-
Upload full trainer checkpoint step=8112 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=7605 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=7098 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=6591 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=6084 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=5577 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=5070 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=4563 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=4056 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=3549 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=3042 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=2535 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=2028 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=1521 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=1014 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=507 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=10557 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=9384 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=8211 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=7038 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=5865 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=4692 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=3519 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=2346 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=1173 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=18 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=9 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=2346 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=1173 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=141 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=24 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=12 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=1 (profile=llama3_2_1b_base_grpo, run_type=grpo)
-
Upload full trainer checkpoint step=29325 (profile=llama3_2_3b_base, run_type=sft)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_3b_base, step=23460)
-
Upload full trainer checkpoint step=23460 (profile=llama3_2_3b_base, run_type=sft)
-
Upload full trainer checkpoint step=17595 (profile=llama3_2_3b_base, run_type=sft)
-
Refresh base adapter to best eval_loss (profile=llama3_2_3b_base, step=8796, eval_loss=0.036849)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_3b_base, step=11730)
-
Upload full trainer checkpoint step=11730 (profile=llama3_2_3b_base, run_type=sft)
-
Upload full trainer checkpoint step=8796 (profile=llama3_2_3b_base, run_type=sft)
-
Upload full trainer checkpoint step=5865 (profile=llama3_2_3b_base, run_type=sft)
-
Upload full trainer checkpoint step=5864 (profile=llama3_2_3b_base, run_type=sft)
-
Upload full trainer checkpoint step=2932 (profile=llama3_2_3b_base, run_type=sft)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=11728, eval_loss=0.039292)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_1b_base, step=11730)
-
Upload full trainer checkpoint step=11728 (profile=llama3_2_1b_base, run_type=sft)
-
Upload full trainer checkpoint step=8796 (profile=llama3_2_1b_base, run_type=sft)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=5864, eval_loss=0.039609)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=5865)