HackerTrans
TopNewTrendsCommentsPastAskShowJobs

DreamGen

no profile record

Submissions

Music Generation by ElevenLabs

twitter.com
2 points·by DreamGen·2 years ago·0 comments

California's Sneate Bill 1047 on AI

leginfo.legislature.ca.gov
2 points·by DreamGen·2 years ago·1 comments

[untitled]

1 points·by DreamGen·2 years ago·0 comments

Open-source passive radar taken down for regulatory reasons (2022)

hackaday.com
67 points·by DreamGen·2 years ago·3 comments

UFO: A UI-Focused AI Agent for Windows OS Interaction

github.com
87 points·by DreamGen·2 years ago·62 comments

Monetize Your Voice with ElevenLabs

twitter.com
1 points·by DreamGen·2 years ago·0 comments

[untitled]

1 points·by DreamGen·2 years ago·0 comments

Large Language Model Course

github.com
2 points·by DreamGen·2 years ago·0 comments

Zero-Shot Identity-Preserving Image Generation in Seconds

github.com
3 points·by DreamGen·2 years ago·0 comments

5x LLM Throughput with SGLang and RadixAttention

lmsys.org
2 points·by DreamGen·2 years ago·0 comments

Transformers Are Multi-State RNNs

arxiv.org
41 points·by DreamGen·2 years ago·9 comments

Efficient and Effective Fine-Tuning Using Mixture-of-Experts PEFT

scale.com
1 points·by DreamGen·3 years ago·0 comments

Make LLM Fine-Tuning 2x Faster with Unsloth and HuggingFace TRL

huggingface.co
2 points·by DreamGen·3 years ago·0 comments

Rabbit: LLM-First Mobile Phone

rabbit.tech
352 points·by DreamGen·3 years ago·512 comments

An e-paper calendar and smart screen

invisible-computers.com
1 points·by DreamGen·3 years ago·0 comments

comments

DreamGen
·2 years ago·discuss
From what I have heard, getting license from them is also far from guaranteed. They are selective about who they want to do business with -- understandable, but something to keep in mind.
DreamGen
·2 years ago·discuss
That would be misleading. They aren't open weight (3B is not available). They aren't compared to Qwen 2.5 which beats them in many of the benchmarks presented while having more permissive license. The closed 3B is not competitive with other API only models, like Gemini Flash 8B which costs less and has better performance.
DreamGen
·2 years ago·discuss
Also, the 3B model, which is API only (so the only thing that matters is price, quality and speed) should be compared to something like Gemini Flash 1.5 8B which is cheaper than this 3B API and also has higher benchmark performance, super long context support, etc.
DreamGen
·2 years ago·discuss
Why I use Llama:

- Ability to self host. This unlocks few things: (1) Customized serving stack with various logit processors, etc. (2) More cost efficient inference.

- Ability to fine tune. Most stock instruct models are quite lame at AI story-writing and role-play and produce slop.

There aren't really any pain points specific to Llama, but if we are creating a wish list:

- Keep the pre-training data diverse. There is a worrying trend where some companies apply heavy handed filtering on the pre-training data that's not just based on quality, but also on content. Quality based filtering is understandable and desirable, but please, keep the pre-training dataset diverse :)

- Efficient inference. Open source is way behind closed source here. TensorRT-LLM is probably the most efficient from what's out there, but it's mostly closed source. Maybe Meta could contribute to some of the open source projects like vLLM (or maybe something lower level...).

- A lot of the improvements we saw recently came from post-training, post-SFT improvements. And it's not just the datasets (which clearly you can't just release), but also algorithms -- and most labs are quite secretive about the details here. The open-source community relies on DPO a lot (and more recently, KTO), since it's easy, but empirically it's not that great.
DreamGen
·2 years ago·discuss
This could have grave impact on AI development. Here are some responses:

- EFF: https://www.context.fund/policy/2024-03-26SB1047EFFSIA.pdf

- Answer AI: https://www.answer.ai/posts/2024-04-29-sb1047.html
DreamGen
·2 years ago·discuss
They were released under Apache 2.0 and there are backups in case they decide to not release them, or to only release them after further alignment:

https://huggingface.co/dreamgen/WizardLM-2-7B

https://huggingface.co/dreamgen/WizardLM-2-8x22B
DreamGen
·2 years ago·discuss
Engaging. But starting over from stage 0 gets old pretty fast.
DreamGen
·2 years ago·discuss
We are in agreement -- that's exactly what I am saying :)
DreamGen
·2 years ago·discuss
What's your source on this? They just very recently reached 100K downloads on Android and according to various SEO tools, they get maybe ~4M visits per-month (and these tend to overestimate, plus it's monthly visits, not DAU).
DreamGen
·2 years ago·discuss
A big distinction is that you can built on top (fine-tune) thus released models as well as if they released the pre-training data.
DreamGen
·2 years ago·discuss
Mistral Instruct v0.2 is 32K.
DreamGen
·2 years ago·discuss
I am not seeing any race-to-zero in the hosted offering space. Most charge multiples of what you would pay on GCP, and the public prices on GCP are already several times what you would pay as an enterprise customer.
DreamGen
·2 years ago·discuss
Great, more competition for the price-gouging platforms like Replicate and Modal is needed. As always with these, I would be curious about the cold-start time -- are you doing anything smart about being able to start (load models into VRAM) quickly? Most platforms that I tested are completely naive in their implementation, often downloading the docker image just-in-time instead of having it ready to be deployed on multiple machines.
DreamGen
·2 years ago·discuss
ChatGPT Vision will do well with this kind of OCR stuff. Just give it the header and a few example rows to get back consistently formatted output.

Or use JSON mode with the API.
DreamGen
·2 years ago·discuss
Curious how this is on the front-page, despite falling down to the second page for a while, and having so many more comments than upvotes (which usually results in demotion of the story).
DreamGen
·2 years ago·discuss
I would not be so quick to jump to conclusions. GPT-4 beats it easily in this simple logic puzzle: https://www.reddit.com/r/singularity/comments/1altttv/bard_a...

We need more data.
DreamGen
·2 years ago·discuss
> Like say the Internet

It's not so clear cut ;) "Research at CERN in Switzerland by the British computer scientist Tim Berners-Lee in 1989–90 resulted in the World Wide Web, linking hypertext documents into an information system, accessible from any node on the network."

https://en.wikipedia.org/wiki/History_of_the_Internet
DreamGen
·2 years ago·discuss
When talking about memory requirements one also needs to mention the sequence length. In case of Mixtral, which supports 32000 tokens, this can be a significant chunk of the memory used.
DreamGen
·2 years ago·discuss
This looks neat. Looking at the video, I would consider putting the "Render component" button at the bottom, so you don't have to scroll back and forth.
DreamGen
·2 years ago·discuss
> Stable LM zephyr is the best 3b chat model

By what measure? Phi 2 seems better as far as I can tell from benchmarks and usage and has much more permissive license.