Update README.md

README.md

@@ -42,7 +42,7 @@ import json
 Specify the base model and the adapter for LoRA fine-tuning. Replace <adapter_id> and <HF_TOKEN> with appropriate values.
 ``` python
 # Base model
-model_name = "kkkeee/llm-jp-3-13b-it"
+model_name = "kkkeee/llm-jp-3-13b-it15"
 
 # Hugging Face Token
 HF_TOKEN = "<your_hf_token>"  # Obtain token from https://huggingface.co/settings/tokens
@@ -65,7 +65,7 @@ FastLanguageModel.for_inference(model)
 ```
 # Step 4: Load Dataset
 Prepare your dataset in .jsonl format and upload it to your environment.
-
+```python
 # Load task data
 datasets = []
 with open("./elyza-tasks-100-TV_0.jsonl", "r") as f:
@@ -76,34 +76,33 @@ with open("./elyza-tasks-100-TV_0.jsonl", "r") as f:
         if item.endswith("}"):
             datasets.append(json.loads(item))
             item = ""
-
+```
 # Step 5: Perform Inference
 Set the model to inference mode and generate predictions.
+```python
+from tqdm import tqdm
 
-# Set model to inference mode
-FastLanguageModel.for_inference(model)
-
+# 推論
 results = []
-for dt in tqdm(datasets):
-    input = dt["input"]
-
-    prompt = f"""### 指示\n{input}\n### 回答\n"""
+for data in tqdm(datasets):
+  input = data["input"]
 
-    inputs = tokenizer([prompt], return_tensors="pt").to(model.device)
+  prompt = f"""### 指示\n{input}\n### 回答\n"""
 
-    outputs = model.generate(
-        **inputs, max_new_tokens=512, use_cache=True, do_sample=False, repetition_penalty=1.2
-    )
-    prediction = tokenizer.decode(outputs[0], skip_special_tokens=True).split('\n### 回答')[-1]
+  inputs = tokenizer([prompt], return_tensors = "pt").to(model.device)
 
-    results.append({"task_id": dt["task_id"], "input": input, "output": prediction})
+  outputs = model.generate(**inputs, max_new_tokens = 512, use_cache = True, do_sample=False, repetition_penalty=1.2)
+  output = tokenizer.decode(outputs[0], skip_special_tokens=True).split('\n### 回答')[-1]
 
+  results.append({"task_id": data["task_id"], "input": input, "output": output})
+```
 # Step 6: Save Results
 Save the inference results to a .jsonl file. Replace <adapter_id> with the appropriate identifier.
-
+```python
 # Save results to JSONL
 json_file_id = re.sub(".*/", "", adapter_id)
 with open(f"/content/{json_file_id}_output.jsonl", 'w', encoding='utf-8') as f:
     for result in results:
         json.dump(result, f, ensure_ascii=False)
-        f.write('\n')
+        f.write('\n')
+```