heliophobicdude

11 ms·

91.
by heliophobicdude 3y ago
I want touch-type on a physical keyboard at my arms and hands resting position. I want pocket keyboards for AR for these reasons below: Allows you to input data without needing to look at the keyboard. The positioning of the keyboard wherev
92.
by heliophobicdude 3y ago
Humaneval needs careful consideration though. In the GPT-4 technical report, they reported contamination of humaneval data in the training data. They did measure against a "non-contaminated" training set but no idea if that can st
93.
by heliophobicdude 3y ago
LoRA probably does not affect the models biggest bottle neck. The attention mechanism. Original transformer was O(n^2d) where n is the query length and d was the cardinal of all the tokens
94.
by heliophobicdude 3y ago
Can you please elaborate on how you prune?
95.
by heliophobicdude 3y ago
It's unfortunate. There are some online tutorials that instruct you to embed all your code and perform top-k cosine similarity searches, populating the responses accordingly. It's quite interesting if you can tweak your search jus
96.
by heliophobicdude 3y ago
No moat? Where's the creativity? Google has the ability to create lock-in for users and lock-out for competitors by integrating their proprietary, existing products with natural language querying and reasoning. That's their advant
97.
by heliophobicdude 3y ago
It's in reference to the article's open source free laborers. https://www.semianalysis.com/p/google-we-have-no-moat-and-ne...
98.
by heliophobicdude 3y ago
You bottled up exactly my disappointment with some large companies' legacy AI offerings. They don't do both: iterate off of telemetry data and fulfill user's needs.
99.
by heliophobicdude 3y ago
I think this is a failure in how we fine-tuned and evaluated them in RLHF. "In theory, the human labeler can include all the context they know with each prompt to teach the model to use only the existing knowledge. However, this is imp
100.
by heliophobicdude 3y ago
These "open source" ai models are more like Obtainable models. You can obtain them. The source is not open, hence open-source. Somewhere open-source got lumped in with free or accessible. Obtainable makes sense to me.
101.
by heliophobicdude 3y ago
I thought this was a good one this week but didn't get popular. https://huyenchip.com/2023/05/02/rlhf.html
102.
by heliophobicdude 3y ago
I think the another great paper is RLHF article from Chip Huyen https://huyenchip.com/2023/05/02/rlhf.html
103.
by heliophobicdude 3y ago
This is a very well written article. Not in the article, but can we still call models like Alpaca RLHF though? What do we call these models finetune on demonstrations created by other chat bots?
104.
by heliophobicdude 3y ago
Important distinction that I'm learning today is that not all LLMs will be interoperable with each others' queries/prompts/inputs. Code LLM right now is not responding how a Chat LLM would respond. ~~~~~ Hats off to the
105.
by heliophobicdude 3y ago
Hard to compare them actually. The thing about ChatGPT is the chat part. It was trained to interact and respond with human conversation. This is more like CodePilot, with code complete based off of actual code
106.
by heliophobicdude 3y ago
Probably not the answer you would like but I think your approach to download them and figure out how to run them on your machine is a good one. You don't need to understand everything to get something working. It can be overwhelming an
107.
by heliophobicdude 3y ago
Hi Jason! I have a few thoughts on this! Refined training is usually updating the weights of usually what's called a foundational model with well structured and numerous data. It's very expensive and can disrupt the usefulness of
108.
by heliophobicdude 3y ago
Thank you!
109.
by heliophobicdude 3y ago
Hi! Could you please share details of how yours works?
110.
by heliophobicdude 3y ago
I switched to typing to Siri instead of using my voice [1]. I also disabled all triggers for Siri. Now I invoke it by pressing power button then type away. It's more useful than dictating when I need something quick and discretely. Whe
111.
by heliophobicdude 3y ago
I recently had a broken phone replaced and had depended on a backup to have my TOTP keys on my new phone. It was not a part of the phone backup. :(
112.
by heliophobicdude 3y ago
https://archive.is/xkM5H
113.
by heliophobicdude 3y ago
Something I am just now noticing is that Barb is able to view the contents of a pdf that I provided a link to. Could be making it up but need to test further...
114.
by heliophobicdude 3y ago
Can you please elaborate on this? I'm interested in knowing more. :)
115.
by heliophobicdude 3y ago
Like Simon said, it's a great language calculator! Also, adapting to its quirks is important. Using the responses from both you and it to scaffold the final response is more effective than giving only one shot
116.
by heliophobicdude 4y ago
Could be a move to pressure OpenAI to do something similar once Publishers are hooked. MS could also change models to something built in-house and cut out OpenAI
117.
by heliophobicdude 4y ago
I'm not sure but the process seems a little bit more involved than other betas. I and others were invited to a private Slack channel and asked to share feedback on a survey mentioned in the slack channel.
118.
by heliophobicdude 4y ago
https://archive.ph/3OQDa
119.
by heliophobicdude 4y ago
Hey 93po, can you please temporarily add your contact details in your bio, I would love to write you and regularly check in on your career pivot! I'm also interested as well!
120.
by heliophobicdude 4y ago
This is the "Composability" concept being pushed by the people behind LangChain. https://github.com/hwchase17/langchain

More ›