-
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)
-
Upload full trainer checkpoint step=5864 (profile=llama3_2_1b_base, run_type=sft)
-
Upload full trainer checkpoint step=4398 (profile=llama3_2_1b_base, run_type=sft)
-
Upload full trainer checkpoint step=2932 (profile=llama3_2_1b_base, run_type=sft)
-
Upload full trainer checkpoint step=1466 (profile=llama3_2_1b_base, run_type=sft)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=5936, eval_loss=0.042181)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_1b_base, step=5938)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=2968, eval_loss=0.044517)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=2969)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=2968, eval_loss=0.038945)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=2969)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=2968, eval_loss=0.041312)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=2969)
-
Upload epoch checkpoint epoch_0003 (profile=llama3_2_3b_base, step=4284)
-
Refresh base adapter to best eval_loss (profile=llama3_2_3b_base, step=2856, eval_loss=0.037086)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_3b_base, step=2856)
-
Refresh base adapter to best eval_loss (profile=llama3_2_3b_base, step=1428, eval_loss=0.038845)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_3b_base, step=1428)
-
Upload epoch checkpoint epoch_0006 (profile=llama3_2_1b_base, step=8568)
-
Upload epoch checkpoint epoch_0005 (profile=llama3_2_1b_base, step=7140)
-
Upload epoch checkpoint epoch_0004 (profile=llama3_2_1b_base, step=5712)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=4284, eval_loss=0.041491)
-
Upload epoch checkpoint epoch_0003 (profile=llama3_2_1b_base, step=4284)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=2856, eval_loss=0.042449)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_1b_base, step=2856)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=1428, eval_loss=0.045955)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=1428)
-
Upload epoch checkpoint epoch_0003 (profile=llama3_2_1b_base, step=861)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=504, eval_loss=0.050252)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_1b_base, step=574)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=216, eval_loss=0.062293)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=287)
-
Refresh base adapter to best eval_loss (profile=llama3_2_3b_base, step=504, eval_loss=0.043426)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_3b_base, step=574)
-
Refresh base adapter to best eval_loss (profile=llama3_2_3b_base, step=216, eval_loss=0.052900)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_3b_base, step=287)
-
Refresh base adapter to best eval_loss (profile=llama3_1_8b_base, step=432, eval_loss=0.040582)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_1_8b_base, step=574)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=5710)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=5408, eval_loss=0.054878)
-
Upload epoch checkpoint epoch_0003 (profile=llama3_2_1b_base, step=8109)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=4056, eval_loss=0.056492)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_1b_base, step=5406)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=1352, eval_loss=0.066586)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=2703)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=3876, eval_loss=0.014388)
-
Upload epoch checkpoint epoch_0002 (profile=llama3_2_1b_base, step=5166)
-
Refresh base adapter to best eval_loss (profile=llama3_2_1b_base, step=1292, eval_loss=0.018476)
-
Upload epoch checkpoint epoch_0001 (profile=llama3_2_1b_base, step=2583)