Update README.md

README.md

@@ -91,13 +91,17 @@ if __name__ == "__main__":
   p = psutil.Process()
   p.cpu_affinity([0, 1, 2, 3])
   torch.set_num_threads(4)
+  transformers.logging.set_verbosity_error()
+  logging.disable(logging.CRITICAL)
 
   tokenizer = AutoTokenizer.from_pretrained("ALMA-Ja-V3-amd-npu")
-  ckpt = "alma_w_bit_4_awq_fa_amd.pt"
+  tokenizer.pad_token = tokenizer.eos_token
+  ckpt = r"ALMA-Ja-V3-amd-npu\alma_w_bit_4_awq_fa_amd.pt"
 
   model = torch.load(ckpt)
   model.eval()
   model = model.to(torch.bfloat16)
+ 
 
   for n, m in model.named_modules():
       if isinstance(m, qlinear.QLinearPerGrp):
@@ -105,20 +109,9 @@ if __name__ == "__main__":
           m.device = "aie"
           m.quantize_weights()
 
-    model = torch.load(ckpt)
-    model.eval()
-    model = model.to(torch.bfloat16)
-
-    for n, m in model.named_modules():
-        if isinstance(m, qlinear.QLinearPerGrp):
-            print(f"Preparing weights of layer : {n}")
-            m.device = "aie"
-            m.quantize_weights()
-
-  print("Translate Japanese to English.", "面白きこともなき世を面白く住みなすものは心なりけり")
-  print("Translate English to Japanese.", "If today were the last day of your life, would you want to do what you are about to do today.")
-
 
+  print(translation("Translate Japanese to English.", "面白きこともなき世を面白く住みなすものは心なりけり"))
+  print(translation("Translate English to Japanese.", "Join me, and together we can rule the galaxy as father and son."))
 ```
 
 ![chat_image](alma-v3.png)