Downloads · 30 days
5
26% of all-time downloads
lf2ar-speech-360m
tiagoCuervo/lf2ar-speech-360m
lf2ar-speech-360m is a machine learning model from tiagoCuervo. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for interleaved-lm. The card lists the license as apache-2.0.
Speech LF²AR checkpoint from LF²AR: Accounting for Layerwise Dynamics to Improve Multimodal Adaptation of Language Models. This model has 401,665,984 parameters and uses dynamic attention residuals.
All-time downloads
19
Public
Parameters
402M
1.6 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.6 GB · 100%
At a glance
- Library
- interleaved-lm
- License
- apache-2.0
- Model type
- interleaved-lm
- Access
- Public
- Created
- Aug 11, 2026
- Updated
- Aug 11, 2026
- SHA
- 1703ec78
Model card
From the Hugging Face model README
lf2ar-speech-360m
Speech LF²AR checkpoint from LF²AR: Accounting for Layerwise Dynamics to Improve Multimodal Adaptation of Language Models. This model has 401,665,984 parameters and uses dynamic attention residuals.
from interleaved_lm import PerceptionExpressionAdaptedTextLM
from interleaved_lm.audio import generate_speech
model = PerceptionExpressionAdaptedTextLM.from_pretrained(
"tiagoCuervo/lf2ar-speech-360m", device="cuda"
)
codes = generate_speech(model, "A short story about the moon", max_new_tokens=250)
Speech is represented by consecutive-run-collapsed 25 Hz mHuBERT layer-11 K-means units (K=500; EOS=500; PAD=501). A vocoder is not bundled.
The text backbone is initialized from SmolLM. Speech-tokenizer and vocoder artifacts are credited in the Interleaved-LM third-party notices.
Evaluation
| Task | Direction | Accuracy |
|---|---|---|
| sstorycloze | audio-audio | 57.19 |
| sstorycloze | text-text | 68.63 |
| sstorycloze | audio-text | 61.89 |
| sstorycloze | text-audio | 62.27 |
| tstorycloze | audio-audio | 84.61 |
| tstorycloze | text-text | 91.39 |
| tstorycloze | audio-text | 84.98 |
| tstorycloze | text-audio | 80.81 |
See provenance.json for exact hashes and evaluation metadata.
- Library
- interleaved-lm
- Type
- interleaved-lm
- License
- apache-2.0
- Created
- Aug 11, 2026
- Updated
- Aug 11, 2026