TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Stable Code 3B: Coding on the Edge

315 点作者 egnehots超过 1 年前

21 条评论

tarruda超过 1 年前
Note that they don&#x27;t compare with deepseek coder 6.7b, which is vastly superior to much bigger coding models. Surpassing codellama 7b is not that big of a deal today.<p>The most impressive thing about these results is how good the 1.3B deepseek coder is.
评论 #39021104 未加载
评论 #39020533 未加载
评论 #39022294 未加载
评论 #39025946 未加载
评论 #39020781 未加载
JCM9超过 1 年前
Don’t entirely understand Stability’s business model. They’ve been putting out a lot of models recently and Stable Diffusion was novel at the time, but now their models consistently seem to be somewhat second rate compared to other things out there. For example Midjourney now seems to have far surpassed them in the image generation front. After raising a ton of funding Stability seems to just be throwing a bunch of stuff out there that’s OK but no longer ground breaking. What am I missing?<p>Many other startups in the space will like face similar issues given the rapid commoditization of these models and the underlying tech. It’s very easy to spend a fortune building a model that offers a short lived incremental improvement at best before one can just quickly swap it out for something else someone else paid to train.
评论 #39022295 未加载
评论 #39021772 未加载
评论 #39022604 未加载
评论 #39024075 未加载
评论 #39021693 未加载
评论 #39021579 未加载
评论 #39026480 未加载
评论 #39021748 未加载
评论 #39027051 未加载
swyx超过 1 年前
&gt; License: Other<p>&gt; Commercial Applications<p>&gt; This model is included in our new Stability AI Membership. Visit our Membership page to take advantage of our commercial Core Model offerings, including SDXL Turbo &amp; Stable Video Diffusion.<p>what exactly is the license lol. can people use this or is this &quot;see dont touch&quot;
评论 #39023729 未加载
keyle超过 1 年前
That is fantastic. I&#x27;m building a small macOS SwiftUI client with llama cpp built in, no server-client model, and it&#x27;s already so useful with models like openhermes chat 7B, and fast.<p>If this opens it to smaller laptops, wow!<p>We truly live in crazy time. The rate of improvement in this field is off the walls.
评论 #39020165 未加载
评论 #39022314 未加载
评论 #39019911 未加载
knicholes超过 1 年前
I&#x27;ve got a machine with 4 3090s-- Anyone know which model would perform the best for programming? It&#x27;s great this can run on a machine w&#x2F;out a graphics card and is only 3B params, but I have the hardware. Might as well use it.
评论 #39022308 未加载
评论 #39020410 未加载
评论 #39020436 未加载
评论 #39020224 未加载
评论 #39022735 未加载
rahimnathwani超过 1 年前
How are people using codellama and this in their workflows?<p>I found one option: <a href="https:&#x2F;&#x2F;github.com&#x2F;xNul&#x2F;code-llama-for-vscode">https:&#x2F;&#x2F;github.com&#x2F;xNul&#x2F;code-llama-for-vscode</a><p>But I&#x27;m guessing there are others, and they might differ in how they provide context to the model.
评论 #39026704 未加载
jjtheblunt超过 1 年前
Jargon naivete question: isn&#x27;t &quot;on the edge&quot; normally implying on a server side with minimal routers hops to the client, not on client side?
评论 #39021077 未加载
评论 #39021143 未加载
outcoldman超过 1 年前
I was able to run this model in <a href="http:&#x2F;&#x2F;lmstudio.ai" rel="nofollow">http:&#x2F;&#x2F;lmstudio.ai</a> as well. Just remove Compatibility Guess in Filters, so you can see all the models. LM Studio can load it and run requests against it.
alwinaugustin超过 1 年前
I&#x27;ve been experimenting with code-llama extensively on my laptop, and from my experience, it seems that these models are still in their early stages. I primarily utilize them through a Web UI, where they can successfully refactor code given an existing snippet. However, it&#x27;s worth noting that they cannot currently analyze entire codebases or packages, refining them based on the most suitable solutions using the most appropriate algorithms. While these models offer assistance to some extent, there is room for improvement in their ability to handle more complex and comprehensive coding scenarios.
评论 #39023134 未加载
评论 #39026693 未加载
lfkdev超过 1 年前
How is this compared to the current GitHub Copilot?
评论 #39020078 未加载
评论 #39020226 未加载
connorgutman超过 1 年前
FYI: This model is already available on Ollama.
评论 #39020622 未加载
artninja1988超过 1 年前
Given the complete failure of the first stable lm, I&#x27;m interested to try this one out. Haven&#x27;t really seen a small language model, except mixtral 7b that&#x27;s really useful for much.<p>I also hope stability comes out with a competitor to the new midjourney and dalle models! That&#x27;s what put them on the map in the first place
评论 #39022330 未加载
评论 #39020304 未加载
评论 #39020097 未加载
mchiang超过 1 年前
It&#x27;s amazing to see more smaller models being released. This creates opportunities for more developers to run it on their local computers, and makes it easier to fine-tune for specific needs.
评论 #39019985 未加载
hospitalJail超过 1 年前
Seems like they caught the Apple Marketing bug and are chasing things noonecares about. Great 3B model, everyone is already running 7B models over here.<p>Maybe one day when I need to do offline coding on my cellphone, it will be really useful.
alastairr超过 1 年前
does anyone have recommendations for addins to integrate these &#x27;smaller&#x27; llms into an IDE like VSCode? I&#x27;m pretty embedded with GH copilot, but curious to explore other options.
herval超过 1 年前
Can anyone explain what’s Stability’s business model (or plan for one)?<p>I get why Meta releases tons of models, but still can’t quite understand what stability is trying to achieve
评论 #39021250 未加载
评论 #39021208 未加载
sytelus超过 1 年前
Why authors miss to compare with Phi-2?
评论 #39025805 未加载
photon_collider超过 1 年前
How reliable are these benchmarks?
评论 #39020119 未加载
ihaag超过 1 年前
Terrible model
akulbe超过 1 年前
I just tried this model with Koboldcpp on my LLM box. I got gibberish back.<p>My prompt - &quot;please show me how to write a web scraper in Python&quot;<p>The response?<p>&lt;blockquote&gt; I&#x27;ve written my first ever python script about 5 months ago and I really don&#x27;t remember anything except for the fact that I used Selenium in order to scrape websites (in this case, Google). So you can probably just copy&#x2F;paste all of these lines from your own Python code which contains logic to determine what value should be returned when called by another piece of software or program. &lt;&#x2F;blockquote&gt;
评论 #39020288 未加载
评论 #39020282 未加载
评论 #39020316 未加载
评论 #39020248 未加载
评论 #39020257 未加载
评论 #39020287 未加载
kleiba超过 1 年前
It&#x27;s quite amazing - I often find that I read quite positive comments towards LLM tools for coding. Yet, an &quot;Ask HN&quot; I posted a while ago (and which admittedly didn&#x27;t gain much traction) seemed to mirror mostly negative&#x2F;pessimistic responses.<p><a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=38803836">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=38803836</a><p>Was it just that my submission didn&#x27;t find enough &#x2F; more balanced commenters?
评论 #39020392 未加载
评论 #39027032 未加载
评论 #39020412 未加载
评论 #39020349 未加载