From 11bda4f4bebcd3aba85a444a72220fcdb8e2613d Mon Sep 17 00:00:00 2001 From: Sourashis Roy Date: Mon, 28 Oct 2024 21:20:20 +0000 Subject: [PATCH] Reverting layer changes --- vllm/attention/layer.py | 2 -- 1 file changed, 2 deletions(-) diff --git a/vllm/attention/layer.py b/vllm/attention/layer.py index 58d6dadbda110..33d05cbd3fe01 100644 --- a/vllm/attention/layer.py +++ b/vllm/attention/layer.py @@ -78,11 +78,9 @@ def __init__( # During model initialization, the default dtype is set as the model # weight and activation dtype. dtype = torch.get_default_dtype() - print('dtype ' + str(dtype)) attn_backend = get_attn_backend(head_size, dtype, kv_cache_dtype, block_size, is_attention_free, blocksparse_params is not None) - print('attn_backend ' + str(attn_backend)) impl_cls = attn_backend.get_impl_cls() self.impl = impl_cls(num_heads, head_size, scale, num_kv_heads, alibi_slopes, sliding_window, kv_cache_dtype,