TheYang

@TheYang@lemmy.world

• ⁨Comment⁩ on ⁨Can local LLMs be as useful and insightful as those widely available?⁩ ⁨⁨13⁩ ⁨hours⁩ ago⁩:

  Because you don’t train your self-hosted LLM.
  As a result you only pay for the electricity of computing your tokens (your request), this can be especially reasonable if the same machine also does local game streaming and or transcoding, and thus already has the requirements to host a LLM.

  If you don’t have rather unreasonable means, your local LLM is just very much more limited in parameters (size), and will not be as good as other, much larger models.

  Privacy, Ethics and personal interest usually are the largest drivers from what I can tell.

• ⁨Comment⁩ on ⁨Any Thoughts, Ideas or Theories of the Bambu H2D that they have been teasing?⁩ ⁨⁨1⁩ ⁨month⁩ ago⁩:

  I doubt that it’ll really have killer features.

  You’ll most likely be able to exchange the 2 Hotend-Toolhead for a Laser-Hotend, it’ll have a heated AMS, it may have a vinyl-cutter head.

  I don’t really think I’d want to Laser on my heated bed, or cut on it either. The Fumes from lasering will impact durability of anything in the printer, without really lots of ventilation it will produce lots of dust (well, ash).
  Cutting on the same head is weird, as a cutter needs to resist a bit or cutting force.

  The dual-nozzle-design is interesting, but I think it’s still vastly inferior to multiple toolheads, with anything over 2 materials there is still cutting required. Depending on how they solved the issue with feeding the two hotends, I’m not sure how there won’t be quite a bit of added complexity for loading the AMS, where you have to think which head needs which filament.
  Using a single extruder gear for both hotends also increases chances and risk of cross-contamination. I’ve never had a printer who didn’t occasionally chew filament.
  Moving the Hotends on linear rails, having a mechanical drop-stopper on the hotend all increase complexity, I’m not sure how bad blob of dooms will get here.

  If they use their touted servo-design actually on the corexy kinematics, that will be interesting, because conventional wisdom says it doesn’t really improve 3d-printing performance. At least not until you get to ridiculous builds (think minuteman)

  Cost will be interesting, as apparently the H2D was touted to “be above current X1 line”, if that were to include X1E and the $2500 price tag it would be… rather expensive.
  But even when it’s “just” more expensive than the X1C at $1200/$1450, coming to… idk, $1500 in it’s bare configuration, that’s rather big chunk of change for a hobbyist. And they will (hopefully) have lost lots of enthusiasts with their firmware-stunt.

• ⁨Comment⁩ on ⁨Is this just crappy Filament⁩ ⁨⁨1⁩ ⁨month⁩ ago⁩:

  I’ve had partial clogs that manifest like that.
  Cold pulls (several) ended up resolving my issue.

  my best explanation was, that there was some debris in the nozzle, which would sometimes (nearly) seal the nozzle, and at other times be retracted with the filament, get stuck somewhere else and filament flows freely.

• ⁨Comment⁩ on ⁨A bit of a warning about the Einstar VEGA 3D-Scanner⁩ ⁨⁨2⁩ ⁨months⁩ ago⁩:

  Shining3D also makes professional-level scanners that cost as much as a car, so I banked on them putting some of that expertise into their consumer models and went with the Einstar.

  Seems like you were correct.