I am building a full fledged RLM harness in rust using fast-rlm
It's coming along nicely. Every context, prompts, file content is a python variable. Every tool runs inside the python repl. It's WIP but super excited about it! PS: deepseek-v4-flash is pretty good as an rlm
We spent three months training a numerology AI that combines current professional tools (such as asking…
Recently, two major events in the AI circle have detonated the industry.
Recently, two major events in the AI circle have detonated the industry. The two top AIs in China and the United States are officially on the same echelon, but their development routes are completely different. On one side, Musk's team launched Grok 4.6, and on the other side, Liang Wenfeng's team released DeepSeek V4-Pro-0813.
Back to basic with AI slop SAAS
TFIS Experiment #13 Finally had some time tokenmaxxing with DeepSeek V4 Flash. Reinstalled OpenCode for this one. Out of the box, minimal settings except for the custom provider. Kinda like the new clean UI in OpenCode. Unfortunately, in terms
You have reached the end of the archive
All of deepseek