Update app.py
Browse files
app.py
CHANGED
@@ -1,42 +1,4 @@
|
|
1 |
-
# -*- coding: utf-8 -*-
|
2 |
-
"""Kopie von Llama 2 Fine-Tuning using QLora
|
3 |
|
4 |
-
Automatically generated by Colaboratory.
|
5 |
-
|
6 |
-
Original file is located at
|
7 |
-
https://colab.research.google.com/drive/13dJqh-1y3KYGi5R82eqXGafkM5Y5k_ff
|
8 |
-
|
9 |
-
**Code Credit: Hugging Face**
|
10 |
-
|
11 |
-
**Dataset Credit: https://twitter.com/Dorialexander/status/1681671177696161794 **
|
12 |
-
|
13 |
-
## Finetune Llama-2-7b on a Google colab
|
14 |
-
|
15 |
-
Welcome to this Google Colab notebook that shows how to fine-tune the recent Llama-2-7b model on a single Google colab and turn it into a chatbot
|
16 |
-
|
17 |
-
We will leverage PEFT library from Hugging Face ecosystem, as well as QLoRA for more memory efficient finetuning
|
18 |
-
|
19 |
-
## Setup
|
20 |
-
|
21 |
-
Run the cells below to setup and install the required libraries. For our experiment we will need `accelerate`, `peft`, `transformers`, `datasets` and TRL to leverage the recent [`SFTTrainer`](https://huggingface.co/docs/trl/main/en/sft_trainer). We will use `bitsandbytes` to [quantize the base model into 4bit](https://huggingface.co/blog/4bit-transformers-bitsandbytes). We will also install `einops` as it is a requirement to load Falcon models.
|
22 |
-
"""
|
23 |
-
|
24 |
-
!pip install -q -U trl transformers accelerate git+https://github.com/huggingface/peft.git
|
25 |
-
!pip install -q datasets bitsandbytes einops wandb
|
26 |
-
|
27 |
-
"""## Dataset
|
28 |
-
|
29 |
-
|
30 |
-
"""
|
31 |
-
|
32 |
-
from datasets import load_dataset
|
33 |
-
|
34 |
-
#dataset_name = "timdettmers/openassistant-guanaco" ###Human ,.,,,,,, ###Assistant
|
35 |
-
|
36 |
-
dataset_name = 'AlexanderDoria/novel17_test' #french novels
|
37 |
-
dataset = load_dataset(dataset_name, split="train")
|
38 |
-
|
39 |
-
"""## Loading the model"""
|
40 |
|
41 |
import torch
|
42 |
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, AutoTokenizer
|
|
|
|
|
|
|
1 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
3 |
import torch
|
4 |
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, AutoTokenizer
|