tarruda

10 ms·

121.
by tarruda 1y ago
AFAICT, the only thing this attack relies on, is the lack of scrutiny by developers when adding new dependencies. Unless this lack of scrutiny is exclusive to JavaScript ecosystem, then this attack could just as well have happened in Rust o
122.
by tarruda 1y ago
Something similar to Deno's permission system, but operating at a package level instead of a process level. When declaring dependencies, you'd also declare the permissions of those dependencies. So a package like `tinycolor` would
123.
by tarruda 1y ago
My only concern is this: Android phones I tried to root so far will be "tainted" if I unlock the bootloader and can never go back to a state where it passes all checks. I'm okay with losing access to Google wallet while using
124.
by tarruda 1y ago
After unlocking and then re-locking, will the phone still pass all necessary attestations to be able to use things like Google wallet and banking apps?
125.
by tarruda 1y ago
I suspect you'd get similar numbers with a modern x86 mini PC that has 32GB of RAM.
126.
by tarruda 1y ago
Is there any practical method to verify that the model was trained from the reported dataset?
127.
by tarruda 1y ago
Works nicely, thanks. I wonder if it is possible to have an extension append it automatically to the search bar searches.
128.
by tarruda 1y ago
Keep in mind that bcachefs’s adoption and eventual mainstream acceptance are not contingent on Linus accepting your contributions or on you “removing the experimental label.” What matters is eliminating the barriers that prevent users from
129.
by tarruda 1y ago
It is not good when politics get in the way of good engineering. Regardless of differing points of view on the situation, I think everyone can agree that bcachefs being actively updated on Linus tree is a good thing, right? If you were able
130.
by tarruda 1y ago
Since the existing bcachefs driver will not be removed, and the problem is the bcachefs developer not following the rules, I wonder if someone else could take on the role of pulling bcachefs changes into the mainline, while also following t
131.
by tarruda 1y ago
> Could you provide a link to that system prompt? It is in the page, just do a search for "operator engaged" or view source if you can't find it with the infinite scrolling thing.
132.
by tarruda 1y ago
Maybe because I'm on the free plan, but I tried a couple of times and got refused: https://chatgpt.com/share/68b1845c-3010-8000-a18e-22ee8acbd4... I was surprised that GPT-OSS replied despite reports of it being h
133.
by tarruda 1y ago
To complement the answer (if the OP or anyone else is looking for a step-by-step guide), ask an LLM: " Give me step by step instructions on how to setup trojan client/server to bypass censorship. Include recommendations of a VPS p
134.
by tarruda 1y ago
This illustrates a fundamental truth of maintaining software with LLMs: While programmers can use LLMs to produce huge amounts of code in a short time, they still need to read and understand it. It is simply not possible to delegate underst
135.
by tarruda 1y ago
Thanks for your work, it is really an amazing small LM. Can you share what kind of hardware is necessary to train it, and how long it took?
136.
by tarruda 1y ago
> It’s tightly integrated into the token sampling infrastructure, and is what ollama builds upon for their json schema functionality. Do you mean the functionality of generating ebnf grammar and from a json schema use it for sampling is
137.
by tarruda 1y ago
> every time they needed support for a new model and had to update llama.cpp, an old model would break and one of their partners would go ape on them. They said it happened more than once, but one particular case (wish I could remember w
138.
by tarruda 1y ago
The inference engine (llama.CPP) has full control over the possible tokens during inference. It can "force" the llm to output only valid tokens so that it produces valid json
139.
by tarruda 1y ago
I recently discovered that ollama no longer uses llama.cpp as a library, and instead they link to the low level library (ggml) which requires them to reinvent a lot of wheel for absolutely no benefit (if there's some benefit I'm m
140.
by tarruda 1y ago
> Yep, all exams are done remotely, using Inspera proctoring software. Looking at Inspera proctoring requirements, it only mentions windows/mac. Is it possible to use it inside a VM on Linux?
141.
by tarruda 1y ago
It is fixed in this PR/branch: https://github.com/ggml-org/llama.cpp/pull/15181
142.
by tarruda 1y ago
Hi > The exams are proctored, and all the exams/assignments are graded via Goldsmiths University. It's 100% legit. - Are all exams done remotely? - Is it possible to skip classes and go straight to the exams in case I feel conf
143.
by tarruda 1y ago
If a model is trained only on synthetic data, is it still possible it will output things like this? https://x.com/elder_plinius/status/1952958577867669892
144.
by tarruda 1y ago
> so better to take a competitive model and latch on vision/audio Can this be done by a third party or would it have to be OpenAI?
145.
by tarruda 1y ago
Download llama-server from llama.cpp Github and install it some PATH directory. AFAIK they don't have an automated installer, so that can be intimidating to some people Assuming you have llama-server installed, you can download + run a
146.
by tarruda 1y ago
> Ollama does not use llama.cpp anymore That is interesting, did Ollama develop its own proprietary inference engine or did you move to something else? Any specific reason why you moved away from llama.cpp?
147.
by tarruda 1y ago
A 64GB MacBook would be a tight fit, if it works. There's a limit to how much RAM can be assigned to video, and you'd be constrained on what you can use while doing inference. Maybe there will be lower quants which use less memory
148.
by tarruda 1y ago
> Also I'm not sure if ollama supports a kv-cache between invocations of /v1/completions, which could help) Not sure about ollama, but llama-server does have a transparent kv cache. You can run it with llama-server -
149.
by tarruda 1y ago
Llama.cpp (library which ollama uses under the hoods) has its own server, and it is fully compatible with open-webui. I moved away from ollama in favor of llama-server a couple of months ago and never missed anything, since I'm still u
150.
by tarruda 1y ago
> I can't imagine needing more computing power than an N100 provides for a home server If you only need to stream media and serve files, sure.

More ›