File size: 1,131 Bytes
98985f3
 
 
b5dfd52
98985f3
 
 
36d9f57
92761a6
98985f3
b5dfd52
39599fb
 
 
 
92761a6
60a0592
92761a6
98985f3
9937e65
b5dfd52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
from transformers import AutoTokenizer
import gradio as gr


gpt2_tokenizer = AutoTokenizer.from_pretrained("gpt2")
gpt_neox_tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
llama_tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/llama-tokenizer")
falcon_tokenizer = AutoTokenizer.from_pretrained("tiiuae/falcon-7b")
phi2_tokenizer = AutoTokenizer.from_pretrained("microsoft/phi-2")

def tokenize(input_text):
    gpt2_tokens = gpt2_tokenizer(input_text, add_special_tokens=True)["input_ids"]
    gpt_neox_tokens = gpt_neox_tokenizer(input_text, add_special_tokens=True)["input_ids"]
    llama_tokens = llama_tokenizer(input_text, add_special_tokens=True)["input_ids"]
    falcon_tokens = falcon_tokenizer(input_text, add_special_tokens=True)["input_ids"]
    phi2_tokens = phi2_tokenizer(input_text, add_special_tokens=True)["input_ids"]

    return f"GPT-2/GPT-J: {len(gpt2_tokens)}\nGPT-NeoX: {len(gpt_neox_tokens)}\nLLaMa: {len(llama_tokens)}\nFalcon: {len(falcon_tokens)}\nPhi-2: {len(phi2_tokens)}"

iface = gr.Interface(fn=tokenize, inputs=gr.Textbox(lines=7), outputs="text")
iface.launch()