gillesjacobs

9 ms·

61.
by gillesjacobs 2y ago
Yeah exactly, existing benchmark datasets available are underutilized (eg KILT, Natural questions, etc.). But it is only natural that different QA use cases require different strategies. I built 3 production RAG systems / virtual assis
62.
BGBye: FOSS Image Background Remover with 10 SotA Models(fyrean.itch.io)
2 points by gillesjacobs 2y ago|1 comments
63.
by gillesjacobs 2y ago
Source: https://github.com/MangoLion/bgbye
64.
by gillesjacobs 2y ago
Poetry worked best in several Python-first companies and clients I have worked for. It's not perfect but the best bet for complex and collaboration-heavy projects.
65.
by gillesjacobs 2y ago
ML engineer, I am up to 3 irreversibly broken environments after simply adding a package. Conda is inexcusably slow and but even mamba can save it. Every conda project I have ever was later switched to poetry. Here's a fun challenge: t
66.
by gillesjacobs 2y ago
The cited reason given is > An investigation by the publisher found a number of articles, including this one, with a number of concerns, including but not limited to compromised editorial handling and peer review process, inappropriate o
67.
by gillesjacobs 2y ago
Exactly, BloombergGPT performed worse on financial sentiment analysis then much smaller fine-tuned Bert-based models. For many extractive tasks BloombergGPT was quite disappointing. A 5-10% performance hit with much larger inference cost co
68.
by gillesjacobs 2y ago
This is indeed often the case, however, my university reviews each thesis, and deemed it can only change to open access in 2026 (+5 years from defense). I think this is default policy for thesis based on publication agreements here. In any
69.
by gillesjacobs 2y ago
Thanks! The fine-tunes where done in 2019-21 on a 4xV100 server with hyperparameter search, so thousands of individual fine-tuned models were trained in the end. I used weights and biased for experiment dashboarding the hyperparam search, b
70.
by gillesjacobs 2y ago
I use LLM information extraction for financial news articles with OpenAI Azure and it is a huge problem for me. 404 Content moderation response in 4% of articles. This is just financial news text. It is a prime reason we are considering ope
71.
by gillesjacobs 2y ago
rovr beat me to it below. Here are more links: https://jacobsgill.es/phdobtained (fun fact: because my thesis contains published papers, I am in breach of a few journal's copyright by uploading my own thesis pdf, but f
72.
by gillesjacobs 2y ago
This is entirely unsurprising and in-line with the finding that even small specialized models do better in information extraction and text classification. So no wonder finetuned large LMs do good too. Personally, my PhD did fine grained ACE
73.
by gillesjacobs 2y ago
Facebook/Meta has been doing open research in ML and NLP for far longer than the current LLM era: Convolutional Neural Networks for Sentence Classification (2014), FastText (2016), PyTorch (2016), fairseq (2017), LASER (2018), RoBERTa
74.
by gillesjacobs 2y ago
I stand corrected. Guess all that posturing about privacy was all empty election promises then. Seems like only the Greens, the left and within ID the AFD voted against.
75.
by gillesjacobs 2y ago
There are many different types and implementations of representative democracy. And the supranational technocracy-driven particracy of the EU is only representative in name. It is on the oligarchic end of the democracy-oligarchy spectrum.
76.
by gillesjacobs 2y ago
Edit: well like someone pointed out instead of party programmes and declarations, we should look at past voting behaviour on the 2021 vote: https://mepwatch.eu/9/vote.html?v=134463&eugroup=ID Seems like only the Gr
77.
by gillesjacobs 2y ago
Supranational/federalist "representative" particracy does not work. Representative democracy works better with increased locality, where policy and politicians are directly beholden to constituents.
78.
by gillesjacobs 2y ago
This isn't necessarily true: the Identity and Democracy faction is explicitly pro-privacy, digital rights and against Chat Control. - https://id-party.eu/program/ (ID Party Official Site) - https://idgr
79.
by gillesjacobs 2y ago
Because it is exactly the EU governance that lacks democratic representation, enabling unfavoured politicians to remain in significant power. You can only elect MEPs (via party lists) locally. The majority of legalislation is drafted by the
80.
by gillesjacobs 2y ago
The field of study is called linguistic typology. During my linguistic studies, the course material was comprised of the Oxford and Cambridge Handbook of Linguistic Typology (and a bunch of papers on specific linguistic features). But the O
81.
by gillesjacobs 2y ago
Dat het verschil in gemiddelde passieve woordenschat tussen Duitse en Nederlandse moedertaalsprekers zo groot is, twijfel ik als linguïst toch wel een beetje aan. Are you sure that there study/studies are sound and use directly compara
82.
by gillesjacobs 3y ago
Ah this takes me back a decade when I used Praat to annotate the prosody of "no wonder" mirative statements for my Computational Linguistics master thesis. Hours spent in Praat and that horrible scripting language... It did lead t
83.
by gillesjacobs 3y ago
Always beware these type of Social Science studies that make for great popsci headlines. The kind of result that makes careers: seemingly elegant and simple study design, common sense hypothesis and overall great marketability. Diederik Sta
84.
by gillesjacobs 3y ago
Used this library quite a bit. I still have no idea if there is a good reason this API is not packaged within huggingface/transformers. Probably historic reasons, anyway solid 9/10 API.
85.
by gillesjacobs 3y ago
Back in 2020, so that's 4 years of active development on both tools since then. Both tools have different scopes though (cf my other comment in the thread).
86.
by gillesjacobs 3y ago
You need to upload your time series (loss, performance metrics) to the cloud. Not your weights or models. Reliance on cloud services is a legitimate worry though for privacy, IP, process control, reliability, etc. The comparison between Opt
87.
by gillesjacobs 3y ago
I have extensively tested Optuna and Weights and Biases (WandB) for hyperparameter tuning on multiple task-specific transformer models back in 2020. Optuna lost out by a long mile back then in feature parity and dashboarding. Optuna did not
88.
The Wrong Way to Make 3DPrinter Filament(youtube.com)
2 points by gillesjacobs 3y ago|0 comments
89.
by gillesjacobs 3y ago
Everything is an engineering trade-off between classification performance, data requirements (quality and quantity for supervised fine-tuning/training), dev time and inference speed/resources. You don't always have to do full
90.
by gillesjacobs 3y ago
The yes-or-no reference answer test is a really bad way to go about this. Maybe take a note out of RAGAS evaluation templates and use an LLM to iteratively summarise the nuanced category.

More ›