Search results

Why "reshard" becomes heavily slower for quantized model?

Status: Open.

#40 In yaof20/Flash-RL;

· LZC-BELIEVER opened on Jan 30, 2026

FP8 doesn't result in speedup over BF16

Status: Open.

#37 In yaof20/Flash-RL;

· nightlessbaron opened on Oct 8, 2025

Discrepancy in validating installation

Status: Open.

#34 In yaof20/Flash-RL;

· nightlessbaron opened on Sep 26, 2025

confuse about the calculate of least_square

Status: Open.

#23 In yaof20/Flash-RL;

· Rudin6 opened on Aug 22, 2025

Process hangs up after installing flash-llm-rl

Status: Open.

#22 In yaof20/Flash-RL;

· zzfoutofspace opened on Aug 22, 2025

Plan for multi-turn RL or agent_loop rollout in latest Verl？

Status: Open.

#19 In yaof20/Flash-RL;

· JianPeng20 opened on Aug 20, 2025

'Attention' object has no attribute 'q_scale'.

Status: Open.

#1 In yaof20/Flash-RL;

· kisseternity opened on Aug 12, 2025