Llama-8B-3.1 model SFT trained for a single epoch.
Muyang Li PRO
li-muyang
AI & ML interests
weakly-supervised learning, data mining
Organizations
None yet
models 187
li-muyang/qwen-7b-dpo_baseline_constant_lr_low
Updated
li-muyang/qwen-7b-dpo_confrank_constant_lr_low
Updated
li-muyang/qwen-7b-dpo_gradrank
Updated
li-muyang/qwen-7b-dpo_confrank_constant_lr
Updated
li-muyang/qwen-7b-dpo_baseline_constant_lr
Updated
li-muyang/qwen-7b-dpo_confrank_no_warmup
Updated
li-muyang/qwen-7b-dpo_baseline_no_warmup
Updated
li-muyang/qwen-7b-dpo_confrank
Updated
li-muyang/qwen-7b-dpo-full
Updated
li-muyang/qwen-sft-full
Updated
datasets 0
None public yet