Tatako Felici

tatakof

AI & ML interests

Transformers, Reinforcement Learning, Differentiable Programming

Recent Activity

updated a model 14 days ago

sandbox-ai/Llama-3.1-Tango-70b-GGUF

updated a model 24 days ago

sandbox-ai/Llama-3.1-Tango-70b

liked a model 28 days ago

sandbox-ai/Llama-3.1-Tango-70b

View all activity

Organizations

tatakof's activity

updated a model 14 days ago

sandbox-ai/Llama-3.1-Tango-70b-GGUF

Updated 13 days ago • 78

updated a model 24 days ago

sandbox-ai/Llama-3.1-Tango-70b

Text Generation • Updated 17 days ago • 97 • 6

liked a model 28 days ago

sandbox-ai/Llama-3.1-Tango-70b

Text Generation • Updated 17 days ago • 97 • 6

updated 2 datasets about 1 month ago

tatakof/messi_mod-v0.0.2

Viewer • Updated Oct 20 • 15.7k • 45

tatakof/messi_mod-v0.0.1

Viewer • Updated Oct 19 • 395k • 46

liked a model 3 months ago

meta-llama/Llama-3.1-70B-Instruct

Text Generation • Updated Sep 25 • 3.23M • • 694

updated a model 6 months ago

tatakof/distillbert-base-spanish-uncased_finetuned_with-Llama2-Knowledge-Distillation

Fill-Mask • Updated May 27 • 18

liked a model 7 months ago

RWKV/v5-Eagle-7B-HF

Text Generation • Updated Feb 25 • 8.47k • 71

Reacted to ordagan's post with 🔥 8 months ago

Post

2161

Excited to introduce Jamba by AI21
ai21labs/Jamba-v0.1

We are thrilled to announce Jamba, the world’s first production-grade Mamba based model.

Key Features:
- First production-grade Mamba based model built on a novel SSM-Transformer hybrid architecture
- 3X throughput on long contexts compared to Mixtral 8x7B
- Democratizes access to a massive 256K context window
- The only model in its size class that fits up to 140K context on a single GPU

Jamba is based on a novel architecture that combines Mamba and Transformer. While our initial results show great efficiency gains, we expect this to be further explored and improved with the help of the community.

Check out our blog post for more info: https://ai21-labs.webflow.io/blog/announcing-jamba