Checkpoints trained on the BBO-Pile dataset. More details: https://arxiv.org/pdf/2605.23417
-
synetune/qwen3_30M_token_800M_lr_5e-3_bsz_4_seed_0
Text Generation • 30.6M • Updated • 12 -
synetune/qwen3_80M_token_400M_lr_5e-3_bsz_4_seed_0
Text Generation • 77.6M • Updated • 15 -
synetune/qwen3_5M_token_400M_lr_1e-2_bsz_8_seed_0
Text Generation • 4.96M • Updated • 15 -
synetune/qwen3_2M_token_1B_lr_1e-2_bsz_8_seed_0
Text Generation • 2.55M • Updated • 16