tanthinhdt
commited on
Commit
•
d6a582f
1
Parent(s):
7e74764
Upload folder using huggingface_hub
Browse files- .gitattributes +3 -0
- test/vasr/audio/cer.132842 +25 -0
- test/vasr/audio/decode.log +0 -0
- test/vasr/audio/decode_test.log +3 -0
- test/vasr/audio/hypo-132842.json +3 -0
- test/vasr/audio/wer.132842 +25 -0
- test/vasr/audio_visual/cer.132842 +25 -0
- test/vasr/audio_visual/decode.log +0 -0
- test/vasr/audio_visual/hypo-132842.json +3 -0
- test/vasr/audio_visual/wer.132842 +25 -0
.gitattributes
CHANGED
@@ -34,3 +34,6 @@ 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 |
test/hypo-132842.json 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 |
test/hypo-132842.json filter=lfs diff=lfs merge=lfs -text
|
37 |
+
test/vasr/audio/decode_test.log filter=lfs diff=lfs merge=lfs -text
|
38 |
+
test/vasr/audio/hypo-132842.json filter=lfs diff=lfs merge=lfs -text
|
39 |
+
test/vasr/audio_visual/hypo-132842.json filter=lfs diff=lfs merge=lfs -text
|
test/vasr/audio/cer.132842
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
CER: 12.286203303986076
|
2 |
+
err / num_ref_chars = 237388 / 1932151
|
3 |
+
|
4 |
+
_name: null
|
5 |
+
max_length: 300
|
6 |
+
max_new_tokens: null
|
7 |
+
min_length: 0
|
8 |
+
min_new_tokens: null
|
9 |
+
max_time: null
|
10 |
+
do_sample: false
|
11 |
+
num_beams: 20
|
12 |
+
num_beam_groups: 1
|
13 |
+
temperature: 1.0
|
14 |
+
top_k: 50
|
15 |
+
top_p: 1.0
|
16 |
+
min_p: null
|
17 |
+
typical_p: 1.0
|
18 |
+
epsilon_cutoff: 0.0
|
19 |
+
eta_cutoff: 0.0
|
20 |
+
diversity_penalty: 0.0
|
21 |
+
repetition_penalty: 1.0
|
22 |
+
encoder_repetition_penalty: 1.0
|
23 |
+
length_penalty: 1.0
|
24 |
+
no_repeat_ngram_size: 0.0
|
25 |
+
exponential_decay_length_penalty: null
|
test/vasr/audio/decode.log
ADDED
File without changes
|
test/vasr/audio/decode_test.log
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3b08548ce64e7b4e25e34e42d2f6e05970496515a2baebdaceed181757489fbb
|
3 |
+
size 17207587
|
test/vasr/audio/hypo-132842.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4d6af839d2aa1d66f13dac1ae6ac1bf67139b3500fa15c9a988f9c9c438d431b
|
3 |
+
size 11961723
|
test/vasr/audio/wer.132842
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
WER: 19.65446517107237
|
2 |
+
err / num_ref_words = 92762 / 471964
|
3 |
+
|
4 |
+
_name: null
|
5 |
+
max_length: 300
|
6 |
+
max_new_tokens: null
|
7 |
+
min_length: 0
|
8 |
+
min_new_tokens: null
|
9 |
+
max_time: null
|
10 |
+
do_sample: false
|
11 |
+
num_beams: 20
|
12 |
+
num_beam_groups: 1
|
13 |
+
temperature: 1.0
|
14 |
+
top_k: 50
|
15 |
+
top_p: 1.0
|
16 |
+
min_p: null
|
17 |
+
typical_p: 1.0
|
18 |
+
epsilon_cutoff: 0.0
|
19 |
+
eta_cutoff: 0.0
|
20 |
+
diversity_penalty: 0.0
|
21 |
+
repetition_penalty: 1.0
|
22 |
+
encoder_repetition_penalty: 1.0
|
23 |
+
length_penalty: 1.0
|
24 |
+
no_repeat_ngram_size: 0.0
|
25 |
+
exponential_decay_length_penalty: null
|
test/vasr/audio_visual/cer.132842
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
CER: 10.567186519066057
|
2 |
+
err / num_ref_chars = 204174 / 1932151
|
3 |
+
|
4 |
+
_name: null
|
5 |
+
max_length: 300
|
6 |
+
max_new_tokens: null
|
7 |
+
min_length: 0
|
8 |
+
min_new_tokens: null
|
9 |
+
max_time: null
|
10 |
+
do_sample: false
|
11 |
+
num_beams: 20
|
12 |
+
num_beam_groups: 1
|
13 |
+
temperature: 1.0
|
14 |
+
top_k: 50
|
15 |
+
top_p: 1.0
|
16 |
+
min_p: null
|
17 |
+
typical_p: 1.0
|
18 |
+
epsilon_cutoff: 0.0
|
19 |
+
eta_cutoff: 0.0
|
20 |
+
diversity_penalty: 0.0
|
21 |
+
repetition_penalty: 1.0
|
22 |
+
encoder_repetition_penalty: 1.0
|
23 |
+
length_penalty: 1.0
|
24 |
+
no_repeat_ngram_size: 0.0
|
25 |
+
exponential_decay_length_penalty: null
|
test/vasr/audio_visual/decode.log
ADDED
File without changes
|
test/vasr/audio_visual/hypo-132842.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5ad25c95b8928172b9fbc885b2b920002ea1ffecad8759466e012b518287a078
|
3 |
+
size 11988955
|
test/vasr/audio_visual/wer.132842
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
WER: 17.284581027366492
|
2 |
+
err / num_ref_words = 81577 / 471964
|
3 |
+
|
4 |
+
_name: null
|
5 |
+
max_length: 300
|
6 |
+
max_new_tokens: null
|
7 |
+
min_length: 0
|
8 |
+
min_new_tokens: null
|
9 |
+
max_time: null
|
10 |
+
do_sample: false
|
11 |
+
num_beams: 20
|
12 |
+
num_beam_groups: 1
|
13 |
+
temperature: 1.0
|
14 |
+
top_k: 50
|
15 |
+
top_p: 1.0
|
16 |
+
min_p: null
|
17 |
+
typical_p: 1.0
|
18 |
+
epsilon_cutoff: 0.0
|
19 |
+
eta_cutoff: 0.0
|
20 |
+
diversity_penalty: 0.0
|
21 |
+
repetition_penalty: 1.0
|
22 |
+
encoder_repetition_penalty: 1.0
|
23 |
+
length_penalty: 1.0
|
24 |
+
no_repeat_ngram_size: 0.0
|
25 |
+
exponential_decay_length_penalty: null
|