All issues

Search results

DSV4 ignores configured prefill limit and retains stale budget after SWA rebuild

bugSomething isn't workingSomething isn't working

Status: Open.

#115 In FlashML-org/FreeToken;

· jpezzulli opened on Aug 23, 2026

RTX 2060 NOT SUPPORTED

Status: Open.

#114 In FlashML-org/FreeToken;

· ak1t4 opened on Aug 23, 2026

Requests longer than the KV pool are queued forever with no error — and --moe-cache-auto leaves only ~8k tokens of KV on a 96GB GPU

Status: Open.

#111 In FlashML-org/FreeToken;

· CygnusBill opened on Aug 23, 2026

Unhandled CUDA OOM in prefill expert workspace kills the engine; server keeps accepting requests that never return (root cause behind #20-style headless state; likely also #72)

Status: Open.

#110 In FlashML-org/FreeToken;

· CygnusBill opened on Aug 23, 2026

FreeToken engine installation doesn't work in Fedora 43 using AppImage

Status: Open.

#107 In FlashML-org/FreeToken;

· nosuko opened on Aug 23, 2026

Support Qwen3.8 27B

Status: Open.

#106 In FlashML-org/FreeToken;

· dyigitpolat opened on Aug 23, 2026

FP4 models are shown on 30XX Hardware

Status: Open.

#100 In FlashML-org/FreeToken;

· builder-main opened on Aug 23, 2026

Support for force selecting a specific GPU on multi-GPU systems

high priority

Status: Open.

#99 In FlashML-org/FreeToken;

· Proryanator opened on Aug 23, 2026

最佳实战配置参数是什么？

Status: Open.

#98 In FlashML-org/FreeToken;

· skun23 opened on Aug 23, 2026

Difficult to load model despite more than enough vram/ram available

Status: Open.

#97 In FlashML-org/FreeToken;

· XeonG opened on Aug 23, 2026

Drag and drop file support?

Desktopproblem related to FreeToken Desktopproblem related to FreeToken Desktop

featureNew feature or requestNew feature or request

Status: Open.

#96 In FlashML-org/FreeToken;

· XeonG opened on Aug 23, 2026

Dual CPU support for tensor Parallelism?

Status: Open.

#95 In FlashML-org/FreeToken;

· rog77 opened on Aug 23, 2026