Am I being unreasonable to find it bizarre that the tutorial begins with subscribing to 3 different SAAS vendors?
Especially seeing these days you can run a vector store on-disk if you have less than 10 million records, pull any free embedding model straight from HuggingFace and run on consumer hardware (your laptop).
Doesn't match how things are run in production these days. As a vendor, you need to target the customer's environment as closely as possible. Even if it's theoretically feasible to serve off a single machine, you should have a cloud-native setup ready to go.
In principle you could totally run this on a single bare-metal node, but most will not be doing that in practice.
> you should have a cloud-native setup ready to go
why is storing the file as a FAISS/LanceDB on-disk vector store not "cloud native"?
I am running this setup in production across dozens of nodes, we migrated all of our infrastructure off Pinecone towards this solution and have seen 10x drop in latency, and the cost improvements have been dramatic (from paid, to totally free).
I have a bit of an axe to grind in the vector DB space, it feels like the industry has gaslit developers over the last year or so into thinking SAAS is necessary for vector retrieval, when low latency on-disk KNN across vectors is a solved problem.
I totally agree that latency of this solution leaves a lot of room to improvement. But that's totally besides the point of the article, which is that people can get a no-cost semantic search for their personal website using those services. They can also use other solutions, of course.
Also I'm experimenting in further integrating things to reduce latency and most likely will publish another article within the month. Stay tuned.
Finally I somewhat agree that many of the players in the vector DB space try to push their cloud offerings. Which is fine, how else should they make money? And if latency matters that much to you, Qdrant offers custom deployments, too. I believe running Qdrant locally will handily beat your LanceDB solution perf-wise unless you're talking about less than 100k entries. We have both docker containers and release binaries for all major OSes, why not give it a try?
That's fantastic! Not all organizations (arguably most) are running their tech/infrastructure so well and competently. For a lot of organizations, it makes sense to externalize anything that's not a core competency directly related to their business. For them, less infra and less code is "better". Depending on how the accounting is done it might also be better to have a "vendor" expense rather than "internal team" expense which requires staffing.
All that is to say, maybe there's a lot of money in the SAAS/big cloud space, and customers willing to run their own setup that requires tuning might not be willing to hand them large sums of money? Just theorizing here!
Oh also "cloud native" is like a marketing term vaguely saying "you can hook this into other cloud stuff" and it works with K8s/whatever cloud thingy.
If you need semantic search locally then it's fine, but serving an embedding model might be still challenging. And if you want to expose it, your laptop might be not enough.
I've hosted embedding models on AWS Lambda (fair that this is a vendor, but 1 vs. 3), if you try an LLM with 1B+ parameters you will struggle, but if the difference between a light-weight BERT-like transformer and an LLM is only a few % of loss, why bother getting your credit card out?
Edit: another thought, skip lambda entirely and run the embedding job on the server as a background process, and use an on-disk vector store (lancedb)
Shameless plug: I built Mighty Inference Server to solve this problem. Fast embeddings with minimal footprint. Better BEIR and MTEB scores using the lightning fast and small E5 V2 models. Scales linearly on CPU, no GPU needed.
There's a bit of a difference between what you see following the 'purchase' link and what you see if you scroll down to 'pricing' on your site. It confused me at first too - I'm just so used to seeing a 'pricing' link in the top bar, I pretty much always go there first to see if there's a reasonable free tier for me to play with something.
That also depends on the scale of the monetary sponsorship, though. If I heard something like "Meta/Google/Whoever is the main sponsor of Python for the next 3 years", I'd assume (perhaps incorrectly) much more money per year than it takes to hire a single engineer. On the other hand, someone just saying "sure, we'll do the silver tier at your next few conferences" is worth a whole lot less than one engineer.
Regardless, the previous commenter's point stands that the title could be a lot more informative.
Thinking about how both The Flash and Indiana Jones 5 are doing disastrously at the box office, and likely spent a lot of ad budget on Twitter that is now getting zero return. Can't imagine Disney and WB are going to be thrilled.
Sure. But I'm a software engineer who is finishing a Ph.D. in applied informatics (coincidently in an area of time series prediction using ML). I'm not a manager.
When I mentioned "AI Winter" in front of them they didn't know what I was speaking about. But they created a nice corporate ladder which anybody can climb and it was based on years of experience with the aforementioned frameworks. Python experience needed, Scala + Spark was an advantage.
I don't know what are you planning to do. But ... I'm buying a huge load of popcorn and I will laugh my ass off when this bubble bursts in a couple of years.
(not parent but -)
It depends on area, generally:
Provider APIs(ChatGPT through OpenAPI etc), langchain, huggingface transformers, pinecone/vector DBs are absolutely taking off.
Lots of specialist ML models which required specific data collected carefully for a business task are no longer needed. Most ML roles until now, and research time, was spent collecting data and training specialist models.
General models like ChatGPT or pretrained image models do better than a fresh model trained from scratch or even a finetuned small/medium model (e.g. BERT/T5) ever will these days.
The special ML (pipeline of data --> train --> deploy --> manage/mlops/drift) used tech like PyTorch, Tensorflow, MLFlow.. and for more applicable levels (e.g. deployment), transformers, sci-kit learn, keras. However these are being replaced wholesale at many companies by langchain, huggingface inference API (for vision tasks) and pinecone/other vector DBs.
Langchain is just a smart way to wrap and order API calls to OpenAI/ChatGPT/other providers really, with some prebuilt use cases. Right now there is less on the metrics/output side than with lets say "bring-your-own-data" ML models, which you could measure things like precision.
Now, the old guard of ML (PyTorch, Tensorflow) is still used for training new models, open source replication attempts etc. But newer frameworks like JAX have not really taken off, as they have been entering the community as the community switched to using providers, rather than training their own models.
There remains a subset still powerful for communication with C-suite: Using simple models like K-means to show clusters with readable axis. They tend to use sci-kit learn or R. But this is more classic data science than ML.
There are also areas of AI so far relatively unaffected by ChatGPT etc - time series prediction (like OP, so it's less surprising they are using the old guard technologies), game engine AIs, non-discrete data, recommendation algorithms, some computer vision algorithms (especially Active learning). Some like HuggingFace (a commercial company running transformers Python module) are sort of inbetween given they serve both data-trained and the newer models.
I was skeptical at first when my doctor recommended it, but since starting Zoloft/Sertraline about 3 weeks ago I have been very productive and have for once in my life (late 20s) I feel like I am actually winning in my battle with anxiety/stress.
I am not one to immediately recommend anti-depressants, and will be looking to wean myself off in the future, but in my current state I have found it extremely beneficial.
+1 to SSRIs. One thing that's under-emphasized is that trying SSRIs is an experiment, rather than a commitment. After 5-6 weeks if they don't work for you or the side effects aren't worth it, you can drop them no problem. So if you feel like you might have clinically significant anxiety or depression, really consider an experiment.
I've had friends that have hated the side effects or not seen much benefit, who stopped. And I've had friends who've start and it's like a cloud has lifted - described as the best decision of their life.
+1 to "experiment, not a commitment," but I would also add that you really need to give them a fair shake if you're going to try them. For most antidepressant drugs, that's going to be a full 6 week trial.
Personally, I learned from my experiments that I can eat Lexapro like candy (not literally, but you get the drift) and it doesn't do anything for me, but Prozac, Wellbutrin, and Effexor have been good to me. IIRC, response rates overall to antidepressant medications are in the 40-50% range, so you might get lucky and find your wonder drug on the first try, or you might have to try more than one. Failing an antidepressant trial can certainly be a frustrating experience, but I would certainly advise anyone who asked to commit to trying up to 3 different antidepressants before throwing in the towel on these medications entirely.
This is extremely unhelpful speculation, Bob Lee had not been a part of Square/Block for many years before the incident, if this was truly connected to the facilitation of money laundering and fraud, why target an ex-CTO of all people?
The report is about the use of user number fraud and illegal transactions to enable key early insiders to cash out and leave the company in decline.
That said, it looks like most of the accusations in the report are after he left the company.
Unmaintained and putting up a pay-wall are not mutually exclusive. Sounds like they are going to layer a new pricing model on-top and leave it at that.
Agreed, don't want to discourage this project as it is helpful to have more tools to learn the rather deep and complex Git interface, but changing Git functionality under the hood is a big no-no in my mind. Using this tool on an existing Git repository and rewriting history could get the book thrown at you by more senior engineers.
> get the book thrown at you by more senior engineers.
We're all professionals working in a professional environment. "Git jail" isn't a thing. The "Git courts" won't take away your ability to use Git.
If anything, this would be an opportunity for a senior engineer, you know the people that are supposed to mentor and role model juniors, to give guidance on how to revert in the future.
I think a more rational approach would be to join a company in the AI field, rather than quitting on the spot because you think the robots are going to shortly take-over.
That's what I'm implying - I'm not retiring with the hopes of AI robots hand feeding me grapes in 5 years. I'm quitting because I think my skills and experience in building CRUD apps on the same data concepts a thousand times over is about to be pretty useless knowledge.
