It's shocking how Azure is doomed to win in AI space. It doesn't matter what happens in this field, how Microsoft can fall behind in development of LLMs. At the end of the day if people want to use it, thay need computation and Azure is a way to go.
Llama 2
31–40 of 860 posts
Re: Llama 2
#32Why doesn't FB create an API around their model and launch OpenAPI competitor? It is not like they don't have resources, and the learnings (I am referring to actual learning from users' prompts) will improve their models over time.
because Facebook is a consumer company and this is an enterprise play.
They enterprisesh plays they've tried Workplace / Parse / Neighborhoods (Nextdoor clone) haven't been super successful compared to their social / consumer plays.
Re: Llama 2
#33Why doesn't FB create an API around their model and launch OpenAPI competitor? It is not like they don't have resources, and the learnings (I am referring to actual learning from users' prompts) will improve their models over time.
There's a million different language model (not wrapper) companies offering APIs already. OpenAI, Anthropic, Cohere, Google, etc. It wouldn't be profitable.
There are really only three companies offering good language model APIs: OpenAI, Anthropic, and Microsoft Azure by serving up OpenAI's models. That is it.
Re: Llama 2
#34Why doesn't FB create an API around their model and launch OpenAPI competitor? It is not like they don't have resources, and the learnings (I am referring to actual learning from users' prompts) will improve their models over time.
They don’t run a cloud services company and get a ton of data elsewhere already. Not worth the effort (yet) imho. I could see them getting into it if the TAM truly proves out but so far it’s speculation that this would be huge for someone outside of selling compute (ex aws/azure)
Re: Llama 2
#35Earlier quoted context omitted.
That's an oddly high number for blocking competition. OpenAI's ChatGPT hit 100 million MAUs in January, and has gone down since. It's essentially a "Amazon and Google don't use this k thx."
> OpenAI's ChatGPT hit 100 million MAUs in January, and has gone down since. poor reading of the numbers. one guy at a bank pulled up similarweb and guesstimated 100m registered users and it went viral. whisper numbers were closer to 50m. but in the 6 months since they have certainly crossed 100m and probably are north of 500m, and only recently dipped.
You are countering whisper numbers with more whisper numbers.
Re: Llama 2
#36Well that is lamer than expected. The RLHF censorship was expected, but no 30B model, and single digit benchmark improvements with 40% more data? Wat. Some of the community fine tunes managed better than that. The 4k context length is nice, but RoPE makes it irrelevant anyway. Edit: Ah wait, it seems like there is a 34B model as per the paper: "We are releasing variants of Llama 2 with 7B, 13B, and 70B parameters. We…
>The 4k context length is nice, but RoPE makes it irrelevant anyway.
Can you elaborate on this?
Re: Llama 2
#37Earlier quoted context omitted.
I think more Apple. It's not like Google or Microsoft would want to use LLaMA when they have fully capable models themselves. I wouldn't be surprised if Amazon does as well. Apple is the big laggard in terms of big tech and complex neural network models.
Google's model is not as capable as llama-derived models, so I think they would actually benefit from this. > I wouldn't be surprised if Amazon does as well. I would - they are not a very major player in this space. TikTok also meets this definition and probably doesn't have LLM.
> Google's model is not as capable as llama-derived models, so I think they would actually benefit from this.
Google's publically available model isn't as capable. But they certainly have models that are far better already in house.
Re: Llama 2
#38Any idea on how it does on other languages? In particular, non-Latin languages like Arabic, Persian, Urdu, Hindi, etc.?
Re: Llama 2
#39The benchmarks look amazing compared to other open source LLMs. Bravo Meta.
Also allowing commercial use? Can be downloaded today? Available on Azure AI model catalog today? This is a very impressive release.
However, if I were starting a company I would be a little worried about the Llama 2 Acceptable Use Policy. Some of the terms in there are a little vague and quite broad. They could, potentially, be weaponized in the future. I get that Meta wants to protect themselves, but I'm a worrier.
Re: Llama 2
#40I wonder when if meta will enable this as a service similar to OpenAI - it seems to me they could monetize this ? Could be a good way for Meta to get into the infrastructure business like google/Amazon?