From 74ecbcce5e1fc3877af2fdbdcfd4a70b76587721 Mon Sep 17 00:00:00 2001
From: Xiang Long <mindsculptor@yeah.net>
Date: Wed, 6 Mar 2024 17:25:41 +0800
Subject: [PATCH] Fix sft_dataset issue and naming error

---
 finetune/finetune.py         | 9 ++++-----
 finetune/lora_finetune.ipynb | 2 +-
 2 files changed, 5 insertions(+), 6 deletions(-)

diff --git a/finetune/finetune.py b/finetune/finetune.py
index 9b0ce1b..1bf3d4e 100644
--- a/finetune/finetune.py
+++ b/finetune/finetune.py
@@ -1,13 +1,13 @@
 # -*- coding: utf-8 -*-
 import json
-from typing import Dict, Optional
 from dataclasses import dataclass, field
+from typing import Dict, Optional
 
 import torch
-from torch.utils.data import Dataset
-
 import transformers
-from transformers import AutoTokenizer, AutoModelForCausalLM, TrainingArguments, Trainer
+from torch.utils.data import Dataset
+from transformers import (AutoModelForCausalLM, AutoTokenizer, Trainer,
+                          TrainingArguments)
 
 
 @dataclass
@@ -90,7 +90,6 @@ class SupervisedDataset(Dataset):
                 label_ids += (
                     [self.ignore_index] * len(self.assistant_tokens)
                     + content_ids
-                    + [self.tokenizer.eos_token_id]
                 )
 
         input_ids = input_ids[: self.model_max_length]
diff --git a/finetune/lora_finetune.ipynb b/finetune/lora_finetune.ipynb
index 6cf5b41..35328be 100644
--- a/finetune/lora_finetune.ipynb
+++ b/finetune/lora_finetune.ipynb
@@ -101,7 +101,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!bash lora_finetune_ds.sh"
+    "!bash lora_finetune.sh"
    ]
   }
  ],