since we now use: https://docs.vllm.ai/en/latest/training/layerwise/, flashrl extras and docs: https://docs.skyrl.ai/docs/examples/flash_rl should be removed. In it's place, we should add docs for quantized training directly with vllm.
since we now use: https://docs.vllm.ai/en/latest/training/layerwise/, flashrl extras and docs: https://docs.skyrl.ai/docs/examples/flash_rl should be removed.
In it's place, we should add docs for quantized training directly with vllm.