Thanks for sharing it, however I have an unrelated comment.
Maybe I am in minority here but just wanted to provide this feedback: The background animation of the blog page is really distracting and making it difficult to focus on the actual content.
I am also using conda and specifically mamba which has a really quick dependency solver.
However, sometimes repos require system level packages as well. Tried to run TRELLIS recently and gave up after 2h of tinkering around to get it to work in Windows.
Also, whenever I try to run some new repo locally, creating a new virtual environment takes a ton of disk space due to CUDA and PyTorch libraries. It adds up quickly to 100s of gigs since most projects use different versions of these libraries.
</rant> Sorry for the rant, can't help myself when it's Python package management...
I did not know that cleanrooms have classes. Apparently they used a 10000 class cleanroom which is one of the "dirtiest" grade.
Surely an astroid sample return mission is scientifically one of the most important and difficult things to accomplish, and I wonder why they did not use a higher class cleanroom for this even though they mentioned "Researchers recommend enhanced contamination control procedures for future sample-return missions to prevent microbial colonization and ensure the integrity of extraterrestrial samples."
Agreed on providing examples is definitely a useful insight vs fine-tuning.
While it is not very important for this toy case, it's good to keep in mind that each provided example in the input will increase the prediction time and cost compared to fine-tuning.
To me it boils down to what is to be measured here. With logprobs we can measure both correctness and not attempted i.e. if LLM is guessing the response.
Similar to exams where both the progress to the solution and the final outcome/value of the calculations are part of the grade.
I am puzzled why they have "asked the model" about the confidence and have not used the logprobs of the output tokens to estimate the confidence in responses.
In my use case and tests, model itself is not capable of giving a reliable confidence value where logprobs almost always provide a better view on calibration.
I would be extremely surprised if that's the case. There are "open-source" multimodal LLMs can extract text from images as a proof that the idea works.
Probably the model is hallucinating and adding "Hungarian language is not installed for Tesseract" to the response.
I am really puzzled how are they able to "quietly" update the terms without notifying their users? Everybody was joking about the emails (We have updated our terms...) raining from every company when GDPR et al. got introduced. What changed?
I am also mostly using VSCode for Notebooks however one big downside is the performance goes down the drain when the notebooks are large in size (i.e. Contains images, plots).
The cell execution speed drastically (up to 10x) slows down w.r.t. Notebook file size. Still looking for a solution to this problem.
20 ppb listed in the EU document is for animal feed materials, not for the actual food. I
EU:
> For maize and all products derived from maize, including processed maize products, maximum limits (μg/kg) of 2,0 ppb for Aflatoxin B1 and 4,0 ppb for sum of Aflatoxins
B1+B2+G1+G2 are set
> For Food (incoming material): Maximum limits (μg/kg) of 5,0 ppb for Aflatoxin B1 and 10,0 ppb for sum of Aflatoxins B1+B2+G1+G2.
> For feed: Feed materials 20 ppb for Aflatoxin B1
They are not noticeable under visible light, so contaminated figs look totally normal except under UV and/or IR light. You can check out the link (2) above or this one:
On a side note, try to keep eating figs in dried form at a moderate amount or avoid altogether if the source is "not reputable".
Dried figs are very susceptible to Aflatoxin B1(1) which is a very potent carcinogen from fungi. US food safety regulations allow 2-10 times more aflatoxin B1 in food compared to EU.
During my work, I had a chance to visit dried fig producers and saw even a couple of contaminated figs spreading to the rest of the stock like wildfire.
What you can do is to check your dried figs under UV light, and it should not shine. Here's an example image(2) I found.
Source: I have worked in the company as a machine vision engineer to develop aflatoxin detection systems with UV light.
Yes, but it is in their interest to monetize this.
Many people I talked with are fully willing to pay for it after it goes behind a paywall. OpenAI improved the user experience of their Playground with ChatGPT and not so many people taking out the wallet for DALL-E 2. It is fair to mention that the free alternatives of DALL-E 2 are equally capable, if not better.
This was a fantastic read! Seeing the German language, that is also foreign to me, and its learning process explained by a master writer, gave me a great new perspective and understanding. Many topics that I felt not completely understood just fits into the place now.
Maybe I am in minority here but just wanted to provide this feedback: The background animation of the blog page is really distracting and making it difficult to focus on the actual content.