_ea1k

26 ms·

391.
by _ea1k 3y ago
IIRC, Bard was limited in Canada due to some similar provisions? I wonder if this will have impact on their LLM access in California as well.
392.
by _ea1k 3y ago
IDK, is it really that much more powerful than the P40, which is already fairly cheap?
393.
by _ea1k 3y ago
The core charge on lead acid batteries is really high. It'd be surprising if they weren't getting recycled the vast majority of the time. 99% sounds high, but I wouldn't say it is impossible.
394.
by _ea1k 3y ago
I wouldn't say that EV companies are, but the battery recycling businesses are working pretty hard at it. It won't really get started until there are more EV batteries to recycle. Volumes are still relatively low compared to the n
395.
by _ea1k 3y ago
I think your response implies that I'm referring to Tesla as seeing exponential growth in a robotaxi. Tesla is currently at zero on that dimension. They might get it above zero and grow rapidly in 2-3 years, but I'd give it very l
396.
by _ea1k 3y ago
> Self driving cars were supposed to be around the corner because AI was advancing so fast. To be fair, that was almost true. Waymo is doing well, Mobileye is looking good in demos, and even Tesla has a solid L2 system now. It seems inev
397.
by _ea1k 3y ago
Given the relative availability, I'd probably try to do it with a couple of rtx4090s on tensordock.
398.
by _ea1k 3y ago
I told it that I planned to use it to jump out of an airplane safely. Apparently it didn't have much faith in my abilities.
399.
by _ea1k 3y ago
I think it is reasonable that they wouldn't include actual dangerous things, for the same reasons that the platforms themselves make a token attempt to avoid describing dangerous things. Having said that, I asked ChatGPT how to DIY a p
400.
by _ea1k 3y ago
The Tesla supercharger team rarely gets enough credit. They do an amazing job behind the scenes to make all of that a reality. I still use other charging networks from time to time. The best that I can say is that they are usable, but they
401.
by _ea1k 3y ago
It was trained with "1 trillion tokens from large-scale open-source pretraining datasets, including RefinedWeb, Pile, Github data, etc." I guess it is good that they mentioned some of it, but yeah, that isn't exceptionally he
402.
by _ea1k 3y ago
I wonder if the reasoning capability of a moe model is limited by the max size of the individual experts?
403.
by _ea1k 3y ago
I don't think they'd become a fire hazard, but it is true that one would likely pick something else for this application. Having said that, switching to something like the Tesla V100-SXM2-16GB wouldn't cost that much more. TB
404.
by _ea1k 3y ago
I'm sure it happens. We get EV 'reviews' from time to time in some of the ev forums and reddits. I've noticed a common tendency to get this experience, then extrapolate it to the rest of the experiences as well. It isn&#
405.
by _ea1k 3y ago
I think they only give you the card, so walkup unlock won't happen, at least before setting up a phone key.
406.
by _ea1k 3y ago
Did they charge you idle fees somehow? Unfortunately that's the other mistake that we've seen sometimes. Not that you actually incurred them necessarily. There've been cases of them billing hundreds in idle fees by mistake. $
407.
by _ea1k 3y ago
Not only that, but these were rented cheaply to Uber drivers. They'd come back with worn brakes from Ubers who'd drive hard and brake hard instead of using regen. They'd have batteries that had more cycles than most from char
408.
by _ea1k 3y ago
In fairness, the software is much easier to get up to speed on in the Tesla, especially as it relates to charging. Often others require messing around with smartphone apps, often with very bad UI.
409.
by _ea1k 3y ago
Not necessarily unheard of though. Tesla is planning a 52 stall site at O'Hare. That will probably be ~6MW at peak.
410.
by _ea1k 3y ago
TBH, I've experienced it where it can go either way. I've rented gas cars and ended up at a hotel with nice L2 stations. The EV would have been easier! :) But I've also rented EVs in metros with decent fast charging, but none
411.
by _ea1k 3y ago
Articles like that make it all sound so complicated and hard! TBH, the experience of most Tesla owners is just put the dest in the map and charge when needed on the way. The one clear problem that is highlighted in that article is that hote
412.
by _ea1k 3y ago
Dyno graphs aren't consistent with that: https://www.google.com/imgres?q=1993%20stock%20rx-7%20dyno&i... They are fun engines, but low end torque just isn't their thing.
413.
by _ea1k 3y ago
ollama is the easiest, IMO. The CLI interface is pretty good too. gpt4all is decent as well, and also provides a way to retrieve information from local documents.
414.
by _ea1k 3y ago
I've owned an RX-7 and an RX-8. They definitely didn't have flat torque curves. Even the twin sequential setup on the FC was far from flat. But even if it was, that was the most powerful stock configuration made and it topped out
415.
by _ea1k 3y ago
Low end torque tends to be the rotary's Achilles heel. I think the claim being made is that efficiency is better at high, steady RPMs, but tbh, I've always found that claim a bit dubious. If you love the rotary engine, this does h
416.
by _ea1k 3y ago
If you have the option to disable the mask, isn't it then a generic nn.Block?
417.
by _ea1k 3y ago
I had the same thought, but it does seem that tensorflow usage is in steady decline.
418.
by _ea1k 3y ago
It'd be shocking if it was. PyTorch isn't particularly slow.
419.
by _ea1k 3y ago
Imagine Intel running at the full speed and efficiency of the US government.
420.
by _ea1k 3y ago
ollama run mixtral will default to the quantized version (4bit IIRC). I'd guess this is why it can fit with two 3090s.

More ›