Upload 10 files
Browse files- .gitattributes +1 -0
- checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=19.ckpt +3 -0
- checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=20-last.ckpt +3 -0
- checkpoints/ASR-Model-Language-ta--val_wer=0.3864-epoch=9.ckpt +3 -0
- checkpoints/ASR-Model-Language-ta.nemo +3 -0
- cmd-args.log +1 -0
- events.out.tfevents.1708036461.0e5e24679ce6.2432.0 +3 -0
- hparams.yaml +620 -0
- lightning_logs.txt +24 -0
- nemo_error_log.txt +42 -0
- nemo_log_globalrank-0_localrank-0.txt +0 -0
.gitattributes
CHANGED
@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
Model-ta.nemo filter=lfs diff=lfs merge=lfs -text
|
|
|
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
Model-ta.nemo filter=lfs diff=lfs merge=lfs -text
|
37 |
+
checkpoints/ASR-Model-Language-ta.nemo filter=lfs diff=lfs merge=lfs -text
|
checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=19.ckpt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e3bd3f517957625e4f6215303565cbad8d61b4047303d5b870498dbc32f47a68
|
3 |
+
size 154197962
|
checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=20-last.ckpt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a3ef97644b2b931b104bab71fb1d952eeb118bedb21c48535e9c2749465ac325
|
3 |
+
size 154197962
|
checkpoints/ASR-Model-Language-ta--val_wer=0.3864-epoch=9.ckpt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:564d0ebe8855b2ac1190654db705ac315cce0d85e2cc3c2e1d3f8926786a78a6
|
3 |
+
size 154197387
|
checkpoints/ASR-Model-Language-ta.nemo
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:14a7ebfb3d783169838f404029db671f15e23cb086d7ee070956cf361ad377d1
|
3 |
+
size 147036160
|
cmd-args.log
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
/usr/local/lib/python3.10/dist-packages/colab_kernel_launcher.py -f /root/.local/share/jupyter/runtime/kernel-af07951b-f896-4671-afe4-af552f6f2ecd.json
|
events.out.tfevents.1708036461.0e5e24679ce6.2432.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:610468695b36c3a80ee06f6809baaed1d2e65256a37bd21d4d52de7234482c22
|
3 |
+
size 800584
|
hparams.yaml
ADDED
@@ -0,0 +1,620 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
cfg:
|
2 |
+
sample_rate: 16000
|
3 |
+
train_ds:
|
4 |
+
manifest_filepath: /content/datasets/ta/yaygomii/Tamil-Speech-Dialect-Corpus-Shuffled-Split/train/train_yaygomii_Tamil-Speech-Dialect-Corpus-Shuffled-Split_manifest_processed.json,/content/datasets/ta/yaygomii/Tamil-Speech-Dialect-Corpus-Shuffled-Split/valid/valid_yaygomii_Tamil-Speech-Dialect-Corpus-Shuffled-Split_manifest_processed.json
|
5 |
+
sample_rate: 16000
|
6 |
+
batch_size: 16
|
7 |
+
trim_silence: true
|
8 |
+
max_duration: 16.7
|
9 |
+
shuffle: true
|
10 |
+
is_tarred: false
|
11 |
+
tarred_audio_filepaths: null
|
12 |
+
num_workers: 8
|
13 |
+
pin_memory: true
|
14 |
+
use_start_end_token: true
|
15 |
+
validation_ds:
|
16 |
+
manifest_filepath: /content/datasets/ta/yaygomii/Tamil-Speech-Dialect-Corpus-Shuffled-Split/test/test_yaygomii_Tamil-Speech-Dialect-Corpus-Shuffled-Split_manifest_processed.json
|
17 |
+
sample_rate: 16000
|
18 |
+
batch_size: 8
|
19 |
+
shuffle: false
|
20 |
+
num_workers: 8
|
21 |
+
pin_memory: true
|
22 |
+
use_start_end_token: true
|
23 |
+
trim_silence: true
|
24 |
+
test_ds:
|
25 |
+
manifest_filepath: /content/datasets/ta/yaygomii/Tamil-Speech-Dialect-Corpus-Shuffled-Split/test/test_yaygomii_Tamil-Speech-Dialect-Corpus-Shuffled-Split_manifest_processed.json
|
26 |
+
sample_rate: 16000
|
27 |
+
batch_size: 8
|
28 |
+
shuffle: false
|
29 |
+
num_workers: 8
|
30 |
+
pin_memory: true
|
31 |
+
use_start_end_token: true
|
32 |
+
trim_silence: true
|
33 |
+
model_defaults:
|
34 |
+
repeat: 5
|
35 |
+
dropout: 0.0
|
36 |
+
separable: true
|
37 |
+
se: true
|
38 |
+
se_context_size: -1
|
39 |
+
tokenizer:
|
40 |
+
dir: tokenizers/ta/tokenizer_spe_bpe_v174/
|
41 |
+
type: bpe
|
42 |
+
model_path: nemo:288fd8b4d3c64e75ac7a6b32b0f3586d_tokenizer.model
|
43 |
+
vocab_path: nemo:6698e9428f25429e8ab2c5238438d52f_vocab.txt
|
44 |
+
spe_tokenizer_vocab: nemo:344d9da9a52049caad7742ae1d994d19_tokenizer.vocab
|
45 |
+
preprocessor:
|
46 |
+
_target_: nemo.collections.asr.modules.AudioToMelSpectrogramPreprocessor
|
47 |
+
sample_rate: 16000
|
48 |
+
normalize: per_feature
|
49 |
+
window_size: 0.025
|
50 |
+
window_stride: 0.01
|
51 |
+
window: hann
|
52 |
+
features: 80
|
53 |
+
n_fft: 512
|
54 |
+
frame_splicing: 1
|
55 |
+
dither: 1.0e-05
|
56 |
+
pad_to: 16
|
57 |
+
stft_conv: false
|
58 |
+
spec_augment:
|
59 |
+
_target_: nemo.collections.asr.modules.SpectrogramAugmentation
|
60 |
+
freq_masks: 2
|
61 |
+
time_masks: 10
|
62 |
+
freq_width: 25
|
63 |
+
time_width: 0.05
|
64 |
+
encoder:
|
65 |
+
_target_: nemo.collections.asr.modules.ConvASREncoder
|
66 |
+
feat_in: 80
|
67 |
+
activation: relu
|
68 |
+
conv_mask: true
|
69 |
+
jasper:
|
70 |
+
- filters: 512
|
71 |
+
repeat: 1
|
72 |
+
kernel:
|
73 |
+
- 5
|
74 |
+
stride:
|
75 |
+
- 1
|
76 |
+
dilation:
|
77 |
+
- 1
|
78 |
+
dropout: 0.0
|
79 |
+
residual: false
|
80 |
+
separable: true
|
81 |
+
se: true
|
82 |
+
se_context_size: -1
|
83 |
+
- filters: 512
|
84 |
+
repeat: 5
|
85 |
+
kernel:
|
86 |
+
- 11
|
87 |
+
stride:
|
88 |
+
- 2
|
89 |
+
dilation:
|
90 |
+
- 1
|
91 |
+
dropout: 0.0
|
92 |
+
residual: true
|
93 |
+
separable: true
|
94 |
+
se: true
|
95 |
+
se_context_size: -1
|
96 |
+
stride_last: true
|
97 |
+
residual_mode: stride_add
|
98 |
+
- filters: 512
|
99 |
+
repeat: 5
|
100 |
+
kernel:
|
101 |
+
- 13
|
102 |
+
stride:
|
103 |
+
- 1
|
104 |
+
dilation:
|
105 |
+
- 1
|
106 |
+
dropout: 0.0
|
107 |
+
residual: true
|
108 |
+
separable: true
|
109 |
+
se: true
|
110 |
+
se_context_size: -1
|
111 |
+
- filters: 512
|
112 |
+
repeat: 5
|
113 |
+
kernel:
|
114 |
+
- 15
|
115 |
+
stride:
|
116 |
+
- 1
|
117 |
+
dilation:
|
118 |
+
- 1
|
119 |
+
dropout: 0.0
|
120 |
+
residual: true
|
121 |
+
separable: true
|
122 |
+
se: true
|
123 |
+
se_context_size: -1
|
124 |
+
- filters: 512
|
125 |
+
repeat: 5
|
126 |
+
kernel:
|
127 |
+
- 17
|
128 |
+
stride:
|
129 |
+
- 1
|
130 |
+
dilation:
|
131 |
+
- 1
|
132 |
+
dropout: 0.0
|
133 |
+
residual: true
|
134 |
+
separable: true
|
135 |
+
se: true
|
136 |
+
se_context_size: -1
|
137 |
+
- filters: 512
|
138 |
+
repeat: 5
|
139 |
+
kernel:
|
140 |
+
- 19
|
141 |
+
stride:
|
142 |
+
- 1
|
143 |
+
dilation:
|
144 |
+
- 1
|
145 |
+
dropout: 0.0
|
146 |
+
residual: true
|
147 |
+
separable: true
|
148 |
+
se: true
|
149 |
+
se_context_size: -1
|
150 |
+
- filters: 512
|
151 |
+
repeat: 5
|
152 |
+
kernel:
|
153 |
+
- 21
|
154 |
+
stride:
|
155 |
+
- 1
|
156 |
+
dilation:
|
157 |
+
- 1
|
158 |
+
dropout: 0.0
|
159 |
+
residual: true
|
160 |
+
separable: true
|
161 |
+
se: true
|
162 |
+
se_context_size: -1
|
163 |
+
- filters: 512
|
164 |
+
repeat: 5
|
165 |
+
kernel:
|
166 |
+
- 13
|
167 |
+
stride:
|
168 |
+
- 2
|
169 |
+
dilation:
|
170 |
+
- 1
|
171 |
+
dropout: 0.0
|
172 |
+
residual: true
|
173 |
+
separable: true
|
174 |
+
se: true
|
175 |
+
se_context_size: -1
|
176 |
+
stride_last: true
|
177 |
+
residual_mode: stride_add
|
178 |
+
- filters: 512
|
179 |
+
repeat: 5
|
180 |
+
kernel:
|
181 |
+
- 15
|
182 |
+
stride:
|
183 |
+
- 1
|
184 |
+
dilation:
|
185 |
+
- 1
|
186 |
+
dropout: 0.0
|
187 |
+
residual: true
|
188 |
+
separable: true
|
189 |
+
se: true
|
190 |
+
se_context_size: -1
|
191 |
+
- filters: 512
|
192 |
+
repeat: 5
|
193 |
+
kernel:
|
194 |
+
- 17
|
195 |
+
stride:
|
196 |
+
- 1
|
197 |
+
dilation:
|
198 |
+
- 1
|
199 |
+
dropout: 0.0
|
200 |
+
residual: true
|
201 |
+
separable: true
|
202 |
+
se: true
|
203 |
+
se_context_size: -1
|
204 |
+
- filters: 512
|
205 |
+
repeat: 5
|
206 |
+
kernel:
|
207 |
+
- 19
|
208 |
+
stride:
|
209 |
+
- 1
|
210 |
+
dilation:
|
211 |
+
- 1
|
212 |
+
dropout: 0.0
|
213 |
+
residual: true
|
214 |
+
separable: true
|
215 |
+
se: true
|
216 |
+
se_context_size: -1
|
217 |
+
- filters: 512
|
218 |
+
repeat: 5
|
219 |
+
kernel:
|
220 |
+
- 21
|
221 |
+
stride:
|
222 |
+
- 1
|
223 |
+
dilation:
|
224 |
+
- 1
|
225 |
+
dropout: 0.0
|
226 |
+
residual: true
|
227 |
+
separable: true
|
228 |
+
se: true
|
229 |
+
se_context_size: -1
|
230 |
+
- filters: 512
|
231 |
+
repeat: 5
|
232 |
+
kernel:
|
233 |
+
- 23
|
234 |
+
stride:
|
235 |
+
- 1
|
236 |
+
dilation:
|
237 |
+
- 1
|
238 |
+
dropout: 0.0
|
239 |
+
residual: true
|
240 |
+
separable: true
|
241 |
+
se: true
|
242 |
+
se_context_size: -1
|
243 |
+
- filters: 512
|
244 |
+
repeat: 5
|
245 |
+
kernel:
|
246 |
+
- 25
|
247 |
+
stride:
|
248 |
+
- 1
|
249 |
+
dilation:
|
250 |
+
- 1
|
251 |
+
dropout: 0.0
|
252 |
+
residual: true
|
253 |
+
separable: true
|
254 |
+
se: true
|
255 |
+
se_context_size: -1
|
256 |
+
- filters: 512
|
257 |
+
repeat: 5
|
258 |
+
kernel:
|
259 |
+
- 25
|
260 |
+
stride:
|
261 |
+
- 2
|
262 |
+
dilation:
|
263 |
+
- 1
|
264 |
+
dropout: 0.0
|
265 |
+
residual: true
|
266 |
+
separable: true
|
267 |
+
se: true
|
268 |
+
se_context_size: -1
|
269 |
+
stride_last: true
|
270 |
+
residual_mode: stride_add
|
271 |
+
- filters: 512
|
272 |
+
repeat: 5
|
273 |
+
kernel:
|
274 |
+
- 27
|
275 |
+
stride:
|
276 |
+
- 1
|
277 |
+
dilation:
|
278 |
+
- 1
|
279 |
+
dropout: 0.0
|
280 |
+
residual: true
|
281 |
+
separable: true
|
282 |
+
se: true
|
283 |
+
se_context_size: -1
|
284 |
+
- filters: 512
|
285 |
+
repeat: 5
|
286 |
+
kernel:
|
287 |
+
- 29
|
288 |
+
stride:
|
289 |
+
- 1
|
290 |
+
dilation:
|
291 |
+
- 1
|
292 |
+
dropout: 0.0
|
293 |
+
residual: true
|
294 |
+
separable: true
|
295 |
+
se: true
|
296 |
+
se_context_size: -1
|
297 |
+
- filters: 512
|
298 |
+
repeat: 5
|
299 |
+
kernel:
|
300 |
+
- 31
|
301 |
+
stride:
|
302 |
+
- 1
|
303 |
+
dilation:
|
304 |
+
- 1
|
305 |
+
dropout: 0.0
|
306 |
+
residual: true
|
307 |
+
separable: true
|
308 |
+
se: true
|
309 |
+
se_context_size: -1
|
310 |
+
- filters: 512
|
311 |
+
repeat: 5
|
312 |
+
kernel:
|
313 |
+
- 33
|
314 |
+
stride:
|
315 |
+
- 1
|
316 |
+
dilation:
|
317 |
+
- 1
|
318 |
+
dropout: 0.0
|
319 |
+
residual: true
|
320 |
+
separable: true
|
321 |
+
se: true
|
322 |
+
se_context_size: -1
|
323 |
+
- filters: 512
|
324 |
+
repeat: 5
|
325 |
+
kernel:
|
326 |
+
- 35
|
327 |
+
stride:
|
328 |
+
- 1
|
329 |
+
dilation:
|
330 |
+
- 1
|
331 |
+
dropout: 0.0
|
332 |
+
residual: true
|
333 |
+
separable: true
|
334 |
+
se: true
|
335 |
+
se_context_size: -1
|
336 |
+
- filters: 512
|
337 |
+
repeat: 5
|
338 |
+
kernel:
|
339 |
+
- 37
|
340 |
+
stride:
|
341 |
+
- 1
|
342 |
+
dilation:
|
343 |
+
- 1
|
344 |
+
dropout: 0.0
|
345 |
+
residual: true
|
346 |
+
separable: true
|
347 |
+
se: true
|
348 |
+
se_context_size: -1
|
349 |
+
- filters: 512
|
350 |
+
repeat: 5
|
351 |
+
kernel:
|
352 |
+
- 39
|
353 |
+
stride:
|
354 |
+
- 1
|
355 |
+
dilation:
|
356 |
+
- 1
|
357 |
+
dropout: 0.0
|
358 |
+
residual: true
|
359 |
+
separable: true
|
360 |
+
se: true
|
361 |
+
se_context_size: -1
|
362 |
+
- filters: 640
|
363 |
+
repeat: 1
|
364 |
+
kernel:
|
365 |
+
- 41
|
366 |
+
stride:
|
367 |
+
- 1
|
368 |
+
dilation:
|
369 |
+
- 1
|
370 |
+
dropout: 0.0
|
371 |
+
residual: false
|
372 |
+
separable: true
|
373 |
+
se: true
|
374 |
+
se_context_size: -1
|
375 |
+
decoder:
|
376 |
+
_target_: nemo.collections.asr.modules.ConvASRDecoder
|
377 |
+
feat_in: 640
|
378 |
+
num_classes: 174
|
379 |
+
vocabulary:
|
380 |
+
- <unk>
|
381 |
+
- ்க
|
382 |
+
- ம்
|
383 |
+
- ▁ப
|
384 |
+
- க்க
|
385 |
+
- ்த
|
386 |
+
- ன்
|
387 |
+
- ்ட
|
388 |
+
- ▁வ
|
389 |
+
- ங்க
|
390 |
+
- ரு
|
391 |
+
- ▁இ
|
392 |
+
- ▁க
|
393 |
+
- ▁அ
|
394 |
+
- ▁எ
|
395 |
+
- க்கு
|
396 |
+
- ▁ச
|
397 |
+
- ல்
|
398 |
+
- ந்த
|
399 |
+
- ட்ட
|
400 |
+
- ப்
|
401 |
+
- ▁ந
|
402 |
+
- த்த
|
403 |
+
- து
|
404 |
+
- ப்ப
|
405 |
+
- ▁ம
|
406 |
+
- ல்ல
|
407 |
+
- ▁த
|
408 |
+
- ்ச
|
409 |
+
- ன்ன
|
410 |
+
- ▁இரு
|
411 |
+
- டி
|
412 |
+
- டு
|
413 |
+
- ▁போ
|
414 |
+
- ும்
|
415 |
+
- ந்து
|
416 |
+
- ட்டு
|
417 |
+
- ான்
|
418 |
+
- ாங்க
|
419 |
+
- ச்ச
|
420 |
+
- ிய
|
421 |
+
- ண்
|
422 |
+
- மா
|
423 |
+
- த்து
|
424 |
+
- ▁வந்து
|
425 |
+
- ர்
|
426 |
+
- ▁பா
|
427 |
+
- ண்ண
|
428 |
+
- ▁ஒ
|
429 |
+
- ல்லா
|
430 |
+
- ண்ட
|
431 |
+
- ▁ஆ
|
432 |
+
- ம்ப
|
433 |
+
- ேன்
|
434 |
+
- னு
|
435 |
+
- க்கா
|
436 |
+
- னா
|
437 |
+
- ைய
|
438 |
+
- ▁மா
|
439 |
+
- ▁இருக்கு
|
440 |
+
- ▁கொ
|
441 |
+
- ஸ்
|
442 |
+
- ரி
|
443 |
+
- ▁என்ன
|
444 |
+
- ▁சொ
|
445 |
+
- ▁சா
|
446 |
+
- ச்சு
|
447 |
+
- ள்
|
448 |
+
- ▁ர
|
449 |
+
- ▁பண்ண
|
450 |
+
- ோம்
|
451 |
+
- லா
|
452 |
+
- ▁அப்ப
|
453 |
+
- ட்
|
454 |
+
- ஞ்ச
|
455 |
+
- ▁கா
|
456 |
+
- யி
|
457 |
+
- ய்
|
458 |
+
- ▁எங்க
|
459 |
+
- ▁ஏ
|
460 |
+
- ▁நா
|
461 |
+
- ▁ஒரு
|
462 |
+
- ▁அவ
|
463 |
+
- ீங்க
|
464 |
+
- ியா
|
465 |
+
- ▁அது
|
466 |
+
- ▁எல்லா
|
467 |
+
- ▁கு
|
468 |
+
- தி
|
469 |
+
- ▁இல்ல
|
470 |
+
- ▁வெ
|
471 |
+
- ▁வே
|
472 |
+
- ▁தான்
|
473 |
+
- யா
|
474 |
+
- ▁பே
|
475 |
+
- றது
|
476 |
+
- ▁செ
|
477 |
+
- ுக்கு
|
478 |
+
- ▁இருக்க
|
479 |
+
- ண்டு
|
480 |
+
- ில
|
481 |
+
- ▁பி
|
482 |
+
- ▁ட
|
483 |
+
- ிரு
|
484 |
+
- ளா
|
485 |
+
- ்
|
486 |
+
- ▁
|
487 |
+
- ு
|
488 |
+
- க
|
489 |
+
- ா
|
490 |
+
- த
|
491 |
+
- ட
|
492 |
+
- ப
|
493 |
+
- ம
|
494 |
+
- ி
|
495 |
+
- ன
|
496 |
+
- ர
|
497 |
+
- ல
|
498 |
+
- வ
|
499 |
+
- ச
|
500 |
+
- ந
|
501 |
+
- ங
|
502 |
+
- ய
|
503 |
+
- ே
|
504 |
+
- ோ
|
505 |
+
- ண
|
506 |
+
- இ
|
507 |
+
- .
|
508 |
+
- அ
|
509 |
+
- எ
|
510 |
+
- ள
|
511 |
+
- ை
|
512 |
+
- ெ
|
513 |
+
- ற
|
514 |
+
- ொ
|
515 |
+
- ீ
|
516 |
+
- ஸ
|
517 |
+
- ூ
|
518 |
+
- ஒ
|
519 |
+
- ஆ
|
520 |
+
- ழ
|
521 |
+
- ஞ
|
522 |
+
- ஏ
|
523 |
+
- ஷ
|
524 |
+
- ஜ
|
525 |
+
- ','
|
526 |
+
- உ
|
527 |
+
- ஊ
|
528 |
+
- ஓ
|
529 |
+
- ஃ
|
530 |
+
- ஹ
|
531 |
+
- ஐ
|
532 |
+
- ௌ
|
533 |
+
- ஈ
|
534 |
+
- '!'
|
535 |
+
- '5'
|
536 |
+
- '2'
|
537 |
+
- '0'
|
538 |
+
- '1'
|
539 |
+
- _
|
540 |
+
- '3'
|
541 |
+
- '6'
|
542 |
+
- ஂ
|
543 |
+
- g
|
544 |
+
- k
|
545 |
+
- m
|
546 |
+
- p
|
547 |
+
- s
|
548 |
+
- '௫'
|
549 |
+
- '7'
|
550 |
+
- '8'
|
551 |
+
- '9'
|
552 |
+
- t
|
553 |
+
- '௯'
|
554 |
+
optim:
|
555 |
+
name: novograd
|
556 |
+
lr: 0.025
|
557 |
+
betas:
|
558 |
+
- 0.8
|
559 |
+
- 0.25
|
560 |
+
weight_decay: 0.001
|
561 |
+
sched:
|
562 |
+
name: CosineAnnealing
|
563 |
+
warmup_steps: null
|
564 |
+
warmup_ratio: 0.1
|
565 |
+
min_lr: 1.0e-09
|
566 |
+
last_epoch: -1
|
567 |
+
target: nemo.collections.asr.models.ctc_bpe_models.EncDecCTCModelBPE
|
568 |
+
nemo_version: 1.22.0
|
569 |
+
decoding:
|
570 |
+
strategy: greedy
|
571 |
+
preserve_alignments: null
|
572 |
+
compute_timestamps: null
|
573 |
+
word_seperator: ' '
|
574 |
+
ctc_timestamp_type: all
|
575 |
+
batch_dim_index: 0
|
576 |
+
greedy:
|
577 |
+
preserve_alignments: false
|
578 |
+
compute_timestamps: false
|
579 |
+
preserve_frame_confidence: false
|
580 |
+
confidence_method_cfg:
|
581 |
+
name: entropy
|
582 |
+
entropy_type: tsallis
|
583 |
+
alpha: 0.33
|
584 |
+
entropy_norm: exp
|
585 |
+
temperature: DEPRECATED
|
586 |
+
beam:
|
587 |
+
beam_size: 4
|
588 |
+
search_type: default
|
589 |
+
preserve_alignments: false
|
590 |
+
compute_timestamps: false
|
591 |
+
return_best_hypothesis: true
|
592 |
+
beam_alpha: 1.0
|
593 |
+
beam_beta: 0.0
|
594 |
+
kenlm_path: null
|
595 |
+
flashlight_cfg:
|
596 |
+
lexicon_path: null
|
597 |
+
boost_path: null
|
598 |
+
beam_size_token: 16
|
599 |
+
beam_threshold: 20.0
|
600 |
+
unk_weight: -.inf
|
601 |
+
sil_weight: 0.0
|
602 |
+
pyctcdecode_cfg:
|
603 |
+
beam_prune_logp: -10.0
|
604 |
+
token_min_logp: -5.0
|
605 |
+
prune_history: false
|
606 |
+
hotwords: null
|
607 |
+
hotword_weight: 10.0
|
608 |
+
confidence_cfg:
|
609 |
+
preserve_frame_confidence: false
|
610 |
+
preserve_token_confidence: false
|
611 |
+
preserve_word_confidence: false
|
612 |
+
exclude_blank: true
|
613 |
+
aggregation: min
|
614 |
+
method_cfg:
|
615 |
+
name: entropy
|
616 |
+
entropy_type: tsallis
|
617 |
+
alpha: 0.33
|
618 |
+
entropy_norm: exp
|
619 |
+
temperature: DEPRECATED
|
620 |
+
temperature: 1.0
|
lightning_logs.txt
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
GPU available: True (cuda), used: True
|
2 |
+
TPU available: False, using: 0 TPU cores
|
3 |
+
IPU available: False, using: 0 IPUs
|
4 |
+
HPU available: False, using: 0 HPUs
|
5 |
+
LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]
|
6 |
+
|
7 |
+
| Name | Type | Params
|
8 |
+
------------------------------------------------------------------------
|
9 |
+
0 | preprocessor | AudioToMelSpectrogramPreprocessor | 0
|
10 |
+
1 | encoder | ConvASREncoder | 36.3 M
|
11 |
+
2 | decoder | ConvASRDecoder | 112 K
|
12 |
+
3 | loss | CTCLoss | 0
|
13 |
+
4 | spec_augmentation | SpectrogramAugmentation | 0
|
14 |
+
5 | wer | WER | 0
|
15 |
+
------------------------------------------------------------------------
|
16 |
+
1.8 M Trainable params
|
17 |
+
34.7 M Non-trainable params
|
18 |
+
36.4 M Total params
|
19 |
+
145.798 Total estimated model params size (MB)
|
20 |
+
Epoch 9, global step 5070: 'val_wer' reached 0.38640 (best 0.38640), saving model to '/content/experiments/lang-ta/ASR-Model-Language-ta/2024-02-15_22-34-08/checkpoints/ASR-Model-Language-ta--val_wer=0.3864-epoch=9.ckpt' as top 3
|
21 |
+
Epoch 19, global step 10140: 'val_wer' reached 0.35093 (best 0.35093), saving model to '/content/experiments/lang-ta/ASR-Model-Language-ta/2024-02-15_22-34-08/checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=19.ckpt' as top 3
|
22 |
+
`Trainer.fit` stopped: `max_epochs=20` reached.
|
23 |
+
Restoring states from the checkpoint path at /content/experiments/lang-ta/ASR-Model-Language-ta/2024-02-15_22-34-08/checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=19.ckpt
|
24 |
+
Restored all states from the checkpoint at /content/experiments/lang-ta/ASR-Model-Language-ta/2024-02-15_22-34-08/checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=19.ckpt
|
nemo_error_log.txt
ADDED
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[NeMo W 2024-02-15 22:31:14 modelPT:165] If you intend to do training or fine-tuning, please call the ModelPT.setup_training_data() method and provide a valid configuration file to setup the train data loader.
|
2 |
+
Train config :
|
3 |
+
manifest_filepath: /content/datasets/ta/yaygomii/Tamil-Speech-Dialect-Corpus-Shuffled-Split/train/train_yaygomii_Tamil-Speech-Dialect-Corpus-Shuffled-Split_manifest_processed.json,/content/datasets/ta/yaygomii/Tamil-Speech-Dialect-Corpus-Shuffled-Split/valid/valid_yaygomii_Tamil-Speech-Dialect-Corpus-Shuffled-Split_manifest_processed.json
|
4 |
+
sample_rate: 16000
|
5 |
+
batch_size: 16
|
6 |
+
trim_silence: true
|
7 |
+
max_duration: 16.7
|
8 |
+
shuffle: true
|
9 |
+
is_tarred: false
|
10 |
+
tarred_audio_filepaths: null
|
11 |
+
num_workers: 8
|
12 |
+
pin_memory: true
|
13 |
+
use_start_end_token: true
|
14 |
+
|
15 |
+
[NeMo W 2024-02-15 22:31:14 modelPT:172] If you intend to do validation, please call the ModelPT.setup_validation_data() or ModelPT.setup_multiple_validation_data() method and provide a valid configuration file to setup the validation data loader(s).
|
16 |
+
Validation config :
|
17 |
+
manifest_filepath: /content/datasets/ta/yaygomii/Tamil-Speech-Dialect-Corpus-Shuffled-Split/test/test_yaygomii_Tamil-Speech-Dialect-Corpus-Shuffled-Split_manifest_processed.json
|
18 |
+
sample_rate: 16000
|
19 |
+
batch_size: 8
|
20 |
+
shuffle: false
|
21 |
+
num_workers: 8
|
22 |
+
pin_memory: true
|
23 |
+
use_start_end_token: true
|
24 |
+
trim_silence: true
|
25 |
+
|
26 |
+
[NeMo W 2024-02-15 22:31:14 modelPT:178] Please call the ModelPT.setup_test_data() or ModelPT.setup_multiple_test_data() method and provide a valid configuration file to setup the test data loader(s).
|
27 |
+
Test config :
|
28 |
+
manifest_filepath: /content/datasets/ta/yaygomii/Tamil-Speech-Dialect-Corpus-Shuffled-Split/test/test_yaygomii_Tamil-Speech-Dialect-Corpus-Shuffled-Split_manifest_processed.json
|
29 |
+
sample_rate: 16000
|
30 |
+
batch_size: 8
|
31 |
+
shuffle: false
|
32 |
+
num_workers: 8
|
33 |
+
pin_memory: true
|
34 |
+
use_start_end_token: true
|
35 |
+
trim_silence: true
|
36 |
+
|
37 |
+
[NeMo W 2024-02-15 22:32:34 nemo_logging:349] /usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 8 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.
|
38 |
+
warnings.warn(_create_warning_msg(
|
39 |
+
|
40 |
+
[NeMo W 2024-02-15 22:34:21 nemo_logging:349] /usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 8 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.
|
41 |
+
warnings.warn(_create_warning_msg(
|
42 |
+
|
nemo_log_globalrank-0_localrank-0.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|