biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[CI] Prune tests/models/decoder_only/language/* tests (#9940)

Browse Source 
Signed-off-by: mgoin <michael@neuralmagic.com>
This commit is contained in:
Michael Goin
2024-11-05 16:02:23 -05:00
committed by GitHub
parent b9c64c0ca7
commit 02462465ea
9 changed files with 70 additions and 270 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
tests/models/decoder_only/language/test_gptq_marlin_24.py
View File

@@ -25,16 +25,16 @@ model_pairs = [
# 4-bit, group_size == 128
ModelPair(model_marlin="alexm-nm/tinyllama-24-marlin24-4bit-g128",
model_gptq="alexm-nm/tinyllama-24-gptq-4bit-g128"),
# 4-bit, group_size == channelwise
ModelPair(model_marlin="alexm-nm/tinyllama-24-marlin24-4bit-channelwise",
model_gptq="alexm-nm/tinyllama-24-gptq-4bit-channelwise"),
# # 4-bit, group_size == channelwise
# ModelPair(model_marlin="alexm-nm/tinyllama-24-marlin24-4bit-channelwise",
# model_gptq="alexm-nm/tinyllama-24-gptq-4bit-channelwise"),
# 8-bit, group_size == 128
ModelPair(model_marlin="alexm-nm/tinyllama-24-marlin24-8bit-g128",
model_gptq="alexm-nm/tinyllama-24-gptq-8bit-g128"),
# 8-bit, group_size == channelwise
ModelPair(model_marlin="alexm-nm/tinyllama-24-marlin24-8bit-channelwise",
model_gptq="alexm-nm/tinyllama-24-gptq-8bit-channelwise"),
# # 8-bit, group_size == channelwise
# ModelPair(model_marlin="alexm-nm/tinyllama-24-marlin24-8bit-channelwise",
# model_gptq="alexm-nm/tinyllama-24-gptq-8bit-channelwise"),
]