HN user
An agent is just an LLM calling tools in a loop. If you're a "show me the code" type person like me, here's a worked example: https://samdobson.uk/posts/how-to-build-an-agent/
Great call, thankyou - I switched to log scale for those metrics - agree that it is much clearer.
This is a much easier way to run the model. I'm going to update the huggingface README to point to this. The one thing that could be improved is the turn-taking between user and assistant, which it sometimes gets confused about. I fixed that in my fork of your gist here: https://gist.github.com/samdobson/975c8b095a71bbdf1488987eac...
I've uploaded the model here: https://huggingface.co/sdobson/nanochat
I didn't get as good results as Karpathy (unlucky seed?)
It's fun to play with though...
User: How many legs does a dog have? Assistant: That's a great question that has been debated by dog enthusiasts for centuries. There's no one "right" answer (...)
There's not much to it... it took longer to spin up the cloud machine than it did to kick off the training run. I'll be writing up a blog post with a step-by-step guide when I get a free moment, but in the meantime, here are the commands I ran: https://pastebin.com/sdKVy0NR
I'm doing a training run right now (started 20min ago). You can follow it at https://api.wandb.ai/links/sjd333-none/dsv4zkij
Will share the resulting model once ready (4 hours from now) for anyone to test inference.
That's probably because fine-tuning is not the right approach for this use-case. A better approach might look more like this: https://github.com/openai/openai-cookbook/blob/main/examples...
Hi HN,
This one is mine. It's a light-hearted digital newspaper of sorts, covering news from local British communities through the medium of verse (generated by LLMs).
Until now I've been using ChatGPT for the generation, with a fairly generic prompt that asks for a poem about the article that follows. ChatGPT's ability to summarise is incredible. It's really not great, though, at rhyme and meter. That means a decent amount of curation and heavy editing is needed for the best to get something passable. Prompt engineering has not seemed to have a meaningful impact. I'm looking to fine-tune a davinci model, which I think will deliver higher quality with less effort.
Some example from the current process:
Poem: https://rhymingreporter.art/farewell-little-red/ | Original article: https://www.cornwalllive.com/whats-on/food-drink/little-red-...
Poem: https://rhymingreporter.art/flowing-frocks-icy-blue/ | Original article: https://www.cornwalllive.com/whats-on/whats-on-news/gallery/...
The quality can mostly be blamed on me, rather than GPT-3. I haven't written a poem since school :)
The accompanying illustrations are created with Stable Diffusion using DiffusionBee. Images take around 30s to generate on my Macbook Air M1. I'm looking to switch to MochiDiffusion to cut generation time a bit.
The blog is running Ghost on a small DigitalOcean VPS, with emails delivered by Mailgun.
The process right now is somewhat labour-intensive: between researching news stories, iterating on the content, and publishing, it takes a decent amount of time for each piece of content. I'm confident in being able to automate a large part of it, in time.
One fun fact I learned when planning the virtual road-trip for this project: in average traffic conditions, it's possible to visit every city in England in less than 48 hours. The near-optimum solution to this formulation of the Traveling Salesmen Problem (starting in the South West), a route taking 47:00:10, was calculated in less than 5 seconds with a Guided Local Search algorithm. [1]
Technology means that I can virtually, learn about, write creatively, and publish regularly, all whilst having a family and a full-time job. What a time to be alive!
Very open to your thoughts, and indeed to feedback on the concept or the execution.
What is your goal?
For short-term career growth, $YOUR_COMPANY's current preferred ETL tool will have the biggest ROI. Focus on design patterns: while APIs will come and go, the concepts, as you rightly say, are transferrable.
If you're looking to land a new role: the market says dbt, databricks and snowflake are pretty strong bets.
If it's personal interest, or a high-risk, high-reward long term play, take your pick from any of the new hotness!
A stark reminder to end users of any service that if you are not paying, you are the product. There is good reason that the sale of financial services through intermediaries is a highly regulated area: it takes a firm counter-force to stop businesses from exploiting customers when all the incentives are there, and there are enough links in the chain for plausible deniability.
Very interesting concept. A couple of initial observations:
* Creating models from a file borks on anything non-UTF8 (i.e. most legacy system outputs)
* `synth model inspect` output does not match the docs - how do I see the JSON?
This looks really cool and haven't seen this before - thanks for sharing. I've updated the description on Kaggle.
Hi HN. Over the last 4 months I have collected 380,000 attempts to guess a number on my Amazon Alexa Skill, and wanted to share with the community. Would love to see what you can do with the data!
Thanks for the tip :) Not sure about the licence - think it was the default option. I'll take a look.
edit: Updated licence to CC-BY-SA (i.e. do what you want as long as you credit)
edit2: Don't seem to be able to re-post :(
Big 4 management consultant in the Financial Services industry, exploring opportunities in Tech. 15 years of coding experience, with GitHub portfolio dating back 5+.
Location: London, UK
Remote: Yes
Willing to relocate: No
Technologies: Python, SQL, Javascript, AWS, Linux
Résumé/CV: on request
Email: sjd followed by three threes (gmail)
Here you go! No need to fork :)
https://samdobson.github.io/stocks/?Banks=GS,MS,JPM&Tech=AAP...
For me 'sharing' is about communicating, about passing on a message. A speech bubble is the best representation in my mind... but that is already used for comments.
Most Germanic languages have constructs similar to English´s phrasal verbs. Learning them, for a foreigner, is really difficult (I know, because I spend most of my time teaching English to Spanish students). I actually just released the Android app I´ve been building to help with this. I didn´t think it was worthy of a Show HN post, but now seems a good time to mention it: https://play.google.com/store/apps/details?id=com.phrasalver...
"Because of my experiences growing up, I have triggers. This means that I’m always scanning for danger; for situations that seem like something from the past that could hurt me. When I recognize something that matches, I can overreact and feel intense fear, anger or anxiety. This is something I’ve worked on a lot. It’s much better now than 10 years ago but there are some things that send me over the edge."
http://butyoureagirl.com/13871/success-against-the-odds-fill...
(site is currently down, may need to view cache)
I'm surprised nobody has picked up on this yet.
"An entire generation of teachers and administrators is embracing the belief that technology will allow teachers and students to work together more effectively, to research more broadly, to share more generously, and to learn more passionately."
As a teacher and member of that generation, thankyou.
Are there any other awesome product like lucidchart available free of charge to K-12 schools?