Sungmin Cha's picture

2

Sungmin Cha

csm9493

·

AI & ML interests

None yet

Recent Activity

upvoted a paper 3 days ago

Hyperparameters in Continual Learning: a Reality Check

upvoted a paper 3 days ago

Towards Robust and Parameter-Efficient Knowledge Unlearning for LLMs

updated a model about 1 month ago

csm9493/67_five_dataset_shuffle_10000_fewshot_lora_all_r16_alpha32_lr_1e5_decay_1e2_cosine_epoch_3_mbs_4

View all activity

Organizations

Papers 2

arxiv:2408.06621

arxiv:2403.09066

models 25

csm9493/67_five_dataset_shuffle_10000_fewshot_lora_all_r16_alpha32_lr_1e5_decay_1e2_cosine_epoch_3_mbs_4

Text Generation • Updated Mar 6 • 3

csm9493/26_one_dataset_cot_lora_all_r16_alpha32_lr_3e5_decay_1e2_cosine_epoch_3_mbs_4

Text Generation • Updated Mar 5 • 14

csm9493/43_five_dataset_shuffle_10000_cot_lora_all_r4_alpha8_lr_1e-05_decay_1e2_cosine_epoch_3_mbs_16

Text Generation • Updated Mar 5 • 13

csm9493/43_five_dataset_shuffle_10000_cot_lora_all_r32_alpha64_lr_1e-05_decay_1e2_cosine_epoch_3_mbs_16

Text Generation • Updated Mar 5 • 14

csm9493/43_five_dataset_shuffle_10000_cot_lora_all_r8_alpha16_lr_1e-05_decay_1e2_cosine_epoch_3_mbs_16

Text Generation • Updated Mar 5 • 14

csm9493/43_five_dataset_shuffle_10000_cot_lora_all_r16_alpha32_lr_1e-05_decay_1e2_cosine_epoch_3_mbs_16

Text Generation • Updated Mar 5 • 14

csm9493/24_three_dataset_shuffle_50000_cot_lora_all_r16_alpha32_lr_3e5_decay_1e2_cosine_epoch_2_mbs_4

Text Generation • Updated Mar 5 • 14

csm9493/41_three_dataset_shuffle_3200_cot_lora_all_r128_alpha256_lr_3e5_decay_1e2_cosine_epoch_3_mbs_4

Text Generation • Updated Mar 5 • 15

csm9493/23_one_dataset_cot_lora_all_r16_alpha32_lr_3e5_decay_1e2_cosine_epoch_2_mbs_4

Text Generation • Updated Mar 5 • 19

csm9493/37_three_dataset_shuffle_3200_cot_lora_all_r64_alpha128_lr_3e5_decay_1e2_cosine_epoch_3_mbs_4

Text Generation • Updated Mar 5 • 19

datasets

None public yet