root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-27 03:44:35 +00:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

remove .rotary_pos_emb.inv_freq and unuse code for chatglm3 model

Browse Source 
Signed-off-by: XingXing Qiao <qiaoxx@dingdao.com>
This commit is contained in:
XingXing Qiao 2024-05-15 11:00:04 +08:00
parent 6630a2da48
commit 5a914ffce0
1 changed files with 8 additions and 75 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

83
convert-hf-to-gguf.py
View File

@ -2790,85 +2790,18 @@ class ChatGLMModel(Model):
self.gguf_writer.add_rope_dimension_count(64) self.gguf_writer.add_rope_dimension_count(64)
self.gguf_writer.add_add_bos_token(False) self.gguf_writer.add_add_bos_token(False)
def write_tensors(self): def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iterable[tuple[str, Tensor]]:
block_count = self.hparams["num_layers"] if name.endswith(".rotary_pos_emb.inv_freq"):
tensors = dict(self.get_tensors()) return []
tensor_map = gguf.get_tensor_name_map(self.model_arch, block_count)
has_lm_head = True
n_head = self.hparams.get("n_head", self.hparams.get("num_attention_heads"))
n_embed = self.hparams.get("hidden_size", self.hparams.get("n_embed"))
for name, data_torch in tensors.items(): del bid # unused
if name.endswith(".rotary_pos_emb.inv_freq"):
continue
if "lm_head.weight" not in tensors.keys() and "output.weight" not in tensors.keys(): name = re.sub(r'transformer\.', '', name)
has_lm_head = False
name = re.sub(r'transformer\.', '', name) if name == "word_embeddings.weight":
assert self.tensor_names is not None
old_dtype = data_torch.dtype return [(self.map_tensor_name(name), data_torch)]
# convert any unsupported data types to float32
if data_torch.dtype not in (torch.float16, torch.float32):
data_torch = data_torch.to(torch.float32)
data = data_torch.squeeze().numpy()
if re.match(r"h\.\d+\.self_attention\.query_key_value\.weight", name):
# Map bloom-style qkv_linear to gpt-style qkv_linear
# bloom: https://github.com/huggingface/transformers/blob/main/src/transformers/models/bloom/modeling_bloom.py#L238-L252 # noqa
# gpt-2: https://github.com/huggingface/transformers/blob/main/src/transformers/models/gpt2/modeling_gpt2.py#L312 # noqa
qkv_weights = data.reshape((n_head, 3, n_embed // n_head, n_embed))
data = np.concatenate(
(
qkv_weights[:, 0, :, :].reshape((-1, n_embed)),
qkv_weights[:, 1, :, :].reshape((-1, n_embed)),
qkv_weights[:, 2, :, :].reshape((-1, n_embed)),
),
axis=0,
)
print("re-format attention.linear_qkv.weight")
elif re.match(r"h\.\d+\.self_attention\.query_key_value\.bias", name):
qkv_bias = data.reshape((n_head, 3, n_embed // n_head))
data = np.concatenate(
(
qkv_bias[:, 0, :].reshape((n_embed,)),
qkv_bias[:, 1, :].reshape((n_embed,)),
qkv_bias[:, 2, :].reshape((n_embed,)),
),
axis=0,
)
print("re-format attention.linear_qkv.bias")
# map tensor names
new_name = tensor_map.get_name(name, try_suffixes=(".weight", ".bias"))
if new_name is None:
print(f"Can not map tensor {name!r}")
sys.exit()
n_dims = len(data.shape)
data_dtype = data.dtype
# if f32 desired, convert any float16 to float32
if self.ftype == 0 and data_dtype == np.float16:
data = data.astype(np.float32)
# TODO: Why cant we use these float16 as-is? There should be not reason to store float16 as float32
if self.ftype == 1 and data_dtype == np.float16 and n_dims == 1:
data = data.astype(np.float32)
# if f16 desired, convert any float32 2-dim weight tensors to float16
if self.ftype == 1 and data_dtype == np.float32 and name.endswith(".weight") and n_dims == 2:
data = data.astype(np.float16)
print(f"=> {new_name}, shape = {data.shape}, {old_dtype} --> {data.dtype}")
self.gguf_writer.add_tensor(new_name, data)
if not has_lm_head and name == "word_embeddings.weight":
self.gguf_writer.add_tensor("output.weight", data)
print(name, f"=> output.weight, shape = {data.shape}, {old_dtype} --> {data.dtype}")
###### CONVERSION LOGIC ###### ###### CONVERSION LOGIC ######