WizardLM VS Mage

This is interesting work, and a good contribution, but there is no need to mislead people.

[1] https://github.com/nlpxucan/WizardLM

If you pair this with the latest WizardCoder models, which have a fairly better performance than the standard Salesforce Codegen2 and Codegen2.5, you have a pretty solid alternative to GitHub Copilot that runs completely locally.

(We will update the demo links in our github.)

In terms of speed, we're talking about 140t/s for 7B models, and 40t/s for 33B models on a 3090/4090 now.[1] (1 token ~= 0.75 word) It's quite zippy. llama.cpp performs close on Nvidia GPUs now (but they don't have a handy chart) and you can get decent performance on 13B models on M1/M2 Macs.

You can take a look at a list of evals here: https://llm-tracker.info/books/evals/page/list-of-evals - for general usage, I think home-rolled evals like llm-jeopardy [2] and local-llm-comparison [3] by hobbyists are more useful than most of the benchmark rankings.

That being said, personally I mostly use GPT-4 for code assistance to that's what I'm most interested in, and the latest code assistants are scoring quite well: https://github.com/abacaj/code-eval - a recent replit-3b fine tune the human-eval results for open models (as a point of reference, GPT-3.5 gets 60.4 on pass@1 and 68.9 on pass@10 [4]) - I've only just started playing around with it since replit model tooling is not as good as llamas (doc here: https://llm-tracker.info/books/howto-guides/page/replit-mode...).

I'm interested in potentially applying reflexion or some of the other techniques that have been tried to even further increase coding abilities. (InterCode in particular has caught my eye https://intercode-benchmark.github.io/)

[1] https://github.com/turboderp/exllama#results-so-far

[2] https://github.com/aigoopy/llm-jeopardy

[3] https://github.com/Troyanovsky/Local-LLM-comparison/tree/mai...

[4] https://github.com/nlpxucan/WizardLM/tree/main/WizardCoder

You talking about this? https://github.com/nlpxucan/WizardLM

The smallest model that is close to competent at code is WizardCoder 15B.. https://github.com/nlpxucan/WizardLM/

WizardCoder: Empowering Code Large Language Models with Evol-Instruct (https://github.com/nlpxucan/WizardLM/tree/main/WizardCoder)

In the coming years, Mage will create a cooperative experience so that developers can build data pipelines with their team and level up together. After that journey, Mage will go on an epic quest to create the 1st open world community experience in the data universe.

Link to original blog: https://www.mage.ai/blog/data-sources-ep-2-aws-s3-to-postgres-data-sync-using-singer

I would recommend the following: - https://www.mage.ai/ - https://dagster.io/ - https://www.prefect.io/ - https://metaflow.org/ - https://zenml.io/home

You're invited to participate in the very first Mage Battlegrounds: Craft insights from real-time customer behavior analysis, a 24-hour virtual hackathon hosted by Shashank Mishra! This data engineering competition will take place on Saturday, April 15, 2023 beginning at 11am (PST). This will be a global event open to all participants who register.

Try this feature: https://github.com/mage-ai/mage-ai/issues/1166

Seeing if we can use it in https://github.com/mage-ai/mage-ai

That being said there’s a solid family of services just breaking ground that make the local pipeline deployment easier (check out https://www.mage.ai, which does have a clear path to cloud deployment of locally developed pipes, it just isn’t well documented yet, and also https://www.neuronsphere.io - which doesn’t have a public solution YET (they’re internally testing an alpha) but they built a cloud deployable solution for their paying customers and working to release one for freemium use)

7️⃣ mage-ai/mage-ai

Yes, check out how Mage does it: https://github.com/mage-ai/mage-ai/tree/master/mage_integrations/mage_integrations/destinations/delta_lake_s3