From b33c92d1bd5c1da325ee7b19254352cce8fd31b6 Mon Sep 17 00:00:00 2001 From: Daniel Han-Chen Date: Fri, 2 Feb 2024 22:22:42 +1100 Subject: [PATCH] Update llama.py --- unsloth/models/llama.py | 2 -- 1 file changed, 2 deletions(-) diff --git a/unsloth/models/llama.py b/unsloth/models/llama.py index e3ec21a3f4..127554ff19 100644 --- a/unsloth/models/llama.py +++ b/unsloth/models/llama.py @@ -681,8 +681,6 @@ import torch._dynamo.config import torch._inductor.config torch._inductor.config.coordinate_descent_tuning = True torch._inductor.config.triton.unique_kernel_names = True -torch._inductor.config.fx_graph_cache = True # Experimental feature to reduce compilation times, will be on by default in future - # https://github.com/huggingface/transformers/blob/main/src/transformers/models/llama/modeling_llama.py#L825 @torch.compile(mode = "reduce-overhead", fullgraph = True)