vLLM says DeepSeek-V4-Flash showed zero logprob difference at every step over 100 RL steps on 32 GB200 GPUs. Rollout throughput averaged 24.22 output tokens/ms, 91.3% of the 26.53 baseline, and 88.6% over the first 89 steps. The post notes that response lengths differed across runs, so the figures are rollout-run measurements rather than an isolated engine benchmark.
Part of a thread · earlier post
3/ DeepSeek-V4-Flash: diff = 0 at every step over 100 RL steps on 32 GB200 GPUs.
Rollout throughput averaged 24.22 vs. 26.53 output tokens/ms: 91.3% of baseline.
Over the first 89 steps, the ratio is 88.6%. Response lengths differ across runs, so these are rollout-run measurements, not an isolated engine benchmark.
Source: vLLM · x.comPublished · added here
