You can deploy this model and start earning money today!

jaswanth99/
Aelvo-4b

Name: Aelvo-4b
Availability: InStock

pytorch

aelvo4b

Views

No views yet

<<<<<<< HEAD

🧠 Aelvo-4B: Unified Multimodal Transformer

Aelvo-4B is a multimodal model integrating Vision, Speech, Text, and TTS in a unified 4B-parameter architecture.
Trained on multimodal datasets (Flickr8k + LJSpeech + text corpora).

🧩 Model Structure

Vision Encoder: 24 Transformer layers (hidden 2048, 16 heads)
Speech Encoder: 24 Transformer layers
Text LLM Core: 24 Transformer layers
TTS Decoder: 14 Transformer layers
Fusion Layer: Unified embedding (2048-d)

🚀 Usage

python

from transformers import AutoModelForCausalLM, BitsAndBytesConfig

# Load normally
model = AutoModelForCausalLM.from_pretrained("jaswanth/aelvo4b")

# Or load in INT4
bnb_config = BitsAndBytesConfig(load_in_4bit=True)
model = AutoModelForCausalLM.from_pretrained("jaswanth/aelvo4b", quantization_config=bnb_config)
=======
---
license: apache-2.0
---
>>>>>>> 6515ba91bd1d691a597d926ea5203a66ad2db50f

jaswanth99/Aelvo-4b

🧠 Aelvo-4B: Unified Multimodal Transformer

🧩 Model Structure

🚀 Usage

jaswanth99/
Aelvo-4b