🥃 Distilling Tiny Embeddings
•
16
AI & ML interests
Applying machine learning to solve everyday problems
Recent Activity
Articles
davidmezzetti
posted
an
update
1 day ago
Post
131
Do you like to "think" in Jupyter Notebooks? Want to integrate OpenCode with that? Then check this project out.
NCoder is an open-source AI coding agent that connects a local running OpenCode server to a Jupyter Notebook via TxtAI.
https://github.com/neuml/ncoder
NCoder is an open-source AI coding agent that connects a local running OpenCode server to a Jupyter Notebook via TxtAI.
https://github.com/neuml/ncoder
davidmezzetti
posted
an
update
6 days ago
Post
1172
Tired of paying a bunch of money for large LLMs? Then read this notebook and/or watch this video on how to distill knowledge into tiny LLMs.
Notebook: https://github.com/neuml/txtai/blob/master/examples/80_Distilling_Knowledge_into_Tiny_LLMs.ipynb
Video: https://www.youtube.com/watch?v=Ol560ktgkf0
Notebook: https://github.com/neuml/txtai/blob/master/examples/80_Distilling_Knowledge_into_Tiny_LLMs.ipynb
Video: https://www.youtube.com/watch?v=Ol560ktgkf0
-
- 6 replies
davidmezzetti
updated
a
Space
11 days ago
davidmezzetti
updated
a
dataset
11 days ago
davidmezzetti
posted
an
update
12 days ago
Post
2179
🥃 Distilling Tiny Embeddings. We're happy to build on the BERT Hash Series of models with this new set of fixed dimensional tiny embeddings models.
Ranging from 244K parameters to 970K and 50 dimensions to 128 dimensions these tiny models pack quite a punch.
Use cases include on-device semantic search, similarity comparisons, LLM chunking and Retrieval Augmented Generation (RAG). The advantage is that data never needs to leave the device while still having solid performance.
https://huggingface.co/blog/NeuML/bert-hash-embeddings
Ranging from 244K parameters to 970K and 50 dimensions to 128 dimensions these tiny models pack quite a punch.
Use cases include on-device semantic search, similarity comparisons, LLM chunking and Retrieval Augmented Generation (RAG). The advantage is that data never needs to leave the device while still having solid performance.
https://huggingface.co/blog/NeuML/bert-hash-embeddings
davidmezzetti
published
an
article
12 days ago
Article
🥃 Distilling Tiny Embeddings
•
16
davidmezzetti
updated
3
models
12 days ago
davidmezzetti
published
3
models
12 days ago
davidmezzetti
updated
a
model
17 days ago
davidmezzetti
published
a
model
17 days ago
davidmezzetti
published
a
dataset
29 days ago
davidmezzetti
updated
2
models
about 1 month ago
davidmezzetti
updated
a
collection
about 1 month ago