LLMS

→ View original post on X — @jiqizhixin

DeepSeek’s nano-vllm: lightweight Python vLLM implementation

By

@jiqizhixin

–

13 June 2025 4h22

DeepSeek engineer Xingkai Yu just dropped nano-vllm — a lightweight vLLM reimplementation from scratch. It's a clean implementation under 1,200 lines of Python code with comparable inference speeds to vLLM. Perfect for learning or building lean! Project:

→ View original post on X — @hardmaru

Hypernetworks Applied to Modern Large Language Models

By

@hardmaru

–

13 June 2025 3h42

Happy to see Hypernetworks applied to modern LLMs

→ View original post on X — @cerebras

Try Qwen3 on Cerebras: AI Model Inference Platform

By

@cerebras

–

13 June 2025 3h36

Think. But don't over think.
Try saying 'hello' to Qwen3 on Cerebras: https://
inference.cerebras.ai/?utm_source=tw
itter
…

Prefix Tuning Approaches: Layer-wise Activation Prepending Comparison

By

–

13 June 2025 3h14

My understanding is that @percyliang
's prefix tuning approach prepended activations to *every* layer, but yours only prepends to the first layer — is that right?

Prefix Tuning Initialization with Real Tokens and DoRA

By

–

13 June 2025 3h10

Is the trick of initializing with real tokens the secret to making all prefix tuning approaches competitive? BTW @EyubogluSabri I don't see DoRA mentioned in your paper – I wonder if that would close the gap a bit?

LoRA vs DoRA vs Prefix Tuning: Fine-tuning Method Effectiveness

By

–

13 June 2025 3h08

The big open question for me is why LoRA (and particularly DoRA) has been the most successful for fine-tuning, but it worked so badly for this use case. Or maybe the opposite – why did prefix tuning work so well?

Claude Limited by Context Window in Research Tasks

By

–

13 June 2025 0h06

Claude not able to continue my research chat about context compression papers because it ran out of context because it doesn't use context compression.