wangzyuan/SafeRLHFV-Qwen35-4B-LoRA-Component-Checkpoint-Rescue download history
wangzyuan/SafeRLHFV-Qwen35-4B-LoRA-Component-Checkpoint-Rescue is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 94 times (18 in the last 7 days), and 94 times in total. It ranks #158,171 among datasets by monthly downloads.
Qwen3.5-4B synthetic component checkpoint Native checkpoint payload: 549972946 bytes / 1 original files, transported in <=64MiB parts. The training-state file embeds all three mutable LoRA banks (actor/reward critic/cost critic), two scalar heads, target ordering, three optimizer stat
Open wangzyuan/SafeRLHFV-Qwen35-4B-LoRA-Component-Checkpoint-Rescue on Hugging Face
Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.