4 ms·
I tried to generate a streamlit dashboard with MACD, RSI, MA(200). 1:0 for qwen3 here. qwen3-coder-30b 4-bit mlx took on the task w/o any hiccups with a fully
by hrpnk 1y ago
I tried to generate a streamlit dashboard with MACD, RSI, MA(200). 1:0 for qwen3 here.
qwen3-coder-30b 4-bit mlx took on the task w/o any hiccups with a fully working dashboard, graphs, and recent data fetched from yfinance.
gpt-oss-20b mxfp4's code had a missing datatime import and when fixed delivered a dashboard without any data and with starting date of Aug 2020. Having adjusted the date, the update methods did not work and displayed error messages.
- teitoklien 1y agofor now, i wouldnt rank any model from openai in coding benchmarks, despite all the false messaging they are giving, almost every single model openai has launched even the high end o3 expensive models are absolutely monumentally horrible at coding tasks. So this is expected. If its decent in other tasks, which i do find openai often being better than others at, then i think its a win, especially a win for the open source community that even AI labs that pionered the hype of Gen AI who didnt want to ever launch open models are now being forced to launch them. That is definitely a win, and not something that was certain before.
- dongobread 1y agoIt is absolutely awful at writing and general knowledge. IMO coding is its greatest strength by far.
- mplewis 1y agoSure sounds like they're not good at anything in particular, then.
- xwolfi 1y agowelcome to 3DTV hype, LLM are useless...
- teitoklien 1y agonot really, claude is amazing, which is why I pay for Claude Max, its insanely amazing and useful, its just OpenAI's one isnt
- pxc 1y agoNVIDIA will probably give us nice, coding-focused fine-tunes of these models at some point, and those might compare more favorably against the smaller Qwen3 Coder.
- iJohnDoe 1y agoWhat is the best local coder model that that can be used with ollama? Maybe a too opened ended question? I can run the deepseek model locally really nicely.