Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -12,6 +12,7 @@ def tokenize(input_text):
|
|
12 |
gpt_neox_tokens = gpt_neox_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
13 |
falcon_tokens = falcon_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
14 |
phi2_tokens = phi2_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
|
|
15 |
t5_tokens = t5_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
16 |
gemma_tokens = gemma_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
17 |
command_r_tokens = command_r_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
@@ -27,6 +28,7 @@ def tokenize(input_text):
|
|
27 |
"GPT-NeoX": gpt_neox_tokens,
|
28 |
"Falcon": falcon_tokens,
|
29 |
"Phi-1/Phi-2": phi2_tokens,
|
|
|
30 |
"T5": t5_tokens,
|
31 |
"Gemma": gemma_tokens,
|
32 |
"Command-R": command_r_tokens,
|
@@ -62,6 +64,9 @@ if __name__ == "__main__":
|
|
62 |
phi2_tokenizer = AutoTokenizer.from_pretrained(
|
63 |
"microsoft/phi-2"
|
64 |
)
|
|
|
|
|
|
|
65 |
t5_tokenizer = AutoTokenizer.from_pretrained(
|
66 |
"google/flan-t5-xxl"
|
67 |
)
|
|
|
12 |
gpt_neox_tokens = gpt_neox_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
13 |
falcon_tokens = falcon_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
14 |
phi2_tokens = phi2_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
15 |
+
phi3_tokens = phi3_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
16 |
t5_tokens = t5_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
17 |
gemma_tokens = gemma_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
18 |
command_r_tokens = command_r_tokenizer(input_text, add_special_tokens=True)["input_ids"]
|
|
|
28 |
"GPT-NeoX": gpt_neox_tokens,
|
29 |
"Falcon": falcon_tokens,
|
30 |
"Phi-1/Phi-2": phi2_tokens,
|
31 |
+
"Phi-3": phi3_tokens,
|
32 |
"T5": t5_tokens,
|
33 |
"Gemma": gemma_tokens,
|
34 |
"Command-R": command_r_tokens,
|
|
|
64 |
phi2_tokenizer = AutoTokenizer.from_pretrained(
|
65 |
"microsoft/phi-2"
|
66 |
)
|
67 |
+
phi3_tokenizer = AutoTokenizer.from_pretrained(
|
68 |
+
"microsoft/Phi-3-mini-4k-instruct"
|
69 |
+
)
|
70 |
t5_tokenizer = AutoTokenizer.from_pretrained(
|
71 |
"google/flan-t5-xxl"
|
72 |
)
|