mirror of
https://github.com/k2-fsa/icefall.git
synced 2025-08-09 01:52:41 +00:00
add small-scaled model for audio tagging (#1604)
This commit is contained in:
parent
368b7d10a7
commit
df36f93bd8
@ -5,6 +5,8 @@ See <https://github.com/k2-fsa/icefall/pull/1421> for more details
|
|||||||
|
|
||||||
[zipformer](./zipformer)
|
[zipformer](./zipformer)
|
||||||
|
|
||||||
|
#### normal-scaled model, number of model parameters: 65549011, i.e., 65.55 M
|
||||||
|
|
||||||
You can find a pretrained model, training logs, decoding logs, and decoding results at:
|
You can find a pretrained model, training logs, decoding logs, and decoding results at:
|
||||||
<https://huggingface.co/marcoyang/icefall-audio-tagging-audioset-zipformer-2024-03-12#/>
|
<https://huggingface.co/marcoyang/icefall-audio-tagging-audioset-zipformer-2024-03-12#/>
|
||||||
|
|
||||||
@ -42,3 +44,52 @@ python zipformer/evaluate.py \
|
|||||||
--exp-dir zipformer/exp_at_as_full \
|
--exp-dir zipformer/exp_at_as_full \
|
||||||
--max-duration 500
|
--max-duration 500
|
||||||
```
|
```
|
||||||
|
|
||||||
|
|
||||||
|
#### small-scaled model, number of model parameters: 22125218, i.e., 22.13 M
|
||||||
|
|
||||||
|
You can find a pretrained model, training logs, decoding logs, and decoding results at:
|
||||||
|
<https://huggingface.co/marcoyang/icefall-audio-tagging-audioset-zipformer-small-2024-04-23#/>
|
||||||
|
|
||||||
|
The model achieves the following mean averaged precision on AudioSet:
|
||||||
|
|
||||||
|
| Model | mAP |
|
||||||
|
| ------ | ------- |
|
||||||
|
| Zipformer-S-AT | 45.1 |
|
||||||
|
|
||||||
|
The training command is:
|
||||||
|
|
||||||
|
```bash
|
||||||
|
export CUDA_VISIBLE_DEVICES="4,5,6,7"
|
||||||
|
subset=full
|
||||||
|
|
||||||
|
python zipformer/train.py \
|
||||||
|
--world-size 4 \
|
||||||
|
--num-epochs 50 \
|
||||||
|
--exp-dir zipformer/exp_small_at_as_${subset} \
|
||||||
|
--start-epoch 1 \
|
||||||
|
--use-fp16 1 \
|
||||||
|
--num-events 527 \
|
||||||
|
--num-encoder-layers 2,2,2,2,2,2 \
|
||||||
|
--feedforward-dim 512,768,768,768,768,768 \
|
||||||
|
--encoder-dim 192,256,256,256,256,256 \
|
||||||
|
--encoder-unmasked-dim 192,192,192,192,192,192 \
|
||||||
|
--audioset-subset $subset \
|
||||||
|
--max-duration 1200 \
|
||||||
|
--enable-musan True \
|
||||||
|
--master-port 13455
|
||||||
|
```
|
||||||
|
|
||||||
|
The evaluation command is:
|
||||||
|
|
||||||
|
```bash
|
||||||
|
python zipformer/evaluate.py \
|
||||||
|
--epoch 31 \
|
||||||
|
--avg 4 \
|
||||||
|
--num-encoder-layers 2,2,2,2,2,2 \
|
||||||
|
--feedforward-dim 512,768,768,768,768,768 \
|
||||||
|
--encoder-dim 192,256,256,256,256,256 \
|
||||||
|
--encoder-unmasked-dim 192,192,192,192,192,192 \
|
||||||
|
--exp-dir zipformer/exp_small_at_as_full \
|
||||||
|
--max-duration 500
|
||||||
|
```
|
Loading…
x
Reference in New Issue
Block a user