[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


File: 1764190845300267.jpg (38 KB, 500x359)
38 KB JPG
I'm building a workflow that will automatically open websites and fill out forms. I tried using my strix halo machine running some top local models like gemma4, qwen3.5. Neither of these with their vision and tool calls could handle the task of opening a website, clicking a button, and entering information provided in a .MD file, into a websites fields. Every one failed on clicking the first button. Both of them were given an MCP server that provided tool calls for reading pages and seeing it with vision screenshots. I switched to Claude sonnet and it did all of what I wanted on the first try. I feel retarded for trusting in local models so much, what the absolute fuck /lmg/ anons.
>>
>YEEEEEEEEEEES
>YEEEEEES
>YESSSSSS
>>
qwen 3.6 is out fwiw
>>
>>108665329
Just wait for the claude mythos leak brah
>>
>>108665329
Everything that has a general thread is a consumerist toy. I think it's quite obvious that the way to enjoy this site is to filter out generals and sort the catalog by creation date.
>>
>>108665329
Why are you using LLMs at all?
You could achieve what you're asking for with a fairly small and simple collection of Ruby scripts reading the info from .MDs, via say Cucumber, piloting a Selenium web driver...
Your air-jordan-shoe-scalping / cryptocurrency daytrader bot will be infinitely slower and more bloated than all of the already existing Ruby and Groovy implementations, underperforming your competition and costing you money in the form of LLM API tokens or whatever...
Are we really at the point where people are too incompetent to do behavior driven development?
>>
>>108665329
skill issue
>>
>>108666422
It's job applications, each site and ERP is different it's not the same scriptable logic
>>
>>108665329
Literally skill issue, we're ordering pizzas with gemma 4 on /lmg/
>>
>>108667799
>each site and ERP is different it's not the same scriptable logic
oh pardon me, I was unaware that HTML destandardized form divs and submit buttons
and forgot that state machines and regex were never invented

job applications are a scam anyway though
you'll see better results from just cold calling recruiters and HR personnel and telling them that you're reaching out to ask about your upcoming interview



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.