README.md

---
license: apache-2.0
datasets:
- mozilla-foundation/common_voice_16_1
- openslr/librispeech_asr
language:
- en
metrics:
- wer
library_name: transformers
model-index:
- name: SpeechLLM
  results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: LibriSpeech (clean)
          type: librispeech_asr
          config: clean
          split: test
          args:
            language: en
        metrics:
          - name: Test WER
            type: wer
            value: 12.3
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: LibriSpeech (other)
          type: librispeech_asr
          config: other
          split: test
          args:
            language: en
        metrics:
          - name: Test WER
            type: wer
            value: 18.9

      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice 16.1
          type: common_voice_16_1
          split: test
          args:
            language: en
        metrics:
          - name: Test WER
            type: wer
            value: 25.01
---

# SpeechLLM

SpeechLLM is a multi-modal LLM trained to predict the metadata of the speaker's turn in a conversation. SpeechLLM model is based on HubertX acoustic encoder and TinyLlama LLM. The model predicts the following:
1. **SpeechActivity** : if the audio signal contains speech (True/False)
2. **Transcript** : ASR transcript of the audio
3. **Gender** of the speaker (Female/Male)
4. **Age** of the speaker (Young/Middle-Age/Senior)
5. **Accent** of the speaker (Africa/America/Celtic/Europe/Oceania/South-Asia/South-East-Asia)
6. **Emotion** of the speaker (Happy/Sad/Anger/Neutral/Frustrated)

## Usage
```python
# Load model directly from huggingface
from transformers import AutoModel
model = AutoModel.from_pretrained("skit-ai/SpeechLLM", trust_remote_code=True)

model.generate_meta(
	audio_path="path-to-audio.wav", 
	instruction="Give me the following information about the audio [SpeechActivity, Transcript, Gender, Emotion, Age, Accent]",
	max_new_tokens=500, 
	return_special_tokens=False
)

# Model Generation
'''
{
  "SpeechActivity" : "True",
  "Transcript": "Yes, I got it. I'll make the payment now.",
  "Gender": "Female",
  "Emotion": "Neutral",
  "Age": "Young",
  "Accent" : "America",
}
'''
```

## Model Details

- Model Size : 2.1 B
- Checkpoint : 2000 k steps (bs=1)
- Adapters : r=4, alpha=8
- lr = 1e-4


## Checkpoint Result

|         Dataset        | Word Error Rate    | Gender Acc|
|:----------------------:|:------------------:|:---------:|
| librispeech-test-clean | 0.1230             | 0.8778    |
| librispeech-test-other | 0.1890             | 0.8908    |
| CommonVoice test       | 0.2501             | 0.8753    |