Only 3.8B parameters according to the paper, so it ought to be quite easy on the hardware as well if they do.
Only 3.8B parameters according to the paper, so it ought to be quite easy on the hardware as well if they do.
Did you follow some guide when replacing the nozzle and bowden tube? Did you hot tighten it when putting them back?
What speed are you printing at, which you think is too slow? What’s your acceleration?
It’s a myth that capricorn tubing is better in any way
I’d call it “successful marketing” :) But yeah, they’re really managed to convince a lot of people that it has near magical properties, and I’ve seen a lot of random people recommend it to others for all sorts of issues.
It sounds like bed adhesion might have got worse, perhaps you have touched the print surface with your fingers while removing prints? You could try removing the plate and washing it with warm water and soap. Some people use IPA but if you do then you need to make sure you really wipe it clean before it evaporates, otherwise the dissolved fats will stay on the bed. If your bed has some kind of anti-stick coating I think there’s also a risk that you damage if you use stronger solvents.
As for warping in general it could be an indication that your flow rate is exceeding your melting capacity. If you have an all metal hotend you could try printing at higher temp, if not then try reducing print speed instead.
Ahh, I thought you meant you had a 0.2mm nozzle, but now I see you probably meant layer height.
Moisture absorbtion is rarely a problem with PLA, but hopefully dehydration won’t hurt, as long as you don’t accidentally overheat it and it deforms. I’ve left rolls of PLA out in the open for 6 months without noticing any deterioration. Both your filaments used to print fine, and then the oozing spontaneously started with both of them?
2mm retraction should be more than enough for a direct drive extruder.
What filament and other slicer settings? Could be too hot. Could be retraction settings. Did the oozing start when you switched nozzles? If it’s a cheap Amazon nozzle it might be faulty and have a different diameter than advertised. Did you follow the correct procedure with hot tightening when switching nozzles? If not, you might have got molten filament in between the nozzle and the heat break.
Assuming they already own a PC, if someone buys two 3090 for it they’ll probably also have to upgrade their PSU so that might be worth including in the budget. But it’s definitely a relatively low cost way to get more VRAM, there are people who run 3 or 4 RTX3090 too.
Probably a joke, since there’s always someone commenting that the item isn’t food safe whenever a model for something food related.
It’s not easy trying to research which 3d printer to buy, there is more click bait and marketing than impartial reviews out there, and search engines tend to promote the garbage. And without a lot of 3d printing experience, it can be difficult to know if a “review” is paid for by the printer’s manufacturer, or just trying to trick you into clicking their affiliate links. There are also no consistently good brands if you’re looking for a cheap printer, pretty much all of them have produced a few good printers and others that have more flaws. For example old Ender 3 and Ender 3 Pro were very good at the time, and Creality built up a lot of brand recognition, but then they switched to low quality components and seemingly stopped doing quality control and made a bunch of crap. Now it might be turning around again, as Creality’s latest printers are starting to look decent again, although perhaps a little overpriced.
Personally I use this spreadsheet to compare pros and cons of budget printers. It’s maintained by a group of users at a 3d printing discord server, and while one cannot know for sure none of them have any ties for example to Sovol (the most recommended budget brand currently), they’ve seemed quite impartial to me so far.
The only issue is (layer height), which of course can change from print to print.
Technically not the only issue 😀 it can also change from layer to layer.
Maybe calculate it from the length of filament being pushed out / the length of the movement * filament cross section? I’m on my phone so I can’t check right now but that info should be possible to extract from the gcode iirc
Never heard of this printer brand, but if it turns out that its mainboard isn’t compatible with what you’re trying to do then it might be an option to replace it with a cheap stepper driver board from AliExpress. You wouldn’t need anything fancy if you’re running Klipper on your RPi
For LLMs it entirely depends on what size models you want to use and how fast you want it to run. Since there’s diminishing returns to increasing model sizes, i.e. a 14B model isn’t twice as good as a 7B model, the best bang for the buck will be achieved with the smallest model you think has acceptable quality. And if you think generation speeds of around 1 token/second are acceptable, you’ll probably get more value for money using partial offloading.
If your answer is “I don’t know what models I want to run” then a second-hand RTX3090 is probably your best bet. If you want to run larger models, building a rig with multiple (used) RTX3090 is probably still the cheapest way to do it.
Is max tokens different from context size?
Might be worth keeping in mind that the generated tokens go into the context, so if you set it to 1k with 4k context you only get 3k left for character card and chat history. I think i usually have it set to 400 tokens or something, and use TGW’s continue button in case a long response gets cut off
llama.cpp uses the gpu if you compile it with gpu support and you tell it to use the gpu…
Never used koboldcpp, so I don’t know why it would it would give you shorter responses if both the model and the prompt are the same (also assuming you’ve generated multiple times, and it’s always the same). If you don’t want to use discord to visit the official koboldcpp server, you might get more answers from a more llm-focused community such as !localllama@sh.itjust.works
Hooray, I can finally play it. Had it on my wish-list for years, when I finally bought it I found out that neither the native Linux or the Windows+Proton version was working.
A static website and Immich
There are tons of options for running LLMs locally nowadays, though none come close to GPT4 or Claude 2 etc. One place to start is /c/localllama@sh.itjust.works
Static html+css page generated with this: https://github.com/maximtrp/tab
Do you mean that you want to build the docker image on one computer, export it to a different computer where it’s going to run, and there shouldn’t be any traces of the build process on the first computer? Perhaps it’s possible with the –output option… Otherwise you could write a small script which combines the commands for docker build, export to file, delete local image, and clean up the system.
Cleaning the print surface with warm water and soap, and then avoiding touching it with your hands, is a good start if you haven’t done so alreayd. Calibrating first layer height, flow rate, temperatures, etc. is generally the way, but if you want a quick and easy solution I gotta say that Magigoo has worked really well for me. It’s a bit expensive, but I’ve reapplied it a few times and never have to wash the bed so that one bottle will likely last a lifetime. I think you can get similar results with a high PVA content glue stick or hair spray. And there are of course other 3d print glues which I haven’t tried, I’m assuming they work equally well.