Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,13 +2,15 @@ from transformers import AutoTokenizer
|
|
| 2 |
import gradio as gr
|
| 3 |
|
| 4 |
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
|
|
|
|
|
|
| 12 |
|
| 13 |
def tokenize(input_text):
|
| 14 |
gpt2_tokens = gpt2_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
|
@@ -21,5 +23,9 @@ def tokenize(input_text):
|
|
| 21 |
|
| 22 |
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)}\nT5: {len(t5_tokens)}\nMistral: {len(mistral_tokens)}"
|
| 23 |
|
| 24 |
-
|
| 25 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
import gradio as gr
|
| 3 |
|
| 4 |
|
| 5 |
+
def load_tokenizers()
|
| 6 |
+
gpt2_tokenizer = AutoTokenizer.from_pretrained("gpt2")
|
| 7 |
+
gpt_neox_tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
|
| 8 |
+
llama_tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/llama-tokenizer")
|
| 9 |
+
falcon_tokenizer = AutoTokenizer.from_pretrained("tiiuae/falcon-7b")
|
| 10 |
+
phi2_tokenizer = AutoTokenizer.from_pretrained("microsoft/phi-2")
|
| 11 |
+
t5_tokenizer = AutoTokenizer.from_pretrained("google/flan-t5-xxl")
|
| 12 |
+
mistral_tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-v0.1")
|
| 13 |
+
|
| 14 |
|
| 15 |
def tokenize(input_text):
|
| 16 |
gpt2_tokens = gpt2_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
|
|
|
| 23 |
|
| 24 |
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)}\nT5: {len(t5_tokens)}\nMistral: {len(mistral_tokens)}"
|
| 25 |
|
| 26 |
+
|
| 27 |
+
if __name__ == "__main__":
|
| 28 |
+
load_tokenizers()
|
| 29 |
+
|
| 30 |
+
iface = gr.Interface(fn=tokenize, inputs=gr.Textbox(lines=7), outputs="text")
|
| 31 |
+
iface.launch()
|