huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-11 04:14:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

use exllamav2QuantLinear instead of exllama1

Browse Source
This commit is contained in:
Florian Zimmermeister 2023-10-25 12:26:20 +02:00 committed by Nicolas Patry
parent a02f6839e9
commit f96d997494
1 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
server/text_generation_server/utils/layers.py
View File

@ -38,6 +38,7 @@ if os.getenv("DISABLE_EXLLAMA") == "True":
elif CAN_EXLLAMA:
try:
from text_generation_server.utils.gptq.exllama import Ex4bitLinear
from text_generation_server.utils.gptq.exllamav2 import QuantLinear as exllamav2QuantLinear
HAS_EXLLAMA = True
except ImportError:
@ -308,7 +309,7 @@ def get_linear(weight, bias, quantize):
)
if use_exllama:
linear = Ex4bitLinear(qweight, qzeros, scales, g_idx, bias, bits, groupsize)
linear = exllamav2QuantLinear(qweight, qzeros, scales, g_idx, bias, bits, groupsize)
else:
linear = QuantLinear(
qweight,