if you're on Strata before 0.1.45 and your client doesn't set sampling, you're running greedy
Two things I ran into while moving my Strata setup from IQ2XS to Unsloth's UD-IQ4XS (RTX 4090, i9-14900K, 64GB DDR5-5800, Arch).
Proof1 community thread
Key points
- Before Strata 0.1.45, a run config without a sampling block decoded greedy whenever the request didn't set its own temperature.
- Greedy is what makes these models loop.
- A request with no sampling now gets the model card's values:
- "sampling": {"temperature": 1.0, "topp": 0.95, "topk": 20}
Sources (1)
- [1]if you're on Strata before 0.1.45 and your client doesn't set sampling, you're running greedyr/LocalLLaMA (top, daily) · Oct 17, 09:34 PM
Two things I ran into while moving my Strata setup from IQ2_XS to Unsloth's UD-IQ4_XS (RTX 4090, i9-14900K, 64GB DDR5-5800, Arch).
Before Strata 0.1.45, a run config without a `sampling` block decoded greedy whenever the request didn't set its own temperature.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 17, 202632GB to 64GB of RAM took Qwen3.8-Flash-Next from IQ2_XS (2-bit) to Unsloth UD-IQ4_XS (4-bit) on a single 4090
- Oct 17, 2026Nemotron 3.5 Bolt, Muse Spark Open, and a Fast LFM2.5 24B
- Oct 16, 2026Improved Qwen3.8-Flash Chat Template (w/ Sharp Template terse option)
- Oct 16, 2026Qwen Flash Next potential SOTA Q2 nobody talks about
- Oct 15, 2026capy-flash: Qwen3.8-Flash-Next (125B MoE) in full NVFP4 on two 16 GB consumer GPUs
- Oct 15, 2026Local and Distributed Large LLM Inference with MLX and RDMA on Mac Studio