Alan Dao

alandao's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes

updated a model 10 days ago

homebrewltd/Ichigo-whisper-v0.1

Audio-Text-to-Text • Updated 10 days ago • 1 • 14

upvoted a paper 18 days ago

Parallelized Autoregressive Visual Generation

Paper • 2412.15119 • Published 24 days ago • 50

updated 2 models about 2 months ago

alandao/f5-tts-mlx-4bit

Updated Nov 21, 2024 • 1

alandao/f5-tts-mlx-8bit

Updated Nov 21, 2024 • 1

liked a model about 2 months ago

bartowski/Ichigo-llama3.1-s-instruct-v0.4-GGUF

Text Generation • Updated Nov 11, 2024 • 50 • 2

upvoted a collection 2 months ago

OpenCoder Datasets

Collection

OpenCoder datasets! • 6 items • Updated Nov 15, 2024 • 38

commented a paper 3 months ago

Ichigo: Mixed-Modal Early-Fusion Realtime Voice Assistant

Paper • 2410.15316 • Published Oct 20, 2024 • 10 •

4

upvoted a paper 3 months ago

Ichigo: Mixed-Modal Early-Fusion Realtime Voice Assistant

Paper • 2410.15316 • Published Oct 20, 2024 • 10

authored a paper 3 months ago

Ichigo: Mixed-Modal Early-Fusion Realtime Voice Assistant

Paper • 2410.15316 • Published Oct 20, 2024 • 10

upvoted a collection 3 months ago

🍓 Ichigo v0.3

Collection

The experimental family designed to train LLMs to understand sound natively. • 6 items • Updated Nov 11, 2024 • 17

updated a model 3 months ago

homebrewltd/mini-Ichigo-llama3.2-3B-s-instruct

Audio-Text-to-Text • Updated Nov 19, 2024 • 47 • 34

updated a collection 3 months ago

🍓 Ichigo v0.3

Collection

The experimental family designed to train LLMs to understand sound natively. • 6 items • Updated Nov 11, 2024 • 17

reacted to reach-vb's post with 😎 3 months ago

view post

Post

5474

Multimodal Ichigo Llama 3.1 - Real Time Voice AI 🔥

> WhisperSpeech X Llama 3.1 8B
> Trained on 50K hours of speech (7 languages)
> Continually trained on 45hrs 10x A1000s
> MLS -> WhisperVQ tokens -> Llama 3.1
> Instruction tuned on 1.89M samples
> 70% speech, 20% transcription, 10% text
> Apache 2.0 licensed ⚡

Architecture:
> WhisperSpeech/ VQ for Semantic Tokens
> Llama 3.1 8B Instruct for Text backbone
> Early fusion (Chameleon)

I'm super bullish on HomeBrew/ Jan and early fusion, audio and text, multimodal models!

(P.S. Play with the demo on Hugging Face: jan-hq/Ichigo-llama3.1-s-instruct)

🔥

16

16

👍

5

5

❤️

2

2

😎

1

1

👀

1

1

🚀

1

1

+

Reply

liked a Space 5 months ago

Running on Zero

111

😻

Llama3.1 S V0.2 Checkpoint 2024 08 20

liked a dataset 5 months ago

homebrewltd/instruction-speech-encodec-v1

Viewer • Updated Aug 19, 2024 • 493k • 699 • 14

updated 2 datasets 5 months ago

alandao/raw_audio_with_audio_tokens_for_pretraining_just_tokens

Viewer • Updated Aug 10, 2024 • 2.42M • 38

alandao/viet_poems

Viewer • Updated Aug 2, 2024 • 3.36k • 33

updated a model 6 months ago

alandao/dummy_sound_instruct_init

Text Generation • Updated Jul 6, 2024 • 24