fix compile (#341)

pytorch · Jun 11, 2024 · 7cddc6e · 7cddc6e
1 parent 79f2c7f
commit 7cddc6e
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/scripts/hf_eval.py b/scripts/hf_eval.py
@@ -22,7 +22,7 @@ def run_evaluation(repo_id, task_list, limit, device, precision, quantization, c
     model = AutoModelForCausalLM.from_pretrained(repo_id).to(device="cuda", dtype=precision)
 
     if compile:
-        torch.compile(model, mode="max-autotune", fullgraph=True)
+        model = torch.compile(model, mode="max-autotune", fullgraph=True)
 
     if quantization == "int8dq":
         change_linear_weights_to_int8_dqtensors(model)