More context: in ToolCallingAgent, smolagents always needs a tool call, so removing the necessity of calling a tool would just break the lib's logic. Indeed this is on vLLM's side to integrate!
OPEN SOURCE
-
smolagents 1.11.0 adds VLLMModel class for local inference
By
–
smolagents now support vLLM! As one of the most popular local inference solutions, the community had been asking us to integrate vLLM: after a heavy refactoring of our LLM classes, we've just released smolagents 1.11.0, with a brand new VLLMModel class. Go try it and tell us
-
Multi-Agent Example with Local Models by RLanceMartin
By
–
Great stuff from @RLanceMartin getting a multi agent example with local models
-
Sovereign Agentic AGI Alpha Agent deployment framework
By
–
SOVEREIGN AGENTIC AGI ALPHA AGENT SOVEREIGN_AGIALPHA/
├─ one_click_install.sh # The all-in-one installer script (Linux/macOS)
├─ run_sovagent.sh # Launch script (invokes Python)
├─ requirements.txt # Python dependencies
├─ -
OLMo 2 32B: Leading Open-Source AI Model
By
–
The latest OLMo 2 32B might be the best open-source AI model right now!
-
Agentic Routing Framework Improves Multi-Agent Efficiency
By
–
Agentic Routing Improves Efficiency! This framework by default comes with 3 agents: General search agent Deep Research agent Financial Analyst It's open-sourced, anyone can add additional agents that are connected to their own data source as well.
-
Manus AI: open source clone Anus AI appears in three days
By
–
That is the best lore I've seen in a long time in open source AI communities. There is a startup, Manus AI, that launched a product accessible only by invitation. And in three days, an open source replica of this startup, named Anus AI, has appeared.
-
DeepScaleR: 1.5B Model Achieves 43.1% Pass@1 on AIME2024
By
–
Developed by Sijun Tan (UC Berkeley) & team, this model smashes benchmarks w/ a 43.1% Pass@1 accuracy on AIME2024, a +14.3% over its base model! DeepScaleR, a 1.5B parameter model fine tuned from Deepseek-R1-Distilled-Qwen-1.5B. Watch for more: https://
bit.ly/4kmLiRf -
Panel on AI Application Building: Infrastructure and Agentic Workflows
By
–
Panel with Replit’s Michele Catasta, Stanford’s @percyliang , Nebius’ Roman Chernin and Hugging Face’s @Thom_Wolf
, moderated by @lmoroney
, on application building. Lots of tips on infra, open source, agentic workflows, benchmarking and code gen. Particular interest in how to take
