mirror of
https://github.com/abetlen/llama-cpp-python.git
synced 2023-09-07 17:34:22 +03:00
Update llama.cpp
This commit is contained in:
@@ -71,7 +71,7 @@ LLAMA_FILE_VERSION = ctypes.c_int(1)
|
|||||||
LLAMA_FILE_MAGIC = b"ggjt"
|
LLAMA_FILE_MAGIC = b"ggjt"
|
||||||
LLAMA_FILE_MAGIC_UNVERSIONED = b"ggml"
|
LLAMA_FILE_MAGIC_UNVERSIONED = b"ggml"
|
||||||
LLAMA_SESSION_MAGIC = b"ggsn"
|
LLAMA_SESSION_MAGIC = b"ggsn"
|
||||||
LLAMA_SESSION_VERSION = ctypes.c_int(0)
|
LLAMA_SESSION_VERSION = ctypes.c_int(1)
|
||||||
|
|
||||||
llama_context_p = c_void_p
|
llama_context_p = c_void_p
|
||||||
|
|
||||||
@@ -239,7 +239,8 @@ _lib.llama_set_rng_seed.argtypes = [llama_context_p, c_int]
|
|||||||
_lib.llama_set_rng_seed.restype = None
|
_lib.llama_set_rng_seed.restype = None
|
||||||
|
|
||||||
|
|
||||||
# Returns the size in bytes of the state (rng, logits, embedding and kv_cache)
|
# Returns the maximum size in bytes of the state (rng, logits, embedding
|
||||||
|
# and kv_cache) - will often be smaller after compacting tokens
|
||||||
def llama_get_state_size(ctx: llama_context_p) -> c_size_t:
|
def llama_get_state_size(ctx: llama_context_p) -> c_size_t:
|
||||||
return _lib.llama_get_state_size(ctx)
|
return _lib.llama_get_state_size(ctx)
|
||||||
|
|
||||||
|
|||||||
2
vendor/llama.cpp
vendored
2
vendor/llama.cpp
vendored
Submodule vendor/llama.cpp updated: 58b367c2d7...e216aa0463
Reference in New Issue
Block a user