huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-06-19 15:52:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

Remove useless modifications

Browse Source 
Signed-off-by: yuanwu <yuan.wu@intel.com>
This commit is contained in:
yuanwu 2025-06-06 07:06:19 +00:00
parent 4a89f59ec7
commit 1505d4687a
2 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
backends/gaudi/server/text_generation_server/layers/attention/hpu.py
View File

@ -88,6 +88,7 @@ def attention(
_, kv_head_num, head_size = key.shape _, kv_head_num, head_size = key.shape
query = query.view(bs, -1, head_num, head_size).transpose(1, 2) query = query.view(bs, -1, head_num, head_size).transpose(1, 2)
key = key.view(bs, -1, kv_head_num, head_size).transpose(1, 2) key = key.view(bs, -1, kv_head_num, head_size).transpose(1, 2)
value = value.view(bs, -1, kv_head_num, head_size).transpose(1, 2)
attn_output = fsdpa_op( attn_output = fsdpa_op(
query, query,
key, key,

1
backends/gaudi/server/text_generation_server/layers/tensor_parallel.py
View File

@ -155,7 +155,6 @@ class TensorParallelColumnLinear(SuperLayer):
@classmethod @classmethod
def load_multi(cls, config, prefixes: List[str], weights, bias: bool, dim: int): def load_multi(cls, config, prefixes: List[str], weights, bias: bool, dim: int):
print(f"bias: {bias}")
if config.quantize == "exl2": if config.quantize == "exl2":
linears = [] linears = []
for prefix in prefixes: for prefix in prefixes: