So by popular demand, we've starting exploring the finetuning of BigCode's recently released code-completion model StarCoder (
https://
x.com/BigCodeProject
/status/1654174941976068119
…) to be a…. Chat Assistant We finetuned it on two high-quality datasets created by the community:
– OpenAssistant’s
@thom_wolf
-
StarCoder Fine-tuned as Chat Assistant Using Community Datasets
By
–
-
Data licensing and commercial AI use ethics
By
–
"having all the data" doesn't mean that the people who wrote the data agreed for you to use it commercially and without crediting, e.g. GPL licences
-
Waiting for GPT-4 detailed specifications and capabilities
By
–
Waiting to read the same details about your GPT4
-
Open-Source Research Project Shares Full Paper and Community Access
By
–
Yes, you can find all the details and ablations in the full 54 pages paper it-self: https://
drive.google.com/file/d/1cN-b9G
nWtHzQRoE7M7gAEyivY0kl4BYs/view
… And more on the project open slack (open access projects FTW!) at https://
bigcode-project.org/docs/about/joi
n/
… (probably more than you even wanted to read ) -
Open Source AI Model Alternative to OpenAI Integration
By
–
It’s an alternative if you’re interested in a model (1) which is only trained on clean data and/or (2) that you can use internally without having to send your data to OpenAI. Here is an example of integration in vscode: https://
marketplace.visualstudio.com/items?itemName
=HuggingFace.huggingface-vscode
… -
StarCoder 15B Model Released: High Performance Open Source Code Generation
By
–
So this week we've finally released StarCoder (
https://
huggingface.co/bigcode/starco
der
…) with @BigCodeProject StarCoder is the first large model (15B) which is both high performance (beating the like of PaLM, LLaMa, CodeGen or OpenAI code-crushman-001 on code generation) and also trained -
Closed-source AI models copyright compliance debate
By
–
I know you don’t talk a lot about them but I’d be curious to know your stance on closed-source models one day Alex Like where do you think models that are being integrated everywhere today like GPT-4, Bard or Claude are standing on the scale of copyright compliance?
-
BigCode announces upcoming Swift large language model
By
–
It’s only the first large model of BigCode, we’ll have Swift at some point , don’t worry 🙂
-
StarCoder: Access Code AI Model via Playground and Extensions
By
–
The best place to try StarCoder is the playground: https://
huggingface.co/spaces/bigcode
/bigcode-playground
… And then you can continue with the vscode extension, the huggingchat interface, or just download it and play locally. All the links in the blog: https://
huggingface.co/blog/starcoder -
StarCoder: Exciting Developments in Open Source Code Generation
By
–
There are a lot of things very very exciting about StarCoder