File size: 1,824 Bytes
5144da6 fca4f90 5144da6 fca4f90 5144da6 fca4f90 5144da6 fca4f90 aea7098 5144da6 fca4f90 5144da6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 |
---
tags:
- generated_from_trainer
model-index:
- name: prosody_gttbsc_distilbert-uncased-best
results:
- task:
type: dialogue act classification
dataset:
name: asapp/slue-phase-2
type: hvb
metrics:
- name: F1 macro E2E
type: F1 macro
value: 66.43
- name: F1 macro GT
type: F1 macro
value: 72.70
datasets:
- asapp/slue-phase-2
language:
- en
metrics:
- f1-macro
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# prosody_gttbsc_distilbert-uncased-best
Ground truth text with prosody encoding cross attention multi-label DAC
## Model description
Prosody encoder: 2 layer transformer encoder with initial dense projection
Backbone: [DistilBert uncased](https://huggingface.co/distilbert/distilbert-base-uncased)
Pooling: Self attention
Multi-label classification head: 2 dense layers with two dropouts 0.3 and Tanh activation inbetween
## Training and evaluation data
Trained on ground truth [slue-phase-2 hvb](https://huggingface.co/datasets/asapp/slue-phase-2).
Evaluated on ground truth (GT) and normalized [Whisper small](https://huggingface.co/openai/whisper-small) transcripts (E2E).
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0004
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
### Framework versions
- Transformers 4.41.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1
|