Update app.py

app.py

@@ -19,12 +19,12 @@ from transformers import AutoTokenizer, AutoModelForCausalLM
 import paramiko
 
 torch.backends.cuda.matmul.allow_tf32 = True
-torch.backends.cuda.matmul.allow_bf16_reduced_precision_reduction = False
+torch.backends.cuda.matmul.allow_bf16_reduced_precision_reduction = True
 torch.backends.cuda.matmul.allow_fp16_reduced_precision_reduction = False
 torch.backends.cudnn.allow_tf32 = True
 torch.backends.cudnn.deterministic = False
 torch.backends.cudnn.benchmark = False
-torch.set_float32_matmul_precision("medium")
+torch.set_float32_matmul_precision("high")
 
 FTP_HOST = "1ink.us"
 FTP_USER = "ford442"