Spaces:
Running
Running
Vineel Pratap
commited on
Commit
•
a4107b1
1
Parent(s):
9f2bd1d
update_model
Browse files- app.py +48 -7
- style.css +6 -0
- upload/mms_zs/config.json +108 -0
- upload/mms_zs/model.safetensors +3 -0
- upload/mms_zs/preprocessor_config.json +10 -0
- upload/mms_zs/special_tokens_map.json +6 -0
- upload/mms_zs/tokenizer_config.json +48 -0
- upload/mms_zs/tokens.txt +32 -0
- upload/mms_zs/vocab.json +34 -0
- zeroshot.py +51 -46
app.py
CHANGED
@@ -1,24 +1,65 @@
|
|
1 |
import gradio as gr
|
2 |
from zeroshot import process, ZS_EXAMPLES
|
3 |
|
4 |
-
with gr.Blocks() as demo:
|
5 |
-
gr.Markdown("")
|
6 |
gr.Markdown(
|
7 |
"<p align='center' style='font-size: 20px;'>MMS Zero-shot ASR Demo. See our arXiV <a href='https://arxiv.org/'>paper</a> for model details.</p>"
|
8 |
)
|
9 |
gr.HTML(
|
10 |
-
"""<center>The demo works on input audio in any language, as long as you provide a list of words for that language and an optional n-gram language model (even a simple 1-gram model will work!) to help with accuracy.</center>"""
|
11 |
)
|
12 |
with gr.Row():
|
13 |
with gr.Column():
|
14 |
audio = gr.Audio(label="Audio Input\n(use microphone or upload a file)")
|
|
|
15 |
with gr.Row():
|
16 |
-
words_file = gr.File(label="
|
17 |
lm_file = gr.File(label="Language Model\n(optional)")
|
18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
with gr.Column():
|
20 |
text = gr.Textbox(label="Transcript")
|
21 |
-
btn.click(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
examples = gr.Examples(examples=ZS_EXAMPLES, inputs=[audio, words_file])
|
23 |
|
24 |
-
demo.launch(
|
|
|
1 |
import gradio as gr
|
2 |
from zeroshot import process, ZS_EXAMPLES
|
3 |
|
4 |
+
with gr.Blocks(css="style.css") as demo:
|
|
|
5 |
gr.Markdown(
|
6 |
"<p align='center' style='font-size: 20px;'>MMS Zero-shot ASR Demo. See our arXiV <a href='https://arxiv.org/'>paper</a> for model details.</p>"
|
7 |
)
|
8 |
gr.HTML(
|
9 |
+
"""<center>The demo works on input audio in any language, as long as you provide a list of words or sentences for that language and an optional n-gram language model (even a simple 1-gram model will work!) to help with accuracy.<br>We recommend having a minimum of 5000 distinct words in the textfile to acheive a good performance.</center>"""
|
10 |
)
|
11 |
with gr.Row():
|
12 |
with gr.Column():
|
13 |
audio = gr.Audio(label="Audio Input\n(use microphone or upload a file)")
|
14 |
+
|
15 |
with gr.Row():
|
16 |
+
words_file = gr.File(label="Text Data")
|
17 |
lm_file = gr.File(label="Language Model\n(optional)")
|
18 |
+
|
19 |
+
with gr.Accordion("Advanced Settings", open=False):
|
20 |
+
gr.Markdown(
|
21 |
+
"The following parameters are used for beam-search decoding. Use the default values if you are not sure."
|
22 |
+
)
|
23 |
+
with gr.Row():
|
24 |
+
wscore = gr.Slider(
|
25 |
+
minimum=-10.0,
|
26 |
+
maximum=10.0,
|
27 |
+
value=0,
|
28 |
+
step=0.1,
|
29 |
+
interactive=True,
|
30 |
+
label="Word Insertion Score",
|
31 |
+
)
|
32 |
+
lmscore = gr.Slider(
|
33 |
+
minimum=-10.0,
|
34 |
+
maximum=10.0,
|
35 |
+
value=0,
|
36 |
+
step=0.1,
|
37 |
+
interactive=True,
|
38 |
+
label="Language Model Score",
|
39 |
+
)
|
40 |
+
with gr.Row():
|
41 |
+
wscore_usedefault = gr.Checkbox(
|
42 |
+
label="Use Default Word Insertion Score", value=True
|
43 |
+
)
|
44 |
+
lmscore_usedefault = gr.Checkbox(
|
45 |
+
label="Use Default Language Model Score", value=True
|
46 |
+
)
|
47 |
+
btn = gr.Button("Submit", elem_id="submit")
|
48 |
with gr.Column():
|
49 |
text = gr.Textbox(label="Transcript")
|
50 |
+
btn.click(
|
51 |
+
process,
|
52 |
+
inputs=[
|
53 |
+
audio,
|
54 |
+
words_file,
|
55 |
+
lm_file,
|
56 |
+
wscore,
|
57 |
+
lmscore,
|
58 |
+
wscore_usedefault,
|
59 |
+
lmscore_usedefault,
|
60 |
+
],
|
61 |
+
outputs=text,
|
62 |
+
)
|
63 |
examples = gr.Examples(examples=ZS_EXAMPLES, inputs=[audio, words_file])
|
64 |
|
65 |
+
demo.launch()
|
style.css
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#submit {
|
2 |
+
margin: auto;
|
3 |
+
color: #fff;
|
4 |
+
background: #1565c0;
|
5 |
+
border-radius: 100vh;
|
6 |
+
}
|
upload/mms_zs/config.json
ADDED
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"activation_dropout": 0.0,
|
3 |
+
"adapter_attn_dim": null,
|
4 |
+
"adapter_kernel_size": 3,
|
5 |
+
"adapter_stride": 2,
|
6 |
+
"add_adapter": false,
|
7 |
+
"apply_spec_augment": true,
|
8 |
+
"architectures": [
|
9 |
+
"Wav2Vec2ForCTC"
|
10 |
+
],
|
11 |
+
"attention_dropout": 0.1,
|
12 |
+
"bos_token_id": 1,
|
13 |
+
"classifier_proj_size": 256,
|
14 |
+
"codevector_dim": 768,
|
15 |
+
"contrastive_logits_temperature": 0.1,
|
16 |
+
"conv_bias": true,
|
17 |
+
"conv_dim": [
|
18 |
+
512,
|
19 |
+
512,
|
20 |
+
512,
|
21 |
+
512,
|
22 |
+
512,
|
23 |
+
512,
|
24 |
+
512
|
25 |
+
],
|
26 |
+
"conv_kernel": [
|
27 |
+
10,
|
28 |
+
3,
|
29 |
+
3,
|
30 |
+
3,
|
31 |
+
3,
|
32 |
+
2,
|
33 |
+
2
|
34 |
+
],
|
35 |
+
"conv_stride": [
|
36 |
+
5,
|
37 |
+
2,
|
38 |
+
2,
|
39 |
+
2,
|
40 |
+
2,
|
41 |
+
2,
|
42 |
+
2
|
43 |
+
],
|
44 |
+
"ctc_loss_reduction": "sum",
|
45 |
+
"ctc_zero_infinity": false,
|
46 |
+
"diversity_loss_weight": 0.1,
|
47 |
+
"do_stable_layer_norm": true,
|
48 |
+
"eos_token_id": 2,
|
49 |
+
"feat_extract_activation": "gelu",
|
50 |
+
"feat_extract_dropout": 0.0,
|
51 |
+
"feat_extract_norm": "layer",
|
52 |
+
"feat_proj_dropout": 0.1,
|
53 |
+
"feat_quantizer_dropout": 0.0,
|
54 |
+
"final_dropout": 0.0,
|
55 |
+
"gradient_checkpointing": false,
|
56 |
+
"hidden_act": "gelu",
|
57 |
+
"hidden_dropout": 0.1,
|
58 |
+
"hidden_size": 1024,
|
59 |
+
"initializer_range": 0.02,
|
60 |
+
"intermediate_size": 4096,
|
61 |
+
"layer_norm_eps": 1e-05,
|
62 |
+
"layerdrop": 0.1,
|
63 |
+
"mask_feature_length": 10,
|
64 |
+
"mask_feature_min_masks": 0,
|
65 |
+
"mask_feature_prob": 0.0,
|
66 |
+
"mask_time_length": 10,
|
67 |
+
"mask_time_min_masks": 2,
|
68 |
+
"mask_time_prob": 0.075,
|
69 |
+
"model_type": "wav2vec2",
|
70 |
+
"num_adapter_layers": 3,
|
71 |
+
"num_attention_heads": 16,
|
72 |
+
"num_codevector_groups": 2,
|
73 |
+
"num_codevectors_per_group": 320,
|
74 |
+
"num_conv_pos_embedding_groups": 16,
|
75 |
+
"num_conv_pos_embeddings": 128,
|
76 |
+
"num_feat_extract_layers": 7,
|
77 |
+
"num_hidden_layers": 24,
|
78 |
+
"num_negatives": 100,
|
79 |
+
"output_hidden_size": 1024,
|
80 |
+
"pad_token_id": 0,
|
81 |
+
"proj_codevector_dim": 768,
|
82 |
+
"tdnn_dilation": [
|
83 |
+
1,
|
84 |
+
2,
|
85 |
+
3,
|
86 |
+
1,
|
87 |
+
1
|
88 |
+
],
|
89 |
+
"tdnn_dim": [
|
90 |
+
512,
|
91 |
+
512,
|
92 |
+
512,
|
93 |
+
512,
|
94 |
+
1500
|
95 |
+
],
|
96 |
+
"tdnn_kernel": [
|
97 |
+
5,
|
98 |
+
3,
|
99 |
+
3,
|
100 |
+
1,
|
101 |
+
1
|
102 |
+
],
|
103 |
+
"torch_dtype": "float32",
|
104 |
+
"transformers_version": "4.42.1",
|
105 |
+
"use_weighted_layer_sum": false,
|
106 |
+
"vocab_size": 32,
|
107 |
+
"xvector_output_dim": 512
|
108 |
+
}
|
upload/mms_zs/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:39baa2c87b9abd9910c1982bf82aabda3dbe3ba615e20d5ee0be1026975dcb8c
|
3 |
+
size 1261938632
|
upload/mms_zs/preprocessor_config.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"do_normalize": true,
|
3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
4 |
+
"feature_size": 1,
|
5 |
+
"padding_side": "right",
|
6 |
+
"padding_value": 0,
|
7 |
+
"processor_class": "Wav2Vec2Processor",
|
8 |
+
"return_attention_mask": true,
|
9 |
+
"sampling_rate": 16000
|
10 |
+
}
|
upload/mms_zs/special_tokens_map.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"eos_token": "</s>",
|
4 |
+
"pad_token": "<pad>",
|
5 |
+
"unk_token": "<unk>"
|
6 |
+
}
|
upload/mms_zs/tokenizer_config.json
ADDED
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "<pad>",
|
5 |
+
"lstrip": true,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": true,
|
8 |
+
"single_word": false,
|
9 |
+
"special": false
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "<s>",
|
13 |
+
"lstrip": true,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": true,
|
16 |
+
"single_word": false,
|
17 |
+
"special": false
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "</s>",
|
21 |
+
"lstrip": true,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": true,
|
24 |
+
"single_word": false,
|
25 |
+
"special": false
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "<unk>",
|
29 |
+
"lstrip": true,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": true,
|
32 |
+
"single_word": false,
|
33 |
+
"special": false
|
34 |
+
}
|
35 |
+
},
|
36 |
+
"bos_token": "<s>",
|
37 |
+
"clean_up_tokenization_spaces": true,
|
38 |
+
"do_lower_case": false,
|
39 |
+
"eos_token": "</s>",
|
40 |
+
"model_max_length": 1000000000000000019884624838656,
|
41 |
+
"pad_token": "<pad>",
|
42 |
+
"processor_class": "Wav2Vec2Processor",
|
43 |
+
"replace_word_delimiter_char": " ",
|
44 |
+
"target_lang": null,
|
45 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
46 |
+
"unk_token": "<unk>",
|
47 |
+
"word_delimiter_token": "|"
|
48 |
+
}
|
upload/mms_zs/tokens.txt
ADDED
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
<s>
|
2 |
+
<pad>
|
3 |
+
</s>
|
4 |
+
<unk>
|
5 |
+
|
|
6 |
+
a
|
7 |
+
i
|
8 |
+
e
|
9 |
+
n
|
10 |
+
o
|
11 |
+
u
|
12 |
+
t
|
13 |
+
k
|
14 |
+
m
|
15 |
+
s
|
16 |
+
r
|
17 |
+
l
|
18 |
+
h
|
19 |
+
g
|
20 |
+
d
|
21 |
+
y
|
22 |
+
b
|
23 |
+
p
|
24 |
+
c
|
25 |
+
w
|
26 |
+
j
|
27 |
+
'
|
28 |
+
v
|
29 |
+
z
|
30 |
+
f
|
31 |
+
q
|
32 |
+
x
|
upload/mms_zs/vocab.json
ADDED
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"'": 26,
|
3 |
+
"</s>": 2,
|
4 |
+
"<pad>": 0,
|
5 |
+
"<s>": 1,
|
6 |
+
"<unk>": 3,
|
7 |
+
"a": 5,
|
8 |
+
"b": 21,
|
9 |
+
"c": 23,
|
10 |
+
"d": 19,
|
11 |
+
"e": 7,
|
12 |
+
"f": 29,
|
13 |
+
"g": 18,
|
14 |
+
"h": 17,
|
15 |
+
"i": 6,
|
16 |
+
"j": 25,
|
17 |
+
"k": 12,
|
18 |
+
"l": 16,
|
19 |
+
"m": 13,
|
20 |
+
"n": 8,
|
21 |
+
"o": 9,
|
22 |
+
"p": 22,
|
23 |
+
"q": 30,
|
24 |
+
"r": 15,
|
25 |
+
"s": 14,
|
26 |
+
"t": 11,
|
27 |
+
"u": 10,
|
28 |
+
"v": 27,
|
29 |
+
"w": 24,
|
30 |
+
"x": 31,
|
31 |
+
"y": 20,
|
32 |
+
"z": 28,
|
33 |
+
"|": 4
|
34 |
+
}
|
zeroshot.py
CHANGED
@@ -16,34 +16,17 @@ UROMAN_PL = os.path.join(uroman_dir, "bin", "uroman.pl")
|
|
16 |
|
17 |
ASR_SAMPLING_RATE = 16_000
|
18 |
|
19 |
-
|
|
|
|
|
|
|
|
|
20 |
|
21 |
processor = AutoProcessor.from_pretrained(MODEL_ID)
|
22 |
model = Wav2Vec2ForCTC.from_pretrained(MODEL_ID)
|
23 |
|
24 |
-
|
25 |
-
lm_decoding_configfile = hf_hub_download(
|
26 |
-
repo_id="facebook/mms-cclms",
|
27 |
-
filename="decoding_config.json",
|
28 |
-
subfolder="mms-1b-all",
|
29 |
-
)
|
30 |
-
|
31 |
-
with open(lm_decoding_configfile) as f:
|
32 |
-
lm_decoding_config = json.loads(f.read())
|
33 |
-
|
34 |
-
decoding_config = lm_decoding_config["eng"]
|
35 |
-
|
36 |
-
lm_file = hf_hub_download(
|
37 |
-
repo_id="facebook/mms-cclms",
|
38 |
-
filename=decoding_config["lmfile"].rsplit("/", 1)[1],
|
39 |
-
subfolder=decoding_config["lmfile"].rsplit("/", 1)[0],
|
40 |
-
)
|
41 |
|
42 |
-
token_file = hf_hub_download(
|
43 |
-
repo_id="facebook/mms-cclms",
|
44 |
-
filename=decoding_config["tokensfile"].rsplit("/", 1)[1],
|
45 |
-
subfolder=decoding_config["tokensfile"].rsplit("/", 1)[0],
|
46 |
-
)
|
47 |
|
48 |
def error_check_file(filepath):
|
49 |
if not isinstance(filepath, str):
|
@@ -53,13 +36,15 @@ def error_check_file(filepath):
|
|
53 |
if not os.path.exists(filepath):
|
54 |
return "Input file '{}' doesn't exists".format(type(filepath))
|
55 |
|
|
|
56 |
def norm_uroman(text):
|
57 |
text = text.lower()
|
58 |
text = text.replace("’", "'")
|
59 |
text = re.sub("([^a-z' ])", " ", text)
|
60 |
-
text = re.sub(
|
61 |
return text.strip()
|
62 |
|
|
|
63 |
def uromanize(words):
|
64 |
iso = "xxx"
|
65 |
with tempfile.NamedTemporaryFile() as tf, tempfile.NamedTemporaryFile() as tf2:
|
@@ -72,24 +57,35 @@ def uromanize(words):
|
|
72 |
lexicon = {}
|
73 |
with open(tf2.name) as f:
|
74 |
for idx, line in enumerate(f):
|
|
|
|
|
75 |
line = re.sub(r"\s+", " ", norm_uroman(line)).strip()
|
76 |
lexicon[words[idx]] = " ".join(line) + " |"
|
77 |
return lexicon
|
78 |
|
79 |
|
80 |
def load_lexicon(filepath):
|
81 |
-
words =
|
82 |
with open(filepath) as f:
|
83 |
for line in f:
|
84 |
line = line.strip()
|
85 |
# ignore invalid words.
|
86 |
if not line or " " in line or len(line) > 50:
|
87 |
continue
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
93 |
if isinstance(audio_data, tuple):
|
94 |
# microphone
|
95 |
sr, audio_samples = audio_data
|
@@ -101,17 +97,18 @@ def process(audio_data, words_file, lm_path=None):
|
|
101 |
audio_samples = librosa.load(audio_data, sr=ASR_SAMPLING_RATE, mono=True)[0]
|
102 |
# print(audio_samples[:10])
|
103 |
# print("I'm here 102")
|
104 |
-
|
105 |
lang_code = "eng"
|
106 |
-
processor.tokenizer.set_target_lang(lang_code)
|
107 |
# print("I'm here 107")
|
108 |
-
model.load_adapter(lang_code)
|
109 |
# print("I'm here 109")
|
110 |
inputs = processor(
|
111 |
audio_samples, sampling_rate=ASR_SAMPLING_RATE, return_tensors="pt"
|
112 |
)
|
113 |
# print("I'm here 106")
|
114 |
-
|
|
|
115 |
# set device
|
116 |
if torch.cuda.is_available():
|
117 |
device = torch.device("cuda")
|
@@ -123,27 +120,37 @@ def process(audio_data, words_file, lm_path=None):
|
|
123 |
device = torch.device("mps")
|
124 |
else:
|
125 |
device = torch.device("cpu")
|
126 |
-
|
127 |
model.to(device)
|
128 |
inputs = inputs.to(device)
|
129 |
# print("I'm here 122")
|
130 |
with torch.no_grad():
|
131 |
outputs = model(**inputs).logits
|
132 |
|
133 |
-
# Setup lexicon and decoder
|
134 |
# print("before uroman")
|
135 |
lexicon = load_lexicon(words_file)
|
136 |
# print("after uroman")
|
137 |
# print("len lexicon", len(lexicon))
|
138 |
with tempfile.NamedTemporaryFile() as lexicon_file:
|
139 |
-
|
140 |
with open(lexicon_file.name, "w") as f:
|
141 |
idx = 10
|
142 |
for word, spelling in lexicon.items():
|
143 |
f.write(word + " " + spelling + "\n")
|
144 |
-
if idx%100 == 0:
|
145 |
print(word, spelling, flush=True)
|
146 |
-
idx+=1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
147 |
beam_search_decoder = ctc_decoder(
|
148 |
lexicon=lexicon_file.name,
|
149 |
tokens=token_file,
|
@@ -151,9 +158,9 @@ def process(audio_data, words_file, lm_path=None):
|
|
151 |
nbest=1,
|
152 |
beam_size=500,
|
153 |
beam_size_token=50,
|
154 |
-
lm_weight=
|
155 |
-
word_score=
|
156 |
-
sil_score=
|
157 |
blank_token="<s>",
|
158 |
)
|
159 |
|
@@ -163,8 +170,6 @@ def process(audio_data, words_file, lm_path=None):
|
|
163 |
return transcription
|
164 |
|
165 |
|
166 |
-
ZS_EXAMPLES = [
|
167 |
-
["upload/english.mp3", "upload/words_top10k.txt"]
|
168 |
-
]
|
169 |
|
170 |
-
|
|
|
16 |
|
17 |
ASR_SAMPLING_RATE = 16_000
|
18 |
|
19 |
+
WORD_SCORE_DEAULT_IF_LM = -0.18
|
20 |
+
WORD_SCORE_DEAULT_IF_NOLM = -3.5
|
21 |
+
LM_SCORE_DEAULT = 1.48
|
22 |
+
|
23 |
+
MODEL_ID = "upload/mms_zs"
|
24 |
|
25 |
processor = AutoProcessor.from_pretrained(MODEL_ID)
|
26 |
model = Wav2Vec2ForCTC.from_pretrained(MODEL_ID)
|
27 |
|
28 |
+
token_file = "upload/mms_zs/tokens.txt"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
|
|
|
|
|
|
|
|
|
|
|
30 |
|
31 |
def error_check_file(filepath):
|
32 |
if not isinstance(filepath, str):
|
|
|
36 |
if not os.path.exists(filepath):
|
37 |
return "Input file '{}' doesn't exists".format(type(filepath))
|
38 |
|
39 |
+
|
40 |
def norm_uroman(text):
|
41 |
text = text.lower()
|
42 |
text = text.replace("’", "'")
|
43 |
text = re.sub("([^a-z' ])", " ", text)
|
44 |
+
text = re.sub(" +", " ", text)
|
45 |
return text.strip()
|
46 |
|
47 |
+
|
48 |
def uromanize(words):
|
49 |
iso = "xxx"
|
50 |
with tempfile.NamedTemporaryFile() as tf, tempfile.NamedTemporaryFile() as tf2:
|
|
|
57 |
lexicon = {}
|
58 |
with open(tf2.name) as f:
|
59 |
for idx, line in enumerate(f):
|
60 |
+
if not line.strip():
|
61 |
+
continue
|
62 |
line = re.sub(r"\s+", " ", norm_uroman(line)).strip()
|
63 |
lexicon[words[idx]] = " ".join(line) + " |"
|
64 |
return lexicon
|
65 |
|
66 |
|
67 |
def load_lexicon(filepath):
|
68 |
+
words = {}
|
69 |
with open(filepath) as f:
|
70 |
for line in f:
|
71 |
line = line.strip()
|
72 |
# ignore invalid words.
|
73 |
if not line or " " in line or len(line) > 50:
|
74 |
continue
|
75 |
+
for w in line.split():
|
76 |
+
words[w.lower()] = True
|
77 |
+
return uromanize(list(words.keys()))
|
78 |
+
|
79 |
+
|
80 |
+
def process(
|
81 |
+
audio_data,
|
82 |
+
words_file,
|
83 |
+
lm_path=None,
|
84 |
+
wscore=None,
|
85 |
+
lmscore=None,
|
86 |
+
wscore_usedefault=True,
|
87 |
+
lmscore_usedefault=True,
|
88 |
+
):
|
89 |
if isinstance(audio_data, tuple):
|
90 |
# microphone
|
91 |
sr, audio_samples = audio_data
|
|
|
97 |
audio_samples = librosa.load(audio_data, sr=ASR_SAMPLING_RATE, mono=True)[0]
|
98 |
# print(audio_samples[:10])
|
99 |
# print("I'm here 102")
|
100 |
+
print("len audio_samples", len(audio_samples))
|
101 |
lang_code = "eng"
|
102 |
+
# processor.tokenizer.set_target_lang(lang_code)
|
103 |
# print("I'm here 107")
|
104 |
+
# model.load_adapter(lang_code)
|
105 |
# print("I'm here 109")
|
106 |
inputs = processor(
|
107 |
audio_samples, sampling_rate=ASR_SAMPLING_RATE, return_tensors="pt"
|
108 |
)
|
109 |
# print("I'm here 106")
|
110 |
+
print("inputs type", type(inputs))
|
111 |
+
# print("inputs size", inputs.size)
|
112 |
# set device
|
113 |
if torch.cuda.is_available():
|
114 |
device = torch.device("cuda")
|
|
|
120 |
device = torch.device("mps")
|
121 |
else:
|
122 |
device = torch.device("cpu")
|
123 |
+
device = torch.device("cpu")
|
124 |
model.to(device)
|
125 |
inputs = inputs.to(device)
|
126 |
# print("I'm here 122")
|
127 |
with torch.no_grad():
|
128 |
outputs = model(**inputs).logits
|
129 |
|
130 |
+
# Setup lexicon and decoder
|
131 |
# print("before uroman")
|
132 |
lexicon = load_lexicon(words_file)
|
133 |
# print("after uroman")
|
134 |
# print("len lexicon", len(lexicon))
|
135 |
with tempfile.NamedTemporaryFile() as lexicon_file:
|
136 |
+
|
137 |
with open(lexicon_file.name, "w") as f:
|
138 |
idx = 10
|
139 |
for word, spelling in lexicon.items():
|
140 |
f.write(word + " " + spelling + "\n")
|
141 |
+
if idx % 100 == 0:
|
142 |
print(word, spelling, flush=True)
|
143 |
+
idx += 1
|
144 |
+
|
145 |
+
if wscore_usedefault:
|
146 |
+
wscore = (
|
147 |
+
WORD_SCORE_DEAULT_IF_LM
|
148 |
+
if lm_path is not None
|
149 |
+
else WORD_SCORE_DEAULT_IF_NOLM
|
150 |
+
)
|
151 |
+
if lmscore_usedefault:
|
152 |
+
lmscore = LM_SCORE_DEAULT if lm_path is not None else 0
|
153 |
+
|
154 |
beam_search_decoder = ctc_decoder(
|
155 |
lexicon=lexicon_file.name,
|
156 |
tokens=token_file,
|
|
|
158 |
nbest=1,
|
159 |
beam_size=500,
|
160 |
beam_size_token=50,
|
161 |
+
lm_weight=lmscore,
|
162 |
+
word_score=wscore,
|
163 |
+
sil_score=0,
|
164 |
blank_token="<s>",
|
165 |
)
|
166 |
|
|
|
170 |
return transcription
|
171 |
|
172 |
|
173 |
+
ZS_EXAMPLES = [["upload/english.mp3", "upload/words_top10k.txt"]]
|
|
|
|
|
174 |
|
175 |
+
print(process("upload/english.mp3", "upload/words_top10k.txt"))
|