chore: fix model card

README.md

@@ -24,10 +24,10 @@ pipeline_tag: translation
 - Template:
 
   ```python
-  prompt = f"Translate this from {src_lang} to {tgt_lang}\n### {src_lang}: {src_text}\n### {tgt_lang}:"
+    prompt = f"Translate this from {src_lang} to {tgt_lang}\n### {src_lang}: {src_text}\n### {tgt_lang}:"
 
-  >>> # src_lang can be 'English', '한국어'
-  >>> # tgt_lang can be '한국어', 'English'
+    >>> # src_lang can be 'English', '한국어'
+    >>> # tgt_lang can be '한국어', 'English'
   ```
 
 - Issue:
@@ -35,11 +35,11 @@ pipeline_tag: translation
   Make sure to use the prompt proposed above.
   
   ```python
-  prompt = f"""Translate this from {src_lang} to {tgt_lang}
-  ### {src_lang}: {src_text}
-  ### {tgt_lang}:"""
+    prompt = f"""Translate this from {src_lang} to {tgt_lang}
+    ### {src_lang}: {src_text}
+    ### {tgt_lang}:"""
 
-  >>> # DO NOT USE this prompt.
+    >>> # DO NOT USE this prompt
   ```
   
   And mind that there is no "space (`_`)" at the end of the prompt.
@@ -52,45 +52,46 @@ pipeline_tag: translation
 
 ### Usage (IMPORTANT)
 - Should remove the EOS token (`<|endoftext|>`, id=46332) at the end of the prompt.
-```python
-# MODEL
-plm_name = 'beomi/open-llama-2-ko-7b'
-adapter_name = 'traintogpb/llama-2-enko-translator-7b-qlora-adapter'
-model = LlamaForCausalLM.from_pretrained(
-    plm_name, 
-    max_length=768,
-    quantization_config=bnb_config, # Use the QLoRA config above
-    attn_implementation='flash_attention_2',
-    torch_dtype=torch.bfloat16
-)
-model = PeftModel.from_pretrained(
-    model, 
-    adapter_name, 
-    torch_dtype=torch.bfloat16
-)
 
-# TOKENIZER
-tokenizer = LlamaTokenizer.from_pretrained(plm_name)
-tokenizer.pad_token = "</s>"
-tokenizer.pad_token_id = 2
-tokenizer.eos_token = "<|endoftext|>" # Must be differentiated from the PAD token
-tokenizer.eos_token_id = 46332
-tokenizer.add_eos_token = True
-tokenizer.model_max_length = 768
+  ```python
+  # MODEL
+  plm_name = 'beomi/open-llama-2-ko-7b'
+  adapter_name = 'traintogpb/llama-2-enko-translator-7b-qlora-adapter'
+  model = LlamaForCausalLM.from_pretrained(
+      plm_name, 
+      max_length=768,
+      quantization_config=bnb_config, # Use the QLoRA config above
+      attn_implementation='flash_attention_2',
+      torch_dtype=torch.bfloat16
+  )
+  model = PeftModel.from_pretrained(
+      model, 
+      adapter_name, 
+      torch_dtype=torch.bfloat16
+  )
 
-# INFERENCE
-text = "NMIXX is the world-best female idol group, who came back with the new song 'DASH'." 
-prompt = f"Translate this from {src_lang} to {tgt_lang}\n### {src_lang}: {src_text}\n### {tgt_lang}:"
+  # TOKENIZER
+  tokenizer = LlamaTokenizer.from_pretrained(plm_name)
+  tokenizer.pad_token = "</s>"
+  tokenizer.pad_token_id = 2
+  tokenizer.eos_token = "<|endoftext|>" # Must be differentiated from the PAD token
+  tokenizer.eos_token_id = 46332
+  tokenizer.add_eos_token = True
+  tokenizer.model_max_length = 768
 
-inputs = tokenizer(prompt, return_tensors="pt", max_length=max_length, truncation=True)
-# REMOVE EOS TOKEN IN THE PROMPT
-inputs['input_ids'] = inputs['input_ids'][0][:-1].unsqueeze(dim=0)
-inputs['attention_mask'] = inputs['attention_mask'][0][:-1].unsqueeze(dim=0)
+  # INFERENCE
+  text = "NMIXX is the world-best female idol group, who came back with the new song 'DASH'." 
+  prompt = f"Translate this from {src_lang} to {tgt_lang}\n### {src_lang}: {src_text}\n### {tgt_lang}:"
 
-outputs = model.generate(**inputs, max_length=max_length, eos_token_id=46332)
+  inputs = tokenizer(prompt, return_tensors="pt", max_length=max_length, truncation=True)
+  # REMOVE EOS TOKEN IN THE PROMPT
+  inputs['input_ids'] = inputs['input_ids'][0][:-1].unsqueeze(dim=0)
+  inputs['attention_mask'] = inputs['attention_mask'][0][:-1].unsqueeze(dim=0)
 
-input_len = len(inputs['input_ids'].squeeze())
-        
-translated_text = tokenizer.decode(outputs[0][input_len:], skip_special_tokens=True)
-print(translated_text)
-```
+  outputs = model.generate(**inputs, max_length=max_length, eos_token_id=46332)
+
+  input_len = len(inputs['input_ids'].squeeze())
+          
+  translated_text = tokenizer.decode(outputs[0][input_len:], skip_special_tokens=True)
+  print(translated_text)
+  ```