Spaces:
Running
Running
Bram Vanroy
commited on
Commit
β’
f8b0e70
1
Parent(s):
55fbc57
add docstrings
Browse files
utils.py
CHANGED
@@ -1,3 +1,5 @@
|
|
|
|
|
|
1 |
import streamlit as st
|
2 |
|
3 |
from torch.quantization import quantize_dynamic
|
@@ -16,7 +18,15 @@ st_hash_funcs = {PreTrainedModel: lambda model: model.name_or_path,
|
|
16 |
|
17 |
|
18 |
@st.cache(show_spinner=False, hash_funcs=st_hash_funcs, allow_output_mutation=True)
|
19 |
-
def get_resources(multilingual: bool, quantize: bool = True):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
20 |
if multilingual:
|
21 |
# Tokenizer src_lang is reset during translation to the right language
|
22 |
tokenizer = AMRMBartTokenizer.from_pretrained("BramVanroy/mbart-en-es-nl-to-amr", src_lang="nl_XX")
|
@@ -37,11 +47,19 @@ def get_resources(multilingual: bool, quantize: bool = True):
|
|
37 |
|
38 |
|
39 |
@st.cache(show_spinner=False, hash_funcs=st_hash_funcs)
|
40 |
-
def translate(text: str, src_lang: str, model: MBartForConditionalGeneration, tokenizer: AMRMBartTokenizer, **gen_kwargs):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
41 |
tokenizer.src_lang = LANGUAGES[src_lang]
|
42 |
encoded = tokenizer(text, return_tensors="pt")
|
43 |
-
print(tokenizer.convert_ids_to_tokens(encoded.input_ids[0]))
|
44 |
-
print(model.name_or_path)
|
45 |
generated = model.generate(**encoded, **gen_kwargs)
|
46 |
return tokenizer.decode_and_fix(generated)[0]
|
47 |
|
|
|
1 |
+
from typing import Tuple
|
2 |
+
|
3 |
import streamlit as st
|
4 |
|
5 |
from torch.quantization import quantize_dynamic
|
|
|
18 |
|
19 |
|
20 |
@st.cache(show_spinner=False, hash_funcs=st_hash_funcs, allow_output_mutation=True)
|
21 |
+
def get_resources(multilingual: bool, quantize: bool = True) -> Tuple[MBartForConditionalGeneration, AMRMBartTokenizer, AMRLogitsProcessor]:
|
22 |
+
"""Get the relevant model, tokenizer and logits_processor. The loaded model depends on whether the multilingual
|
23 |
+
model is requested, or not. If not, an English-only model is loaded. The model can be optionally quantized
|
24 |
+
for better performance.
|
25 |
+
|
26 |
+
:param multilingual: whether or not to load the multilingual model. If not, loads the English-only model
|
27 |
+
:param quantize: whether to quantize the model with PyTorch's 'quantize_dynamic'
|
28 |
+
:return: the loaded model, tokenizer, and logits processor
|
29 |
+
"""
|
30 |
if multilingual:
|
31 |
# Tokenizer src_lang is reset during translation to the right language
|
32 |
tokenizer = AMRMBartTokenizer.from_pretrained("BramVanroy/mbart-en-es-nl-to-amr", src_lang="nl_XX")
|
|
|
47 |
|
48 |
|
49 |
@st.cache(show_spinner=False, hash_funcs=st_hash_funcs)
|
50 |
+
def translate(text: str, src_lang: str, model: MBartForConditionalGeneration, tokenizer: AMRMBartTokenizer, **gen_kwargs) -> str:
|
51 |
+
"""Translates a given text of a given source language with a given model and tokenizer. The generation is guided by
|
52 |
+
potential keyword-arguments, which can include arguments such as max length, logits processors, etc.
|
53 |
+
|
54 |
+
:param text: source text to translate
|
55 |
+
:param src_lang: source language
|
56 |
+
:param model: MBART model
|
57 |
+
:param tokenizer: MBART tokenizer
|
58 |
+
:param gen_kwargs: potential keyword arguments for the generation process
|
59 |
+
:return: the translation (linearized AMR graph)
|
60 |
+
"""
|
61 |
tokenizer.src_lang = LANGUAGES[src_lang]
|
62 |
encoded = tokenizer(text, return_tensors="pt")
|
|
|
|
|
63 |
generated = model.generate(**encoded, **gen_kwargs)
|
64 |
return tokenizer.decode_and_fix(generated)[0]
|
65 |
|