Upload 92 files
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/47epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/48epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/50epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/52epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/54epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/55epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/57epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/58epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/59epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/60epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/RESULTS.md +108 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/checkpoint.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/config.yaml +227 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/acc.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/backward_time.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/cer.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/clip.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/forward_time.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/gpu_max_cached_mem_GB.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/grad_norm.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/iter_time.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/loss.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/loss_att.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/loss_scale.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/optim0_lr0.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/optim_step_time.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/train_time.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/wer.png +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/latest.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/run.sh +1 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/tensorboard/train/events.out.tfevents.1704228297.de-74279-k2-train-7-1218101249-5bcbfb5567-jsftr.1813079.0 +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/tensorboard/train/events.out.tfevents.1705380444.de-74279-k2-train-7-1218101249-5bcbfb5567-jsftr.3520216.0 +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/tensorboard/valid/events.out.tfevents.1704228297.de-74279-k2-train-7-1218101249-5bcbfb5567-jsftr.1813079.1 +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/tensorboard/valid/events.out.tfevents.1705380444.de-74279-k2-train-7-1218101249-5bcbfb5567-jsftr.3520216.1 +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/train.1.log +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/train.log +0 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/valid.acc.ave.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/valid.acc.ave_10best.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/valid.acc.best.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/43epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/44epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/48epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/49epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/51epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/52epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/54epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/56epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/57epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/60epoch.pth +3 -0
- large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/RESULTS.md +63 -0
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/47epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ec2e4fcd4f8b304e0b86667cd1868eaa7505bfd6cab4bb8e3ba25a4bec14441b
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/48epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:723f61d6ae38c4dfe06a8c5d0770c057d8b8e17909909b64117beaad445ea733
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/50epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:72f45ac4204e62012096f8f811c7fed27dab5140166b6d33c50ef0fbcab3439d
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/52epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cfac6a2bc79e49b74cd49d3ef954c755f36885277d1dcdc57c84dd497543f6ac
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/54epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0107f21cc4ec3208e81bc4065b1b7c2d6e9ebea3fa0df09518531a6268430783
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/55epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5b7e15c311a51848a22efa5626215db5b3c95a4f0e02ef25d0975eaa715f7d06
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/57epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ec381eb186b0df466f4334af0870480ee358b3641d0fb26ac8b0e2ed39b51bb0
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/58epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fbbb194b9c615c467052711edbac4d5e8626affe4298ead189570fa309e3c7f7
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/59epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:10dd51d66a258909a8f6aab8ab41d6eaa08dc2dd0b08c1e4845ad4705befd817
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/60epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4c215ed61c217a50608410e89c77ae5f208ca92c388249d897ad93e33561e3ba
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/RESULTS.md
ADDED
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
<!-- Generated by scripts/utils/show_asr_result.sh -->
|
2 |
+
# RESULTS
|
3 |
+
## Environments
|
4 |
+
- date: `Tue Mar 12 04:03:00 CST 2024`
|
5 |
+
- python version: `3.9.18 (main, Sep 11 2023, 13:41:44) [GCC 11.2.0]`
|
6 |
+
- espnet version: `espnet 202308`
|
7 |
+
- pytorch version: `pytorch 1.12.1+cu116`
|
8 |
+
- Git hash: `884659f9ee95374811015381c976fa3b4f6e01db`
|
9 |
+
- Commit date: `Thu Nov 23 00:23:29 2023 +0800`
|
10 |
+
|
11 |
+
## exp/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new
|
12 |
+
### WER
|
13 |
+
|
14 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
15 |
+
|---|---|---|---|---|---|---|---|---|
|
16 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk|3315|226216|81.3|7.3|11.4|11.0|29.7|96.8|
|
17 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_spk1|3315|226216|40.0|14.4|45.6|3.1|63.1|100.0|
|
18 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_spk2|3315|226216|38.6|15.5|45.9|3.1|64.5|100.0|
|
19 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_sys7_8khz_spk1_large|1606|135101|41.7|11.2|47.0|2.5|60.8|100.0|
|
20 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_sys7_8khz_spk2_large|1606|135101|39.9|13.0|47.1|2.5|62.6|100.0|
|
21 |
+
|decode_sot_asr_model_valid.acc.best/dev_3spk|2059|209679|69.9|14.7|15.4|10.0|40.1|100.0|
|
22 |
+
|decode_sot_asr_model_valid.acc.best/dev_3spk_kaldi_fmt|1004|124462|73.7|11.7|14.5|6.0|32.3|99.9|
|
23 |
+
|decode_sot_asr_model_valid.acc.best/dev_4spk|1467|200029|57.8|21.1|21.0|8.2|50.3|100.0|
|
24 |
+
|decode_sot_asr_model_valid.acc.best/dev_4spk_kaldi_fmt|721|119166|61.5|16.5|22.0|5.4|43.9|100.0|
|
25 |
+
|decode_sot_asr_model_valid.acc.best/dev_kaldi_fmt|605|47659|84.5|10.0|5.5|4.2|19.6|93.6|
|
26 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk|4570|301042|81.9|6.6|11.5|10.2|28.3|97.6|
|
27 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk_spk1|4570|301042|38.2|15.0|46.8|3.0|64.9|100.0|
|
28 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk_spk2|4570|301042|37.4|14.5|48.1|2.3|65.0|100.0|
|
29 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_3spk|2072|212871|71.0|13.2|15.7|10.9|39.8|99.9|
|
30 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_4spk|1326|185394|59.8|19.4|20.8|9.1|49.3|100.0|
|
31 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk|4663|336490|81.0|8.8|10.2|10.8|29.8|99.1|
|
32 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk_spk1|4663|336490|34.2|18.6|47.2|3.5|69.3|100.0|
|
33 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk_spk2|4663|336490|40.5|14.8|44.8|3.0|62.5|100.0|
|
34 |
+
|decode_sot_asr_model_valid.acc.best/test-other_3spk|2453|266074|68.4|17.2|14.4|11.0|42.6|100.0|
|
35 |
+
|decode_sot_asr_model_valid.acc.best/test-other_4spk|1795|259138|55.9|23.8|20.3|8.7|52.8|100.0|
|
36 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_2spk_kaldi_fmt|2180|178761|85.1|5.1|9.7|5.9|20.8|95.1|
|
37 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_2spk_sys7_8khz_spk1_large|2180|178761|39.3|12.4|48.3|2.5|63.2|100.0|
|
38 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_2spk_sys7_8khz_spk2_large|2180|178761|39.4|12.6|48.0|2.0|62.6|100.0|
|
39 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_3spk_kaldi_fmt|977|124741|74.1|10.8|15.1|7.1|33.0|99.8|
|
40 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_4spk_kaldi_fmt|632|109072|63.2|15.2|21.6|6.1|42.9|100.0|
|
41 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_kaldi_fmt|961|64007|86.8|8.4|4.7|4.0|17.2|94.8|
|
42 |
+
|decode_sot_asr_model_valid.acc.best/test_other_2spk_kaldi_fmt|2363|205496|83.9|7.3|8.8|6.5|22.6|98.4|
|
43 |
+
|decode_sot_asr_model_valid.acc.best/test_other_2spk_sys7_8khz_spk1_large|2363|205496|35.2|16.2|48.6|2.9|67.6|100.0|
|
44 |
+
|decode_sot_asr_model_valid.acc.best/test_other_2spk_sys7_8khz_spk2_large|2363|205496|41.9|12.8|45.3|2.5|60.6|100.0|
|
45 |
+
|decode_sot_asr_model_valid.acc.best/test_other_3spk_kaldi_fmt|1246|162996|70.6|15.3|14.1|7.6|37.0|100.0|
|
46 |
+
|decode_sot_asr_model_valid.acc.best/test_other_4spk_kaldi_fmt|901|157123|59.0|19.4|21.7|6.4|47.4|100.0|
|
47 |
+
|decode_sot_asr_model_valid.acc.best/test_other_kaldi_fmt|992|80370|81.6|12.2|6.2|5.9|24.3|97.5|
|
48 |
+
|
49 |
+
### CER
|
50 |
+
|
51 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
52 |
+
|---|---|---|---|---|---|---|---|---|
|
53 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk|3315|1230801|85.8|4.2|10.0|9.8|24.0|96.8|
|
54 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_spk1|3315|1230801|47.9|6.1|46.0|3.4|55.5|100.0|
|
55 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_spk2|3315|1230801|47.2|6.5|46.2|3.4|56.2|100.0|
|
56 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_sys7_8khz_spk1_large|1606|735694|48.0|4.5|47.4|2.6|54.5|100.0|
|
57 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_sys7_8khz_spk2_large|1606|735694|47.3|5.3|47.3|2.6|55.2|100.0|
|
58 |
+
|decode_sot_asr_model_valid.acc.best/dev_3spk|2059|1140428|77.7|8.0|14.3|8.6|30.9|100.0|
|
59 |
+
|decode_sot_asr_model_valid.acc.best/dev_3spk_kaldi_fmt|1004|677017|80.1|6.0|13.9|5.0|24.9|99.9|
|
60 |
+
|decode_sot_asr_model_valid.acc.best/dev_4spk|1467|1087409|68.5|10.4|21.1|7.4|39.0|100.0|
|
61 |
+
|decode_sot_asr_model_valid.acc.best/dev_4spk_kaldi_fmt|721|647884|70.2|7.5|22.3|4.8|34.6|100.0|
|
62 |
+
|decode_sot_asr_model_valid.acc.best/dev_kaldi_fmt|605|258151|90.1|4.1|5.8|4.0|13.9|93.6|
|
63 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk|4570|1550429|85.9|4.2|9.9|8.8|22.9|97.6|
|
64 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk_spk1|4570|1550429|46.2|6.4|47.4|3.1|56.9|100.0|
|
65 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk_spk2|4570|1550429|45.8|5.8|48.4|2.7|56.9|100.0|
|
66 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_3spk|2072|1084475|77.9|8.0|14.1|9.1|31.2|99.9|
|
67 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_4spk|1326|938467|69.5|10.1|20.4|8.2|38.7|100.0|
|
68 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk|4663|1742136|86.1|4.9|9.0|9.6|23.5|99.1|
|
69 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk_spk1|4663|1742136|44.1|7.9|48.0|3.6|59.5|100.0|
|
70 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk_spk2|4663|1742136|48.7|6.2|45.1|3.3|54.6|100.0|
|
71 |
+
|decode_sot_asr_model_valid.acc.best/test-other_3spk|2453|1381987|77.3|9.2|13.6|9.6|32.3|100.0|
|
72 |
+
|decode_sot_asr_model_valid.acc.best/test-other_4spk|1795|1346646|67.6|11.6|20.8|8.0|40.4|100.0|
|
73 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_2spk_kaldi_fmt|2180|921344|88.2|2.9|8.9|5.0|16.8|95.1|
|
74 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_2spk_sys7_8khz_spk1_large|2180|921344|46.1|5.0|48.9|2.5|56.4|100.0|
|
75 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_2spk_sys7_8khz_spk2_large|2180|921344|46.9|4.8|48.2|2.2|55.3|100.0|
|
76 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_3spk_kaldi_fmt|977|635802|79.7|6.3|14.0|5.7|26.1|99.8|
|
77 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_4spk_kaldi_fmt|632|552325|71.1|7.2|21.7|5.5|34.4|100.0|
|
78 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_kaldi_fmt|961|329390|91.8|3.4|4.8|4.0|12.2|94.8|
|
79 |
+
|decode_sot_asr_model_valid.acc.best/test_other_2spk_kaldi_fmt|2363|1064868|88.0|3.8|8.2|5.6|17.6|98.4|
|
80 |
+
|decode_sot_asr_model_valid.acc.best/test_other_2spk_sys7_8khz_spk1_large|2363|1064868|44.0|6.7|49.3|2.8|58.7|100.0|
|
81 |
+
|decode_sot_asr_model_valid.acc.best/test_other_2spk_sys7_8khz_spk2_large|2363|1064868|49.1|5.2|45.7|2.6|53.5|100.0|
|
82 |
+
|decode_sot_asr_model_valid.acc.best/test_other_3spk_kaldi_fmt|1246|847159|78.5|7.7|13.8|6.6|28.1|100.0|
|
83 |
+
|decode_sot_asr_model_valid.acc.best/test_other_4spk_kaldi_fmt|901|817228|68.9|8.8|22.3|5.9|37.0|100.0|
|
84 |
+
|decode_sot_asr_model_valid.acc.best/test_other_kaldi_fmt|992|416899|88.3|5.2|6.5|5.7|17.4|97.5|
|
85 |
+
|
86 |
+
### TER
|
87 |
+
|
88 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
89 |
+
|---|---|---|---|---|---|---|---|---|
|
90 |
+
## exp/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/decode_sot_asr_model_valid.acc.best
|
91 |
+
### WER
|
92 |
+
|
93 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
94 |
+
|---|---|---|---|---|---|---|---|---|
|
95 |
+
|org/dev_2spk_kaldi_fmt|1606|135101|84.1|5.7|10.2|6.8|22.7|93.5|
|
96 |
+
|org/tt_mix_clean_reverb_max_16k|3000|3000|0.0|100.0|0.0|3712.3|3812.3|100.0|
|
97 |
+
|
98 |
+
### CER
|
99 |
+
|
100 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
101 |
+
|---|---|---|---|---|---|---|---|---|
|
102 |
+
|org/dev_2spk_kaldi_fmt|1606|735694|87.5|3.0|9.4|5.9|18.4|93.5|
|
103 |
+
|org/tt_mix_clean_reverb_max_16k|3000|143026|17.0|82.9|0.1|370.7|453.8|100.0|
|
104 |
+
|
105 |
+
### TER
|
106 |
+
|
107 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
108 |
+
|---|---|---|---|---|---|---|---|---|
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/checkpoint.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c781ea07d1d9c8d8325bbfc5c127a3fe0ffb838d3f2c07a826f804c91a4b339a
|
3 |
+
size 516979614
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/config.yaml
ADDED
@@ -0,0 +1,227 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
config: conf/tuning/train_sot_asr_conformer_large.yaml
|
2 |
+
print_config: false
|
3 |
+
log_level: INFO
|
4 |
+
drop_last_iter: false
|
5 |
+
dry_run: false
|
6 |
+
iterator_type: sequence
|
7 |
+
valid_iterator_type: null
|
8 |
+
output_dir: exp/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new
|
9 |
+
ngpu: 1
|
10 |
+
seed: 0
|
11 |
+
num_workers: 16
|
12 |
+
num_att_plot: 3
|
13 |
+
dist_backend: nccl
|
14 |
+
dist_init_method: env://
|
15 |
+
dist_world_size: 4
|
16 |
+
dist_rank: 0
|
17 |
+
local_rank: 0
|
18 |
+
dist_master_addr: localhost
|
19 |
+
dist_master_port: 58141
|
20 |
+
dist_launcher: null
|
21 |
+
multiprocessing_distributed: true
|
22 |
+
unused_parameters: false
|
23 |
+
sharded_ddp: false
|
24 |
+
cudnn_enabled: true
|
25 |
+
cudnn_benchmark: false
|
26 |
+
cudnn_deterministic: true
|
27 |
+
collect_stats: false
|
28 |
+
write_collected_feats: false
|
29 |
+
max_epoch: 60
|
30 |
+
patience: null
|
31 |
+
val_scheduler_criterion:
|
32 |
+
- valid
|
33 |
+
- loss
|
34 |
+
early_stopping_criterion:
|
35 |
+
- valid
|
36 |
+
- loss
|
37 |
+
- min
|
38 |
+
best_model_criterion:
|
39 |
+
- - valid
|
40 |
+
- acc
|
41 |
+
- max
|
42 |
+
keep_nbest_models: 10
|
43 |
+
nbest_averaging_interval: 0
|
44 |
+
grad_clip: 5.0
|
45 |
+
grad_clip_type: 2.0
|
46 |
+
grad_noise: false
|
47 |
+
accum_grad: 4
|
48 |
+
no_forward_run: false
|
49 |
+
resume: true
|
50 |
+
train_dtype: float32
|
51 |
+
use_amp: false
|
52 |
+
log_interval: null
|
53 |
+
use_matplotlib: true
|
54 |
+
use_tensorboard: true
|
55 |
+
create_graph_in_tensorboard: false
|
56 |
+
use_wandb: false
|
57 |
+
wandb_project: null
|
58 |
+
wandb_id: null
|
59 |
+
wandb_entity: null
|
60 |
+
wandb_name: null
|
61 |
+
wandb_model_log_interval: -1
|
62 |
+
detect_anomaly: false
|
63 |
+
pretrain_path: null
|
64 |
+
init_param:
|
65 |
+
- /star-home/jinzengrui/dev/espnet/egs2/librimix/sot_asr1_pretrain/exp/asr_train_sot_asr_conformer_raw_en_char_sp/45epoch.pth
|
66 |
+
ignore_init_mismatch: false
|
67 |
+
freeze_param: []
|
68 |
+
num_iters_per_epoch: null
|
69 |
+
batch_size: 20
|
70 |
+
valid_batch_size: null
|
71 |
+
batch_bins: 16000000
|
72 |
+
valid_batch_bins: null
|
73 |
+
train_shape_file:
|
74 |
+
- exp/asr_stats_raw_en_char/train/speech_shape
|
75 |
+
- exp/asr_stats_raw_en_char/train/text_shape.char
|
76 |
+
valid_shape_file:
|
77 |
+
- exp/asr_stats_raw_en_char/valid/speech_shape
|
78 |
+
- exp/asr_stats_raw_en_char/valid/text_shape.char
|
79 |
+
batch_type: numel
|
80 |
+
valid_batch_type: null
|
81 |
+
fold_length:
|
82 |
+
- 80000
|
83 |
+
- 150
|
84 |
+
sort_in_batch: descending
|
85 |
+
shuffle_within_batch: false
|
86 |
+
sort_batch: descending
|
87 |
+
multiple_iterator: false
|
88 |
+
chunk_length: 500
|
89 |
+
chunk_shift_ratio: 0.5
|
90 |
+
num_cache_chunks: 1024
|
91 |
+
chunk_excluded_key_prefixes: []
|
92 |
+
train_data_path_and_name_and_type:
|
93 |
+
- - dump/raw/train_large_kaldi_fmt/wav.scp
|
94 |
+
- speech
|
95 |
+
- kaldi_ark
|
96 |
+
- - dump/raw/train_large_kaldi_fmt/text
|
97 |
+
- text
|
98 |
+
- text
|
99 |
+
valid_data_path_and_name_and_type:
|
100 |
+
- - dump/raw/dev_2spk_kaldi_fmt/wav.scp
|
101 |
+
- speech
|
102 |
+
- kaldi_ark
|
103 |
+
- - dump/raw/dev_2spk_kaldi_fmt/text
|
104 |
+
- text
|
105 |
+
- text
|
106 |
+
allow_variable_data_keys: false
|
107 |
+
max_cache_size: 0.0
|
108 |
+
max_cache_fd: 32
|
109 |
+
valid_max_cache_size: null
|
110 |
+
exclude_weight_decay: false
|
111 |
+
exclude_weight_decay_conf: {}
|
112 |
+
optim: adam
|
113 |
+
optim_conf:
|
114 |
+
lr: 0.002
|
115 |
+
weight_decay: 1.0e-06
|
116 |
+
scheduler: warmuplr
|
117 |
+
scheduler_conf:
|
118 |
+
warmup_steps: 20000
|
119 |
+
token_list:
|
120 |
+
- <blank>
|
121 |
+
- <unk>
|
122 |
+
- <sc>
|
123 |
+
- <space>
|
124 |
+
- E
|
125 |
+
- T
|
126 |
+
- A
|
127 |
+
- O
|
128 |
+
- N
|
129 |
+
- I
|
130 |
+
- H
|
131 |
+
- S
|
132 |
+
- R
|
133 |
+
- D
|
134 |
+
- L
|
135 |
+
- U
|
136 |
+
- M
|
137 |
+
- C
|
138 |
+
- W
|
139 |
+
- F
|
140 |
+
- G
|
141 |
+
- Y
|
142 |
+
- P
|
143 |
+
- B
|
144 |
+
- V
|
145 |
+
- K
|
146 |
+
- ''''
|
147 |
+
- X
|
148 |
+
- J
|
149 |
+
- Q
|
150 |
+
- Z
|
151 |
+
- <sos/eos>
|
152 |
+
init: null
|
153 |
+
input_size: null
|
154 |
+
ctc_conf:
|
155 |
+
dropout_rate: 0.0
|
156 |
+
ctc_type: builtin
|
157 |
+
reduce: true
|
158 |
+
ignore_nan_grad: null
|
159 |
+
zero_infinity: true
|
160 |
+
joint_net_conf: null
|
161 |
+
use_preprocessor: true
|
162 |
+
token_type: char
|
163 |
+
bpemodel: null
|
164 |
+
non_linguistic_symbols: null
|
165 |
+
cleaner: null
|
166 |
+
g2p: null
|
167 |
+
speech_volume_normalize: null
|
168 |
+
rir_scp: null
|
169 |
+
rir_apply_prob: 1.0
|
170 |
+
noise_scp: null
|
171 |
+
noise_apply_prob: 1.0
|
172 |
+
noise_db_range: '13_15'
|
173 |
+
short_noise_thres: 0.5
|
174 |
+
aux_ctc_tasks: []
|
175 |
+
frontend: default
|
176 |
+
frontend_conf:
|
177 |
+
fs: 16k
|
178 |
+
specaug: null
|
179 |
+
specaug_conf: {}
|
180 |
+
normalize: global_mvn
|
181 |
+
normalize_conf:
|
182 |
+
stats_file: exp/asr_stats_raw_en_char/train/feats_stats.npz
|
183 |
+
model: espnet
|
184 |
+
model_conf:
|
185 |
+
ctc_weight: 0.0
|
186 |
+
lsm_weight: 0.1
|
187 |
+
length_normalized_loss: false
|
188 |
+
preencoder: null
|
189 |
+
preencoder_conf: {}
|
190 |
+
encoder: conformer
|
191 |
+
encoder_conf:
|
192 |
+
output_size: 256
|
193 |
+
attention_heads: 4
|
194 |
+
linear_units: 2048
|
195 |
+
num_blocks: 12
|
196 |
+
dropout_rate: 0.1
|
197 |
+
positional_dropout_rate: 0.1
|
198 |
+
attention_dropout_rate: 0.1
|
199 |
+
input_layer: conv2d
|
200 |
+
normalize_before: true
|
201 |
+
macaron_style: true
|
202 |
+
rel_pos_type: latest
|
203 |
+
pos_enc_layer_type: rel_pos
|
204 |
+
selfattention_layer_type: rel_selfattn
|
205 |
+
activation_type: swish
|
206 |
+
use_cnn_module: true
|
207 |
+
cnn_module_kernel: 31
|
208 |
+
postencoder: null
|
209 |
+
postencoder_conf: {}
|
210 |
+
decoder: transformer
|
211 |
+
decoder_conf:
|
212 |
+
attention_heads: 4
|
213 |
+
linear_units: 2048
|
214 |
+
num_blocks: 6
|
215 |
+
dropout_rate: 0.1
|
216 |
+
positional_dropout_rate: 0.1
|
217 |
+
self_attention_dropout_rate: 0.1
|
218 |
+
src_attention_dropout_rate: 0.1
|
219 |
+
preprocessor: multi
|
220 |
+
preprocessor_conf:
|
221 |
+
speaker_change_symbol:
|
222 |
+
- <sc>
|
223 |
+
required:
|
224 |
+
- output_dir
|
225 |
+
- token_list
|
226 |
+
version: '202308'
|
227 |
+
distributed: true
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/acc.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/backward_time.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/cer.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/clip.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/forward_time.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/gpu_max_cached_mem_GB.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/grad_norm.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/iter_time.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/loss.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/loss_att.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/loss_scale.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/optim0_lr0.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/optim_step_time.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/train_time.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/images/wer.png
ADDED
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/latest.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4c215ed61c217a50608410e89c77ae5f208ca92c388249d897ad93e33561e3ba
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/run.sh
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
./asr.sh --lang en --audio_format flac.ark --stage 2 --feats_type raw --token_type char --sot_asr true --max_wav_duration 50 --feats_normalize global_mvn --use_lm false --pretrained_model /star-home/jinzengrui/dev/espnet/egs2/librimix/sot_asr1_pretrain/exp/asr_train_sot_asr_conformer_raw_en_char_sp/45epoch.pth --asr_config conf/tuning/train_sot_asr_conformer_large.yaml --lm_config conf/tuning/train_lm_transformer.yaml --inference_config conf/tuning/decode_sot.yaml --train_set train_large_kaldi_fmt --valid_set dev_2spk_kaldi_fmt --test_sets 'dev_kaldi_fmt test_clean_kaldi_fmt test_other_kaldi_fmt' --ngpu 4 --asr_tag train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new --lm_train_text data/local/other_text/text --bpe_train_text data/train_large_kaldi_fmt/text --stage 10 --stage 11 "$@"; exit $?
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/tensorboard/train/events.out.tfevents.1704228297.de-74279-k2-train-7-1218101249-5bcbfb5567-jsftr.1813079.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c8f651320400a9314034267172216336b3e21b9e518a65553fa75cbfa69d5955
|
3 |
+
size 667008502
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/tensorboard/train/events.out.tfevents.1705380444.de-74279-k2-train-7-1218101249-5bcbfb5567-jsftr.3520216.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:86beae4916acd5f20539b2daeb461e38342baf385c48af42d18f8b0e979504e7
|
3 |
+
size 193681524
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/tensorboard/valid/events.out.tfevents.1704228297.de-74279-k2-train-7-1218101249-5bcbfb5567-jsftr.1813079.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a2c0ffbc3c49d3f0d972dd24e2a750899a4c9c1e7df1ba0b28bcec22aefee522
|
3 |
+
size 12778
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/tensorboard/valid/events.out.tfevents.1705380444.de-74279-k2-train-7-1218101249-5bcbfb5567-jsftr.3520216.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fd9de87f3d29c3dfcbbfa6762a59e413eee1bfaf26e3327da03f5327430c76f2
|
3 |
+
size 4330
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/train.1.log
ADDED
The diff for this file is too large to render.
See raw diff
|
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/train.log
ADDED
The diff for this file is too large to render.
See raw diff
|
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/valid.acc.ave.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:75c15394b6137b25ea2b7609fd48046efa36924ddd91279111fd96b3b98162c7
|
3 |
+
size 172358249
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/valid.acc.ave_10best.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:75c15394b6137b25ea2b7609fd48046efa36924ddd91279111fd96b3b98162c7
|
3 |
+
size 172358249
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new/valid.acc.best.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5b7e15c311a51848a22efa5626215db5b3c95a4f0e02ef25d0975eaa715f7d06
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/43epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:85e6aa2f4634193567c8875dffed462acb4ddee25a64e6d2cc6784bb731b5118
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/44epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c8eab89d55429339128f95488d43c1130146d94388fe0f035b18944d0486db9d
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/48epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5beda5f12025256fd094495ca64605c78a61acaaeb8fea4f3afacedecf54ff51
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/49epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:770bb8e30bb8128f2f12b52495b484509fad7b46921e2882a89a58694f1a1313
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/51epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c414cae0c18b1d880232d1c56249ed3ad00d3001ea345e7205a6e0d0d76a192f
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/52epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:669d29c06a8814a43482c1c8483b0a02c3ce58e966971a81fe7e686ce5c9240f
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/54epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2a38c3b7350afab15a254eced20881144b8c58356a06b08ac7b8dca587363465
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/56epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9aa51356d8e20d16ce16cf03478783f3c94872757f700b7e6f6709b6d828ec8b
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/57epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1389731de517bd849a0f81bcde4f8fab312ff3019b3dd9c9ddd479d5346481b3
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/60epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bf36c914d7d71528a338cd3a769884e121a1ee7eb57709cf3fd9f79a73f3dd04
|
3 |
+
size 172367337
|
large/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr/RESULTS.md
ADDED
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
<!-- Generated by scripts/utils/show_asr_result.sh -->
|
2 |
+
# RESULTS
|
3 |
+
## Environments
|
4 |
+
- date: `Tue Mar 12 04:49:59 CST 2024`
|
5 |
+
- python version: `3.9.18 (main, Sep 11 2023, 13:41:44) [GCC 11.2.0]`
|
6 |
+
- espnet version: `espnet 202308`
|
7 |
+
- pytorch version: `pytorch 1.12.1+cu116`
|
8 |
+
- Git hash: `884659f9ee95374811015381c976fa3b4f6e01db`
|
9 |
+
- Commit date: `Thu Nov 23 00:23:29 2023 +0800`
|
10 |
+
|
11 |
+
## exp/asr_train_sot_asr_conformer_raw_en_char_sp_finetune_ls100_45epoch_new_whamr
|
12 |
+
### WER
|
13 |
+
|
14 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
15 |
+
|---|---|---|---|---|---|---|---|---|
|
16 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk|3315|226216|80.6|7.3|12.1|10.5|29.9|97.3|
|
17 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_kaldi_fmt|1606|135101|83.1|5.7|11.2|6.6|23.5|94.5|
|
18 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_spk1_whamr|3315|226216|39.3|15.3|45.5|2.6|63.4|100.0|
|
19 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_spk2_whamr|3315|226216|40.2|13.6|46.2|2.8|62.6|100.0|
|
20 |
+
|decode_sot_asr_model_valid.acc.best/dev_3spk|2059|209679|68.5|14.2|17.2|10.0|41.4|99.9|
|
21 |
+
|decode_sot_asr_model_valid.acc.best/dev_4spk|1467|200029|57.4|19.8|22.7|8.8|51.3|100.0|
|
22 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk|4570|301042|81.4|6.6|12.0|9.9|28.5|97.6|
|
23 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk_spk1_whamr|4570|301042|38.2|14.3|47.5|2.1|64.0|100.0|
|
24 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk_spk2_whamr|4570|301042|38.7|13.7|47.7|2.5|63.8|100.0|
|
25 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_3spk|2072|212871|70.0|13.0|17.0|10.7|40.7|99.9|
|
26 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_4spk|1326|185394|59.1|18.8|22.1|8.9|49.8|100.0|
|
27 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk|4663|336490|80.4|8.4|11.1|10.4|30.0|99.0|
|
28 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk_spk1_whamr|4663|336490|34.6|17.9|47.5|3.0|68.4|100.0|
|
29 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk_spk2_whamr|4663|336490|40.9|14.3|44.8|2.7|61.8|100.0|
|
30 |
+
|decode_sot_asr_model_valid.acc.best/test-other_3spk|2453|266074|67.2|17.0|15.8|10.5|43.4|100.0|
|
31 |
+
|decode_sot_asr_model_valid.acc.best/test-other_4spk|1795|259138|54.9|23.2|22.0|8.4|53.5|100.0|
|
32 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_2spk_kaldi_fmt|2180|178761|84.4|4.9|10.7|5.9|21.5|95.0|
|
33 |
+
|decode_sot_asr_model_valid.acc.best/test_other_2spk_kaldi_fmt|2363|205496|82.8|7.2|10.0|6.7|23.9|98.1|
|
34 |
+
|decode_sot_asr_model_valid.acc.best/tt_mix_clean_reverb_max_16k|3000|3000|0.0|100.0|0.0|3118.7|3218.7|100.0|
|
35 |
+
|
36 |
+
### CER
|
37 |
+
|
38 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
39 |
+
|---|---|---|---|---|---|---|---|---|
|
40 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk|3315|1230801|85.0|4.6|10.5|9.1|24.1|97.3|
|
41 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_kaldi_fmt|1606|735694|86.4|3.3|10.3|5.6|19.2|94.5|
|
42 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_spk1_whamr|3315|1230801|47.6|6.5|45.9|2.9|55.3|100.0|
|
43 |
+
|decode_sot_asr_model_valid.acc.best/dev_2spk_spk2_whamr|3315|1230801|47.8|5.8|46.4|2.9|55.1|100.0|
|
44 |
+
|decode_sot_asr_model_valid.acc.best/dev_3spk|2059|1140428|76.1|7.9|16.0|8.5|32.4|99.9|
|
45 |
+
|decode_sot_asr_model_valid.acc.best/dev_4spk|1467|1087409|67.5|10.0|22.5|7.6|40.1|100.0|
|
46 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk|4570|1550429|85.3|4.3|10.4|8.5|23.1|97.6|
|
47 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk_spk1_whamr|4570|1550429|46.6|5.8|47.6|2.6|56.1|100.0|
|
48 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_2spk_spk2_whamr|4570|1550429|46.2|5.8|48.0|2.7|56.5|100.0|
|
49 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_3spk|2072|1084475|76.8|7.8|15.4|9.0|32.1|99.9|
|
50 |
+
|decode_sot_asr_model_valid.acc.best/test-clean_4spk|1326|938467|68.6|9.8|21.6|8.0|39.4|100.0|
|
51 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk|4663|1742136|85.4|4.8|9.8|9.2|23.8|99.0|
|
52 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk_spk1_whamr|4663|1742136|44.2|7.7|48.1|3.2|59.0|100.0|
|
53 |
+
|decode_sot_asr_model_valid.acc.best/test-other_2spk_spk2_whamr|4663|1742136|48.9|5.9|45.1|3.0|54.1|100.0|
|
54 |
+
|decode_sot_asr_model_valid.acc.best/test-other_3spk|2453|1381987|76.0|8.9|15.1|9.2|33.2|100.0|
|
55 |
+
|decode_sot_asr_model_valid.acc.best/test-other_4spk|1795|1346646|66.2|11.3|22.4|7.7|41.5|100.0|
|
56 |
+
|decode_sot_asr_model_valid.acc.best/test_clean_2spk_kaldi_fmt|2180|921344|87.3|2.8|9.9|5.1|17.7|95.0|
|
57 |
+
|decode_sot_asr_model_valid.acc.best/test_other_2spk_kaldi_fmt|2363|1064868|86.9|3.8|9.3|5.8|18.9|98.1|
|
58 |
+
|decode_sot_asr_model_valid.acc.best/tt_mix_clean_reverb_max_16k|3000|143026|16.3|83.6|0.2|299.2|382.9|100.0|
|
59 |
+
|
60 |
+
### TER
|
61 |
+
|
62 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
63 |
+
|---|---|---|---|---|---|---|---|---|
|