@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / fe470ae5ad
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git

Synced Nov 17, 06:45 UTC

[Minor] Fix code style for baichuan (#2003)

Woosuk Kwon committed Dec 9, 2023 at 19:24 UTC fe470ae5ade3d4b04ec5ee6f26ffb681d66a8ca1
1 file changed +4 -3
vllm/model_executor/models/baichuan.py
+4 -3
@@ -366,8 +366,8 @@ class BaiChuanBaseForCausalLM(nn.Module):
366 weight_loader(param, loaded_weight)
367
368
369 -class BaichuanForCausalLM(BaiChuanBaseForCausalLM
370 - ): # baichuan 13b, baichuan2 13b, baichuan2 7b
369 +class BaichuanForCausalLM(BaiChuanBaseForCausalLM):
370 + """Baichuan 13B and Baichuan2 7B/13B."""
371
372 def __init__(self,
373 config,
@@ -378,7 +378,8 @@ class BaichuanForCausalLM(BaiChuanBaseForCausalLM
378 super().__init__(config, "ALIBI", linear_method)
379
380
381 -class BaiChuanForCausalLM(BaiChuanBaseForCausalLM): # baichuan 7b
381 +class BaiChuanForCausalLM(BaiChuanBaseForCausalLM):
382 + """Baichuan 7B."""
383
384 def __init__(self,
385 config,