tokens = model.tokenize(b"Hello, world!", add_bos=True, special=True)