PyTorch
Chinese
alpaca
Chinese-Vicuna
llama
Facico commited on
Commit
c53974f
·
1 Parent(s): cca810e

Upload 6 files

Browse files
Files changed (6) hide show
  1. optimizer.pt +3 -0
  2. pytorch_model.bin +3 -0
  3. scaler.pt +3 -0
  4. scheduler.pt +3 -0
  5. trainer_state.json +3960 -0
  6. training_args.bin +3 -0
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38ccc11f6ce0dec57cfc55c480235f130f1f38bcdd345f6e8fa3e3ede32667d2
3
+ size 8556805
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35d28c3716ed97b74f9056243f7103927d67717e140fffa894f9dbf2f2bf70c1
3
+ size 16822989
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4589151aea8ed46a82534a0ba3218415960c8e634a5d6f6b4d08425e37e874d3
3
+ size 557
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8653a90e40ac3737e454e1218d2e4b9813c4bafc9fa945c5c592e3c4a893cd25
3
+ size 627
trainer_state.json ADDED
@@ -0,0 +1,3960 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7965201735496521,
3
+ "best_model_checkpoint": "lora-Vicuna/checkpoint-1000",
4
+ "epoch": 2.1394319439321285,
5
+ "global_step": 11600,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 5.9999999999999995e-05,
13
+ "loss": 2.041,
14
+ "step": 20
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 0.00011999999999999999,
19
+ "loss": 1.5971,
20
+ "step": 40
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 0.00017999999999999998,
25
+ "loss": 1.2446,
26
+ "step": 60
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 0.00023999999999999998,
31
+ "loss": 1.1737,
32
+ "step": 80
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 0.0003,
37
+ "loss": 1.1234,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 0.0002996287817855596,
43
+ "loss": 1.0991,
44
+ "step": 120
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 0.0002992575635711192,
49
+ "loss": 1.089,
50
+ "step": 140
51
+ },
52
+ {
53
+ "epoch": 0.03,
54
+ "learning_rate": 0.0002988863453566788,
55
+ "loss": 1.0685,
56
+ "step": 160
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 0.0002985151271422384,
61
+ "loss": 1.0466,
62
+ "step": 180
63
+ },
64
+ {
65
+ "epoch": 0.04,
66
+ "learning_rate": 0.00029814390892779805,
67
+ "loss": 1.0441,
68
+ "step": 200
69
+ },
70
+ {
71
+ "epoch": 0.04,
72
+ "eval_loss": 0.8649595379829407,
73
+ "eval_runtime": 1.3492,
74
+ "eval_samples_per_second": 0.741,
75
+ "eval_steps_per_second": 0.741,
76
+ "step": 200
77
+ },
78
+ {
79
+ "epoch": 0.04,
80
+ "learning_rate": 0.0002977726907133576,
81
+ "loss": 1.0417,
82
+ "step": 220
83
+ },
84
+ {
85
+ "epoch": 0.04,
86
+ "learning_rate": 0.00029740147249891725,
87
+ "loss": 1.0269,
88
+ "step": 240
89
+ },
90
+ {
91
+ "epoch": 0.05,
92
+ "learning_rate": 0.0002970302542844769,
93
+ "loss": 1.015,
94
+ "step": 260
95
+ },
96
+ {
97
+ "epoch": 0.05,
98
+ "learning_rate": 0.0002966590360700365,
99
+ "loss": 1.0089,
100
+ "step": 280
101
+ },
102
+ {
103
+ "epoch": 0.06,
104
+ "learning_rate": 0.0002962878178555961,
105
+ "loss": 1.0269,
106
+ "step": 300
107
+ },
108
+ {
109
+ "epoch": 0.06,
110
+ "learning_rate": 0.0002959165996411557,
111
+ "loss": 1.0187,
112
+ "step": 320
113
+ },
114
+ {
115
+ "epoch": 0.06,
116
+ "learning_rate": 0.00029554538142671533,
117
+ "loss": 1.0142,
118
+ "step": 340
119
+ },
120
+ {
121
+ "epoch": 0.07,
122
+ "learning_rate": 0.0002951741632122749,
123
+ "loss": 1.0077,
124
+ "step": 360
125
+ },
126
+ {
127
+ "epoch": 0.07,
128
+ "learning_rate": 0.00029480294499783453,
129
+ "loss": 1.004,
130
+ "step": 380
131
+ },
132
+ {
133
+ "epoch": 0.07,
134
+ "learning_rate": 0.00029443172678339416,
135
+ "loss": 1.0042,
136
+ "step": 400
137
+ },
138
+ {
139
+ "epoch": 0.07,
140
+ "eval_loss": 0.8426274061203003,
141
+ "eval_runtime": 1.3288,
142
+ "eval_samples_per_second": 0.753,
143
+ "eval_steps_per_second": 0.753,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 0.08,
148
+ "learning_rate": 0.00029406050856895373,
149
+ "loss": 1.0061,
150
+ "step": 420
151
+ },
152
+ {
153
+ "epoch": 0.08,
154
+ "learning_rate": 0.00029368929035451335,
155
+ "loss": 1.0152,
156
+ "step": 440
157
+ },
158
+ {
159
+ "epoch": 0.08,
160
+ "learning_rate": 0.000293318072140073,
161
+ "loss": 1.0049,
162
+ "step": 460
163
+ },
164
+ {
165
+ "epoch": 0.09,
166
+ "learning_rate": 0.0002929468539256326,
167
+ "loss": 1.0011,
168
+ "step": 480
169
+ },
170
+ {
171
+ "epoch": 0.09,
172
+ "learning_rate": 0.00029257563571119223,
173
+ "loss": 0.9932,
174
+ "step": 500
175
+ },
176
+ {
177
+ "epoch": 0.1,
178
+ "learning_rate": 0.0002922044174967518,
179
+ "loss": 0.9926,
180
+ "step": 520
181
+ },
182
+ {
183
+ "epoch": 0.1,
184
+ "learning_rate": 0.00029183319928231143,
185
+ "loss": 0.9971,
186
+ "step": 540
187
+ },
188
+ {
189
+ "epoch": 0.1,
190
+ "learning_rate": 0.000291461981067871,
191
+ "loss": 0.9866,
192
+ "step": 560
193
+ },
194
+ {
195
+ "epoch": 0.11,
196
+ "learning_rate": 0.00029109076285343063,
197
+ "loss": 0.9871,
198
+ "step": 580
199
+ },
200
+ {
201
+ "epoch": 0.11,
202
+ "learning_rate": 0.00029071954463899026,
203
+ "loss": 0.9906,
204
+ "step": 600
205
+ },
206
+ {
207
+ "epoch": 0.11,
208
+ "eval_loss": 0.8352808356285095,
209
+ "eval_runtime": 1.3169,
210
+ "eval_samples_per_second": 0.759,
211
+ "eval_steps_per_second": 0.759,
212
+ "step": 600
213
+ },
214
+ {
215
+ "epoch": 0.11,
216
+ "learning_rate": 0.0002903483264245499,
217
+ "loss": 0.9797,
218
+ "step": 620
219
+ },
220
+ {
221
+ "epoch": 0.12,
222
+ "learning_rate": 0.0002899771082101095,
223
+ "loss": 0.989,
224
+ "step": 640
225
+ },
226
+ {
227
+ "epoch": 0.12,
228
+ "learning_rate": 0.0002896058899956691,
229
+ "loss": 0.9931,
230
+ "step": 660
231
+ },
232
+ {
233
+ "epoch": 0.13,
234
+ "learning_rate": 0.0002892346717812287,
235
+ "loss": 0.9859,
236
+ "step": 680
237
+ },
238
+ {
239
+ "epoch": 0.13,
240
+ "learning_rate": 0.00028886345356678834,
241
+ "loss": 0.9874,
242
+ "step": 700
243
+ },
244
+ {
245
+ "epoch": 0.13,
246
+ "learning_rate": 0.0002884922353523479,
247
+ "loss": 0.9888,
248
+ "step": 720
249
+ },
250
+ {
251
+ "epoch": 0.14,
252
+ "learning_rate": 0.00028812101713790754,
253
+ "loss": 0.9798,
254
+ "step": 740
255
+ },
256
+ {
257
+ "epoch": 0.14,
258
+ "learning_rate": 0.00028774979892346716,
259
+ "loss": 0.981,
260
+ "step": 760
261
+ },
262
+ {
263
+ "epoch": 0.14,
264
+ "learning_rate": 0.00028737858070902674,
265
+ "loss": 0.9824,
266
+ "step": 780
267
+ },
268
+ {
269
+ "epoch": 0.15,
270
+ "learning_rate": 0.00028700736249458636,
271
+ "loss": 0.9922,
272
+ "step": 800
273
+ },
274
+ {
275
+ "epoch": 0.15,
276
+ "eval_loss": 0.8162809610366821,
277
+ "eval_runtime": 1.2603,
278
+ "eval_samples_per_second": 0.793,
279
+ "eval_steps_per_second": 0.793,
280
+ "step": 800
281
+ },
282
+ {
283
+ "epoch": 0.15,
284
+ "learning_rate": 0.000286636144280146,
285
+ "loss": 0.9942,
286
+ "step": 820
287
+ },
288
+ {
289
+ "epoch": 0.15,
290
+ "learning_rate": 0.0002862649260657056,
291
+ "loss": 0.9878,
292
+ "step": 840
293
+ },
294
+ {
295
+ "epoch": 0.16,
296
+ "learning_rate": 0.00028589370785126524,
297
+ "loss": 0.9855,
298
+ "step": 860
299
+ },
300
+ {
301
+ "epoch": 0.16,
302
+ "learning_rate": 0.0002855224896368248,
303
+ "loss": 0.9803,
304
+ "step": 880
305
+ },
306
+ {
307
+ "epoch": 0.17,
308
+ "learning_rate": 0.00028515127142238444,
309
+ "loss": 0.9793,
310
+ "step": 900
311
+ },
312
+ {
313
+ "epoch": 0.17,
314
+ "learning_rate": 0.000284780053207944,
315
+ "loss": 0.9939,
316
+ "step": 920
317
+ },
318
+ {
319
+ "epoch": 0.17,
320
+ "learning_rate": 0.00028440883499350364,
321
+ "loss": 0.9806,
322
+ "step": 940
323
+ },
324
+ {
325
+ "epoch": 0.18,
326
+ "learning_rate": 0.00028403761677906327,
327
+ "loss": 0.9864,
328
+ "step": 960
329
+ },
330
+ {
331
+ "epoch": 0.18,
332
+ "learning_rate": 0.0002836663985646229,
333
+ "loss": 0.9839,
334
+ "step": 980
335
+ },
336
+ {
337
+ "epoch": 0.18,
338
+ "learning_rate": 0.0002832951803501825,
339
+ "loss": 0.9816,
340
+ "step": 1000
341
+ },
342
+ {
343
+ "epoch": 0.18,
344
+ "eval_loss": 0.7965201735496521,
345
+ "eval_runtime": 1.3017,
346
+ "eval_samples_per_second": 0.768,
347
+ "eval_steps_per_second": 0.768,
348
+ "step": 1000
349
+ },
350
+ {
351
+ "epoch": 0.19,
352
+ "learning_rate": 0.0002829239621357421,
353
+ "loss": 0.9848,
354
+ "step": 1020
355
+ },
356
+ {
357
+ "epoch": 0.19,
358
+ "learning_rate": 0.0002825527439213017,
359
+ "loss": 0.9908,
360
+ "step": 1040
361
+ },
362
+ {
363
+ "epoch": 0.2,
364
+ "learning_rate": 0.00028218152570686134,
365
+ "loss": 0.9885,
366
+ "step": 1060
367
+ },
368
+ {
369
+ "epoch": 0.2,
370
+ "learning_rate": 0.0002818103074924209,
371
+ "loss": 0.9772,
372
+ "step": 1080
373
+ },
374
+ {
375
+ "epoch": 0.2,
376
+ "learning_rate": 0.00028143908927798054,
377
+ "loss": 0.9948,
378
+ "step": 1100
379
+ },
380
+ {
381
+ "epoch": 0.21,
382
+ "learning_rate": 0.00028106787106354017,
383
+ "loss": 0.9778,
384
+ "step": 1120
385
+ },
386
+ {
387
+ "epoch": 0.21,
388
+ "learning_rate": 0.00028069665284909974,
389
+ "loss": 0.9909,
390
+ "step": 1140
391
+ },
392
+ {
393
+ "epoch": 0.21,
394
+ "learning_rate": 0.00028032543463465937,
395
+ "loss": 0.9802,
396
+ "step": 1160
397
+ },
398
+ {
399
+ "epoch": 0.22,
400
+ "learning_rate": 0.000279954216420219,
401
+ "loss": 0.9849,
402
+ "step": 1180
403
+ },
404
+ {
405
+ "epoch": 0.22,
406
+ "learning_rate": 0.0002795829982057786,
407
+ "loss": 0.9788,
408
+ "step": 1200
409
+ },
410
+ {
411
+ "epoch": 0.22,
412
+ "eval_loss": 0.8706400990486145,
413
+ "eval_runtime": 1.3141,
414
+ "eval_samples_per_second": 0.761,
415
+ "eval_steps_per_second": 0.761,
416
+ "step": 1200
417
+ },
418
+ {
419
+ "epoch": 0.23,
420
+ "learning_rate": 0.00027921177999133825,
421
+ "loss": 0.9766,
422
+ "step": 1220
423
+ },
424
+ {
425
+ "epoch": 0.23,
426
+ "learning_rate": 0.0002788405617768978,
427
+ "loss": 0.9853,
428
+ "step": 1240
429
+ },
430
+ {
431
+ "epoch": 0.23,
432
+ "learning_rate": 0.00027846934356245745,
433
+ "loss": 0.9814,
434
+ "step": 1260
435
+ },
436
+ {
437
+ "epoch": 0.24,
438
+ "learning_rate": 0.000278098125348017,
439
+ "loss": 0.9747,
440
+ "step": 1280
441
+ },
442
+ {
443
+ "epoch": 0.24,
444
+ "learning_rate": 0.00027772690713357665,
445
+ "loss": 0.9715,
446
+ "step": 1300
447
+ },
448
+ {
449
+ "epoch": 0.24,
450
+ "learning_rate": 0.0002773556889191363,
451
+ "loss": 0.9903,
452
+ "step": 1320
453
+ },
454
+ {
455
+ "epoch": 0.25,
456
+ "learning_rate": 0.0002769844707046959,
457
+ "loss": 0.9862,
458
+ "step": 1340
459
+ },
460
+ {
461
+ "epoch": 0.25,
462
+ "learning_rate": 0.0002766132524902555,
463
+ "loss": 0.9791,
464
+ "step": 1360
465
+ },
466
+ {
467
+ "epoch": 0.25,
468
+ "learning_rate": 0.0002762420342758151,
469
+ "loss": 0.9834,
470
+ "step": 1380
471
+ },
472
+ {
473
+ "epoch": 0.26,
474
+ "learning_rate": 0.0002758708160613747,
475
+ "loss": 0.9752,
476
+ "step": 1400
477
+ },
478
+ {
479
+ "epoch": 0.26,
480
+ "eval_loss": 0.8170408010482788,
481
+ "eval_runtime": 1.2722,
482
+ "eval_samples_per_second": 0.786,
483
+ "eval_steps_per_second": 0.786,
484
+ "step": 1400
485
+ },
486
+ {
487
+ "epoch": 0.26,
488
+ "learning_rate": 0.00027549959784693435,
489
+ "loss": 0.9803,
490
+ "step": 1420
491
+ },
492
+ {
493
+ "epoch": 0.27,
494
+ "learning_rate": 0.0002751283796324939,
495
+ "loss": 0.9968,
496
+ "step": 1440
497
+ },
498
+ {
499
+ "epoch": 0.27,
500
+ "learning_rate": 0.00027475716141805355,
501
+ "loss": 0.98,
502
+ "step": 1460
503
+ },
504
+ {
505
+ "epoch": 0.27,
506
+ "learning_rate": 0.0002743859432036132,
507
+ "loss": 0.9819,
508
+ "step": 1480
509
+ },
510
+ {
511
+ "epoch": 0.28,
512
+ "learning_rate": 0.00027401472498917275,
513
+ "loss": 0.9855,
514
+ "step": 1500
515
+ },
516
+ {
517
+ "epoch": 0.28,
518
+ "learning_rate": 0.0002736435067747324,
519
+ "loss": 0.9852,
520
+ "step": 1520
521
+ },
522
+ {
523
+ "epoch": 0.28,
524
+ "learning_rate": 0.000273272288560292,
525
+ "loss": 0.977,
526
+ "step": 1540
527
+ },
528
+ {
529
+ "epoch": 0.29,
530
+ "learning_rate": 0.00027290107034585163,
531
+ "loss": 0.9918,
532
+ "step": 1560
533
+ },
534
+ {
535
+ "epoch": 0.29,
536
+ "learning_rate": 0.00027252985213141126,
537
+ "loss": 0.9965,
538
+ "step": 1580
539
+ },
540
+ {
541
+ "epoch": 0.3,
542
+ "learning_rate": 0.00027215863391697083,
543
+ "loss": 0.9721,
544
+ "step": 1600
545
+ },
546
+ {
547
+ "epoch": 0.3,
548
+ "eval_loss": 0.8074629902839661,
549
+ "eval_runtime": 1.2561,
550
+ "eval_samples_per_second": 0.796,
551
+ "eval_steps_per_second": 0.796,
552
+ "step": 1600
553
+ },
554
+ {
555
+ "epoch": 0.3,
556
+ "learning_rate": 0.00027178741570253045,
557
+ "loss": 0.984,
558
+ "step": 1620
559
+ },
560
+ {
561
+ "epoch": 0.3,
562
+ "learning_rate": 0.0002714533193095341,
563
+ "loss": 0.9858,
564
+ "step": 1640
565
+ },
566
+ {
567
+ "epoch": 0.31,
568
+ "learning_rate": 0.0002711006620058157,
569
+ "loss": 1.011,
570
+ "step": 1660
571
+ },
572
+ {
573
+ "epoch": 0.31,
574
+ "learning_rate": 0.00027072944379137533,
575
+ "loss": 0.9876,
576
+ "step": 1680
577
+ },
578
+ {
579
+ "epoch": 0.31,
580
+ "learning_rate": 0.00027035822557693495,
581
+ "loss": 0.9941,
582
+ "step": 1700
583
+ },
584
+ {
585
+ "epoch": 0.32,
586
+ "learning_rate": 0.0002699870073624946,
587
+ "loss": 0.9935,
588
+ "step": 1720
589
+ },
590
+ {
591
+ "epoch": 0.32,
592
+ "learning_rate": 0.00026961578914805415,
593
+ "loss": 0.9818,
594
+ "step": 1740
595
+ },
596
+ {
597
+ "epoch": 0.32,
598
+ "learning_rate": 0.0002692445709336138,
599
+ "loss": 0.9876,
600
+ "step": 1760
601
+ },
602
+ {
603
+ "epoch": 0.33,
604
+ "learning_rate": 0.0002688733527191734,
605
+ "loss": 0.9628,
606
+ "step": 1780
607
+ },
608
+ {
609
+ "epoch": 0.33,
610
+ "learning_rate": 0.00026850213450473303,
611
+ "loss": 0.983,
612
+ "step": 1800
613
+ },
614
+ {
615
+ "epoch": 0.33,
616
+ "eval_loss": 0.8400484919548035,
617
+ "eval_runtime": 1.261,
618
+ "eval_samples_per_second": 0.793,
619
+ "eval_steps_per_second": 0.793,
620
+ "step": 1800
621
+ },
622
+ {
623
+ "epoch": 0.34,
624
+ "learning_rate": 0.0002681309162902926,
625
+ "loss": 0.9845,
626
+ "step": 1820
627
+ },
628
+ {
629
+ "epoch": 0.34,
630
+ "learning_rate": 0.00026775969807585223,
631
+ "loss": 0.98,
632
+ "step": 1840
633
+ },
634
+ {
635
+ "epoch": 0.34,
636
+ "learning_rate": 0.0002673884798614118,
637
+ "loss": 0.9649,
638
+ "step": 1860
639
+ },
640
+ {
641
+ "epoch": 0.35,
642
+ "learning_rate": 0.00026701726164697143,
643
+ "loss": 0.9847,
644
+ "step": 1880
645
+ },
646
+ {
647
+ "epoch": 0.35,
648
+ "learning_rate": 0.00026664604343253106,
649
+ "loss": 0.9824,
650
+ "step": 1900
651
+ },
652
+ {
653
+ "epoch": 0.35,
654
+ "learning_rate": 0.0002662748252180907,
655
+ "loss": 0.9777,
656
+ "step": 1920
657
+ },
658
+ {
659
+ "epoch": 0.36,
660
+ "learning_rate": 0.0002659036070036503,
661
+ "loss": 0.9781,
662
+ "step": 1940
663
+ },
664
+ {
665
+ "epoch": 0.36,
666
+ "learning_rate": 0.00026553238878920994,
667
+ "loss": 0.9795,
668
+ "step": 1960
669
+ },
670
+ {
671
+ "epoch": 0.37,
672
+ "learning_rate": 0.0002651611705747695,
673
+ "loss": 0.9875,
674
+ "step": 1980
675
+ },
676
+ {
677
+ "epoch": 0.37,
678
+ "learning_rate": 0.00026478995236032914,
679
+ "loss": 0.9741,
680
+ "step": 2000
681
+ },
682
+ {
683
+ "epoch": 0.37,
684
+ "eval_loss": 0.8619368672370911,
685
+ "eval_runtime": 1.2019,
686
+ "eval_samples_per_second": 0.832,
687
+ "eval_steps_per_second": 0.832,
688
+ "step": 2000
689
+ },
690
+ {
691
+ "epoch": 0.37,
692
+ "learning_rate": 0.0002644187341458887,
693
+ "loss": 0.9841,
694
+ "step": 2020
695
+ },
696
+ {
697
+ "epoch": 0.38,
698
+ "learning_rate": 0.00026404751593144834,
699
+ "loss": 0.9813,
700
+ "step": 2040
701
+ },
702
+ {
703
+ "epoch": 0.38,
704
+ "learning_rate": 0.00026367629771700796,
705
+ "loss": 0.9777,
706
+ "step": 2060
707
+ },
708
+ {
709
+ "epoch": 0.38,
710
+ "learning_rate": 0.0002633050795025676,
711
+ "loss": 0.972,
712
+ "step": 2080
713
+ },
714
+ {
715
+ "epoch": 0.39,
716
+ "learning_rate": 0.00026293386128812716,
717
+ "loss": 0.97,
718
+ "step": 2100
719
+ },
720
+ {
721
+ "epoch": 0.39,
722
+ "learning_rate": 0.0002625626430736868,
723
+ "loss": 0.9866,
724
+ "step": 2120
725
+ },
726
+ {
727
+ "epoch": 0.39,
728
+ "learning_rate": 0.0002621914248592464,
729
+ "loss": 0.9742,
730
+ "step": 2140
731
+ },
732
+ {
733
+ "epoch": 0.4,
734
+ "learning_rate": 0.00026182020664480604,
735
+ "loss": 0.9843,
736
+ "step": 2160
737
+ },
738
+ {
739
+ "epoch": 0.4,
740
+ "learning_rate": 0.0002614489884303656,
741
+ "loss": 0.9616,
742
+ "step": 2180
743
+ },
744
+ {
745
+ "epoch": 0.41,
746
+ "learning_rate": 0.00026107777021592524,
747
+ "loss": 0.9813,
748
+ "step": 2200
749
+ },
750
+ {
751
+ "epoch": 0.41,
752
+ "eval_loss": 0.8606493473052979,
753
+ "eval_runtime": 1.2145,
754
+ "eval_samples_per_second": 0.823,
755
+ "eval_steps_per_second": 0.823,
756
+ "step": 2200
757
+ },
758
+ {
759
+ "epoch": 0.41,
760
+ "learning_rate": 0.0002607065520014848,
761
+ "loss": 0.9716,
762
+ "step": 2220
763
+ },
764
+ {
765
+ "epoch": 0.41,
766
+ "learning_rate": 0.00026033533378704444,
767
+ "loss": 0.9822,
768
+ "step": 2240
769
+ },
770
+ {
771
+ "epoch": 0.42,
772
+ "learning_rate": 0.00025996411557260407,
773
+ "loss": 0.9857,
774
+ "step": 2260
775
+ },
776
+ {
777
+ "epoch": 0.42,
778
+ "learning_rate": 0.0002595928973581637,
779
+ "loss": 0.9977,
780
+ "step": 2280
781
+ },
782
+ {
783
+ "epoch": 0.42,
784
+ "learning_rate": 0.0002592216791437233,
785
+ "loss": 0.9715,
786
+ "step": 2300
787
+ },
788
+ {
789
+ "epoch": 0.43,
790
+ "learning_rate": 0.00025885046092928294,
791
+ "loss": 0.9623,
792
+ "step": 2320
793
+ },
794
+ {
795
+ "epoch": 0.43,
796
+ "learning_rate": 0.0002584792427148425,
797
+ "loss": 0.9637,
798
+ "step": 2340
799
+ },
800
+ {
801
+ "epoch": 0.44,
802
+ "learning_rate": 0.00025810802450040214,
803
+ "loss": 0.978,
804
+ "step": 2360
805
+ },
806
+ {
807
+ "epoch": 0.44,
808
+ "learning_rate": 0.0002577368062859617,
809
+ "loss": 0.9802,
810
+ "step": 2380
811
+ },
812
+ {
813
+ "epoch": 0.44,
814
+ "learning_rate": 0.00025736558807152134,
815
+ "loss": 0.9807,
816
+ "step": 2400
817
+ },
818
+ {
819
+ "epoch": 0.44,
820
+ "eval_loss": 0.8513050079345703,
821
+ "eval_runtime": 1.2201,
822
+ "eval_samples_per_second": 0.82,
823
+ "eval_steps_per_second": 0.82,
824
+ "step": 2400
825
+ },
826
+ {
827
+ "epoch": 0.45,
828
+ "learning_rate": 0.00025699436985708097,
829
+ "loss": 0.9841,
830
+ "step": 2420
831
+ },
832
+ {
833
+ "epoch": 0.45,
834
+ "learning_rate": 0.0002566231516426406,
835
+ "loss": 0.9714,
836
+ "step": 2440
837
+ },
838
+ {
839
+ "epoch": 0.45,
840
+ "learning_rate": 0.00025625193342820017,
841
+ "loss": 0.9693,
842
+ "step": 2460
843
+ },
844
+ {
845
+ "epoch": 0.46,
846
+ "learning_rate": 0.0002558807152137598,
847
+ "loss": 0.9763,
848
+ "step": 2480
849
+ },
850
+ {
851
+ "epoch": 0.46,
852
+ "learning_rate": 0.0002555094969993194,
853
+ "loss": 0.9779,
854
+ "step": 2500
855
+ },
856
+ {
857
+ "epoch": 0.46,
858
+ "learning_rate": 0.00025513827878487905,
859
+ "loss": 0.9736,
860
+ "step": 2520
861
+ },
862
+ {
863
+ "epoch": 0.47,
864
+ "learning_rate": 0.0002547670605704386,
865
+ "loss": 0.9788,
866
+ "step": 2540
867
+ },
868
+ {
869
+ "epoch": 0.47,
870
+ "learning_rate": 0.00025439584235599825,
871
+ "loss": 0.9789,
872
+ "step": 2560
873
+ },
874
+ {
875
+ "epoch": 0.48,
876
+ "learning_rate": 0.0002540246241415578,
877
+ "loss": 0.9729,
878
+ "step": 2580
879
+ },
880
+ {
881
+ "epoch": 0.48,
882
+ "learning_rate": 0.00025365340592711745,
883
+ "loss": 0.9752,
884
+ "step": 2600
885
+ },
886
+ {
887
+ "epoch": 0.48,
888
+ "eval_loss": 0.8586015105247498,
889
+ "eval_runtime": 1.2961,
890
+ "eval_samples_per_second": 0.772,
891
+ "eval_steps_per_second": 0.772,
892
+ "step": 2600
893
+ },
894
+ {
895
+ "epoch": 0.48,
896
+ "learning_rate": 0.00025328218771267707,
897
+ "loss": 0.9734,
898
+ "step": 2620
899
+ },
900
+ {
901
+ "epoch": 0.49,
902
+ "learning_rate": 0.0002529109694982367,
903
+ "loss": 0.9767,
904
+ "step": 2640
905
+ },
906
+ {
907
+ "epoch": 0.49,
908
+ "learning_rate": 0.0002525397512837963,
909
+ "loss": 0.962,
910
+ "step": 2660
911
+ },
912
+ {
913
+ "epoch": 0.49,
914
+ "learning_rate": 0.00025216853306935595,
915
+ "loss": 0.9706,
916
+ "step": 2680
917
+ },
918
+ {
919
+ "epoch": 0.5,
920
+ "learning_rate": 0.0002517973148549155,
921
+ "loss": 0.975,
922
+ "step": 2700
923
+ },
924
+ {
925
+ "epoch": 0.5,
926
+ "learning_rate": 0.00025142609664047515,
927
+ "loss": 0.9653,
928
+ "step": 2720
929
+ },
930
+ {
931
+ "epoch": 0.51,
932
+ "learning_rate": 0.0002510548784260347,
933
+ "loss": 0.976,
934
+ "step": 2740
935
+ },
936
+ {
937
+ "epoch": 0.51,
938
+ "learning_rate": 0.00025068366021159435,
939
+ "loss": 0.9768,
940
+ "step": 2760
941
+ },
942
+ {
943
+ "epoch": 0.51,
944
+ "learning_rate": 0.000250312441997154,
945
+ "loss": 0.9764,
946
+ "step": 2780
947
+ },
948
+ {
949
+ "epoch": 0.52,
950
+ "learning_rate": 0.0002499412237827136,
951
+ "loss": 0.9741,
952
+ "step": 2800
953
+ },
954
+ {
955
+ "epoch": 0.52,
956
+ "eval_loss": 0.8531299829483032,
957
+ "eval_runtime": 1.2788,
958
+ "eval_samples_per_second": 0.782,
959
+ "eval_steps_per_second": 0.782,
960
+ "step": 2800
961
+ },
962
+ {
963
+ "epoch": 0.52,
964
+ "learning_rate": 0.0002495700055682732,
965
+ "loss": 0.9786,
966
+ "step": 2820
967
+ },
968
+ {
969
+ "epoch": 0.52,
970
+ "learning_rate": 0.0002491987873538328,
971
+ "loss": 0.9737,
972
+ "step": 2840
973
+ },
974
+ {
975
+ "epoch": 0.53,
976
+ "learning_rate": 0.00024882756913939243,
977
+ "loss": 0.9648,
978
+ "step": 2860
979
+ },
980
+ {
981
+ "epoch": 0.53,
982
+ "learning_rate": 0.00024845635092495206,
983
+ "loss": 0.9717,
984
+ "step": 2880
985
+ },
986
+ {
987
+ "epoch": 0.53,
988
+ "learning_rate": 0.00024808513271051163,
989
+ "loss": 0.9706,
990
+ "step": 2900
991
+ },
992
+ {
993
+ "epoch": 0.54,
994
+ "learning_rate": 0.00024771391449607125,
995
+ "loss": 0.9662,
996
+ "step": 2920
997
+ },
998
+ {
999
+ "epoch": 0.54,
1000
+ "learning_rate": 0.0002473426962816308,
1001
+ "loss": 0.9719,
1002
+ "step": 2940
1003
+ },
1004
+ {
1005
+ "epoch": 0.55,
1006
+ "learning_rate": 0.00024697147806719045,
1007
+ "loss": 0.9729,
1008
+ "step": 2960
1009
+ },
1010
+ {
1011
+ "epoch": 0.55,
1012
+ "learning_rate": 0.0002466002598527501,
1013
+ "loss": 0.9679,
1014
+ "step": 2980
1015
+ },
1016
+ {
1017
+ "epoch": 0.55,
1018
+ "learning_rate": 0.0002462290416383097,
1019
+ "loss": 0.9782,
1020
+ "step": 3000
1021
+ },
1022
+ {
1023
+ "epoch": 0.55,
1024
+ "eval_loss": 0.8357999324798584,
1025
+ "eval_runtime": 1.3862,
1026
+ "eval_samples_per_second": 0.721,
1027
+ "eval_steps_per_second": 0.721,
1028
+ "step": 3000
1029
+ },
1030
+ {
1031
+ "epoch": 0.56,
1032
+ "learning_rate": 0.00024585782342386933,
1033
+ "loss": 0.9636,
1034
+ "step": 3020
1035
+ },
1036
+ {
1037
+ "epoch": 0.56,
1038
+ "learning_rate": 0.00024548660520942896,
1039
+ "loss": 0.9779,
1040
+ "step": 3040
1041
+ },
1042
+ {
1043
+ "epoch": 0.56,
1044
+ "learning_rate": 0.00024511538699498853,
1045
+ "loss": 0.9718,
1046
+ "step": 3060
1047
+ },
1048
+ {
1049
+ "epoch": 0.57,
1050
+ "learning_rate": 0.00024474416878054816,
1051
+ "loss": 0.9756,
1052
+ "step": 3080
1053
+ },
1054
+ {
1055
+ "epoch": 0.57,
1056
+ "learning_rate": 0.00024437295056610773,
1057
+ "loss": 0.983,
1058
+ "step": 3100
1059
+ },
1060
+ {
1061
+ "epoch": 0.58,
1062
+ "learning_rate": 0.00024400173235166736,
1063
+ "loss": 0.9735,
1064
+ "step": 3120
1065
+ },
1066
+ {
1067
+ "epoch": 0.58,
1068
+ "learning_rate": 0.00024363051413722698,
1069
+ "loss": 0.976,
1070
+ "step": 3140
1071
+ },
1072
+ {
1073
+ "epoch": 0.58,
1074
+ "learning_rate": 0.0002432592959227866,
1075
+ "loss": 0.9691,
1076
+ "step": 3160
1077
+ },
1078
+ {
1079
+ "epoch": 0.59,
1080
+ "learning_rate": 0.0002428880777083462,
1081
+ "loss": 0.9644,
1082
+ "step": 3180
1083
+ },
1084
+ {
1085
+ "epoch": 0.59,
1086
+ "learning_rate": 0.0002425168594939058,
1087
+ "loss": 0.9768,
1088
+ "step": 3200
1089
+ },
1090
+ {
1091
+ "epoch": 0.59,
1092
+ "eval_loss": 0.8302448987960815,
1093
+ "eval_runtime": 1.362,
1094
+ "eval_samples_per_second": 0.734,
1095
+ "eval_steps_per_second": 0.734,
1096
+ "step": 3200
1097
+ },
1098
+ {
1099
+ "epoch": 0.59,
1100
+ "learning_rate": 0.0002421456412794654,
1101
+ "loss": 0.9709,
1102
+ "step": 3220
1103
+ },
1104
+ {
1105
+ "epoch": 0.6,
1106
+ "learning_rate": 0.00024177442306502504,
1107
+ "loss": 0.9696,
1108
+ "step": 3240
1109
+ },
1110
+ {
1111
+ "epoch": 0.6,
1112
+ "learning_rate": 0.00024140320485058466,
1113
+ "loss": 0.9669,
1114
+ "step": 3260
1115
+ },
1116
+ {
1117
+ "epoch": 0.6,
1118
+ "learning_rate": 0.00024103198663614426,
1119
+ "loss": 0.965,
1120
+ "step": 3280
1121
+ },
1122
+ {
1123
+ "epoch": 0.61,
1124
+ "learning_rate": 0.0002406607684217039,
1125
+ "loss": 0.9757,
1126
+ "step": 3300
1127
+ },
1128
+ {
1129
+ "epoch": 0.61,
1130
+ "learning_rate": 0.00024028955020726346,
1131
+ "loss": 0.9717,
1132
+ "step": 3320
1133
+ },
1134
+ {
1135
+ "epoch": 0.62,
1136
+ "learning_rate": 0.0002399183319928231,
1137
+ "loss": 0.963,
1138
+ "step": 3340
1139
+ },
1140
+ {
1141
+ "epoch": 0.62,
1142
+ "learning_rate": 0.00023954711377838271,
1143
+ "loss": 0.9519,
1144
+ "step": 3360
1145
+ },
1146
+ {
1147
+ "epoch": 0.62,
1148
+ "learning_rate": 0.0002391758955639423,
1149
+ "loss": 0.964,
1150
+ "step": 3380
1151
+ },
1152
+ {
1153
+ "epoch": 0.63,
1154
+ "learning_rate": 0.00023880467734950194,
1155
+ "loss": 0.967,
1156
+ "step": 3400
1157
+ },
1158
+ {
1159
+ "epoch": 0.63,
1160
+ "eval_loss": 0.87740159034729,
1161
+ "eval_runtime": 1.4495,
1162
+ "eval_samples_per_second": 0.69,
1163
+ "eval_steps_per_second": 0.69,
1164
+ "step": 3400
1165
+ },
1166
+ {
1167
+ "epoch": 0.63,
1168
+ "learning_rate": 0.00023843345913506157,
1169
+ "loss": 0.9666,
1170
+ "step": 3420
1171
+ },
1172
+ {
1173
+ "epoch": 0.63,
1174
+ "learning_rate": 0.00023806224092062114,
1175
+ "loss": 0.9616,
1176
+ "step": 3440
1177
+ },
1178
+ {
1179
+ "epoch": 0.64,
1180
+ "learning_rate": 0.00023769102270618077,
1181
+ "loss": 0.9663,
1182
+ "step": 3460
1183
+ },
1184
+ {
1185
+ "epoch": 0.64,
1186
+ "learning_rate": 0.00023731980449174036,
1187
+ "loss": 0.9665,
1188
+ "step": 3480
1189
+ },
1190
+ {
1191
+ "epoch": 0.65,
1192
+ "learning_rate": 0.0002369485862773,
1193
+ "loss": 0.9662,
1194
+ "step": 3500
1195
+ },
1196
+ {
1197
+ "epoch": 0.65,
1198
+ "learning_rate": 0.00023657736806285962,
1199
+ "loss": 0.9656,
1200
+ "step": 3520
1201
+ },
1202
+ {
1203
+ "epoch": 0.65,
1204
+ "learning_rate": 0.00023620614984841922,
1205
+ "loss": 0.9587,
1206
+ "step": 3540
1207
+ },
1208
+ {
1209
+ "epoch": 0.66,
1210
+ "learning_rate": 0.00023583493163397882,
1211
+ "loss": 0.9639,
1212
+ "step": 3560
1213
+ },
1214
+ {
1215
+ "epoch": 0.66,
1216
+ "learning_rate": 0.00023546371341953842,
1217
+ "loss": 0.9727,
1218
+ "step": 3580
1219
+ },
1220
+ {
1221
+ "epoch": 0.66,
1222
+ "learning_rate": 0.00023509249520509804,
1223
+ "loss": 0.954,
1224
+ "step": 3600
1225
+ },
1226
+ {
1227
+ "epoch": 0.66,
1228
+ "eval_loss": 0.8274036645889282,
1229
+ "eval_runtime": 1.399,
1230
+ "eval_samples_per_second": 0.715,
1231
+ "eval_steps_per_second": 0.715,
1232
+ "step": 3600
1233
+ },
1234
+ {
1235
+ "epoch": 0.67,
1236
+ "learning_rate": 0.00023472127699065767,
1237
+ "loss": 0.9614,
1238
+ "step": 3620
1239
+ },
1240
+ {
1241
+ "epoch": 0.67,
1242
+ "learning_rate": 0.00023435005877621727,
1243
+ "loss": 0.9573,
1244
+ "step": 3640
1245
+ },
1246
+ {
1247
+ "epoch": 0.68,
1248
+ "learning_rate": 0.0002339788405617769,
1249
+ "loss": 0.9606,
1250
+ "step": 3660
1251
+ },
1252
+ {
1253
+ "epoch": 0.68,
1254
+ "learning_rate": 0.00023360762234733647,
1255
+ "loss": 0.962,
1256
+ "step": 3680
1257
+ },
1258
+ {
1259
+ "epoch": 0.68,
1260
+ "learning_rate": 0.0002332364041328961,
1261
+ "loss": 0.9591,
1262
+ "step": 3700
1263
+ },
1264
+ {
1265
+ "epoch": 0.69,
1266
+ "learning_rate": 0.00023286518591845572,
1267
+ "loss": 0.9547,
1268
+ "step": 3720
1269
+ },
1270
+ {
1271
+ "epoch": 0.69,
1272
+ "learning_rate": 0.00023249396770401532,
1273
+ "loss": 0.9627,
1274
+ "step": 3740
1275
+ },
1276
+ {
1277
+ "epoch": 0.69,
1278
+ "learning_rate": 0.00023212274948957495,
1279
+ "loss": 0.9578,
1280
+ "step": 3760
1281
+ },
1282
+ {
1283
+ "epoch": 0.7,
1284
+ "learning_rate": 0.00023175153127513457,
1285
+ "loss": 0.9623,
1286
+ "step": 3780
1287
+ },
1288
+ {
1289
+ "epoch": 0.7,
1290
+ "learning_rate": 0.00023138031306069415,
1291
+ "loss": 0.946,
1292
+ "step": 3800
1293
+ },
1294
+ {
1295
+ "epoch": 0.7,
1296
+ "eval_loss": 0.8205245137214661,
1297
+ "eval_runtime": 1.3808,
1298
+ "eval_samples_per_second": 0.724,
1299
+ "eval_steps_per_second": 0.724,
1300
+ "step": 3800
1301
+ },
1302
+ {
1303
+ "epoch": 0.7,
1304
+ "learning_rate": 0.00023100909484625377,
1305
+ "loss": 0.9734,
1306
+ "step": 3820
1307
+ },
1308
+ {
1309
+ "epoch": 0.71,
1310
+ "learning_rate": 0.00023063787663181337,
1311
+ "loss": 0.964,
1312
+ "step": 3840
1313
+ },
1314
+ {
1315
+ "epoch": 0.71,
1316
+ "learning_rate": 0.000230266658417373,
1317
+ "loss": 0.9576,
1318
+ "step": 3860
1319
+ },
1320
+ {
1321
+ "epoch": 0.72,
1322
+ "learning_rate": 0.00022989544020293263,
1323
+ "loss": 0.9512,
1324
+ "step": 3880
1325
+ },
1326
+ {
1327
+ "epoch": 0.72,
1328
+ "learning_rate": 0.00022952422198849222,
1329
+ "loss": 0.9555,
1330
+ "step": 3900
1331
+ },
1332
+ {
1333
+ "epoch": 0.72,
1334
+ "learning_rate": 0.00022915300377405182,
1335
+ "loss": 0.95,
1336
+ "step": 3920
1337
+ },
1338
+ {
1339
+ "epoch": 0.73,
1340
+ "learning_rate": 0.00022878178555961142,
1341
+ "loss": 0.9617,
1342
+ "step": 3940
1343
+ },
1344
+ {
1345
+ "epoch": 0.73,
1346
+ "learning_rate": 0.00022841056734517105,
1347
+ "loss": 0.9636,
1348
+ "step": 3960
1349
+ },
1350
+ {
1351
+ "epoch": 0.73,
1352
+ "learning_rate": 0.00022803934913073068,
1353
+ "loss": 0.955,
1354
+ "step": 3980
1355
+ },
1356
+ {
1357
+ "epoch": 0.74,
1358
+ "learning_rate": 0.00022766813091629028,
1359
+ "loss": 0.9516,
1360
+ "step": 4000
1361
+ },
1362
+ {
1363
+ "epoch": 0.74,
1364
+ "eval_loss": 0.8116470575332642,
1365
+ "eval_runtime": 1.3156,
1366
+ "eval_samples_per_second": 0.76,
1367
+ "eval_steps_per_second": 0.76,
1368
+ "step": 4000
1369
+ },
1370
+ {
1371
+ "epoch": 0.74,
1372
+ "learning_rate": 0.0002272969127018499,
1373
+ "loss": 0.9517,
1374
+ "step": 4020
1375
+ },
1376
+ {
1377
+ "epoch": 0.75,
1378
+ "learning_rate": 0.00022692569448740948,
1379
+ "loss": 0.9551,
1380
+ "step": 4040
1381
+ },
1382
+ {
1383
+ "epoch": 0.75,
1384
+ "learning_rate": 0.0002265544762729691,
1385
+ "loss": 0.9578,
1386
+ "step": 4060
1387
+ },
1388
+ {
1389
+ "epoch": 0.75,
1390
+ "learning_rate": 0.00022618325805852873,
1391
+ "loss": 0.9613,
1392
+ "step": 4080
1393
+ },
1394
+ {
1395
+ "epoch": 0.76,
1396
+ "learning_rate": 0.00022581203984408833,
1397
+ "loss": 0.9526,
1398
+ "step": 4100
1399
+ },
1400
+ {
1401
+ "epoch": 0.76,
1402
+ "learning_rate": 0.00022544082162964795,
1403
+ "loss": 0.966,
1404
+ "step": 4120
1405
+ },
1406
+ {
1407
+ "epoch": 0.76,
1408
+ "learning_rate": 0.00022506960341520755,
1409
+ "loss": 0.9504,
1410
+ "step": 4140
1411
+ },
1412
+ {
1413
+ "epoch": 0.77,
1414
+ "learning_rate": 0.00022469838520076715,
1415
+ "loss": 0.9458,
1416
+ "step": 4160
1417
+ },
1418
+ {
1419
+ "epoch": 0.77,
1420
+ "learning_rate": 0.00022432716698632678,
1421
+ "loss": 0.9693,
1422
+ "step": 4180
1423
+ },
1424
+ {
1425
+ "epoch": 0.77,
1426
+ "learning_rate": 0.00022395594877188638,
1427
+ "loss": 0.9566,
1428
+ "step": 4200
1429
+ },
1430
+ {
1431
+ "epoch": 0.77,
1432
+ "eval_loss": 0.8783697485923767,
1433
+ "eval_runtime": 1.3242,
1434
+ "eval_samples_per_second": 0.755,
1435
+ "eval_steps_per_second": 0.755,
1436
+ "step": 4200
1437
+ },
1438
+ {
1439
+ "epoch": 0.78,
1440
+ "learning_rate": 0.000223584730557446,
1441
+ "loss": 0.9736,
1442
+ "step": 4220
1443
+ },
1444
+ {
1445
+ "epoch": 0.78,
1446
+ "learning_rate": 0.0002232135123430056,
1447
+ "loss": 0.9454,
1448
+ "step": 4240
1449
+ },
1450
+ {
1451
+ "epoch": 0.79,
1452
+ "learning_rate": 0.00022284229412856523,
1453
+ "loss": 0.9547,
1454
+ "step": 4260
1455
+ },
1456
+ {
1457
+ "epoch": 0.79,
1458
+ "learning_rate": 0.00022247107591412483,
1459
+ "loss": 0.9521,
1460
+ "step": 4280
1461
+ },
1462
+ {
1463
+ "epoch": 0.79,
1464
+ "learning_rate": 0.00022209985769968443,
1465
+ "loss": 0.9546,
1466
+ "step": 4300
1467
+ },
1468
+ {
1469
+ "epoch": 0.8,
1470
+ "learning_rate": 0.00022172863948524406,
1471
+ "loss": 0.9557,
1472
+ "step": 4320
1473
+ },
1474
+ {
1475
+ "epoch": 0.8,
1476
+ "learning_rate": 0.00022135742127080366,
1477
+ "loss": 0.947,
1478
+ "step": 4340
1479
+ },
1480
+ {
1481
+ "epoch": 0.8,
1482
+ "learning_rate": 0.00022098620305636328,
1483
+ "loss": 0.9576,
1484
+ "step": 4360
1485
+ },
1486
+ {
1487
+ "epoch": 0.81,
1488
+ "learning_rate": 0.0002206149848419229,
1489
+ "loss": 0.9506,
1490
+ "step": 4380
1491
+ },
1492
+ {
1493
+ "epoch": 0.81,
1494
+ "learning_rate": 0.00022024376662748248,
1495
+ "loss": 0.9537,
1496
+ "step": 4400
1497
+ },
1498
+ {
1499
+ "epoch": 0.81,
1500
+ "eval_loss": 0.8563671112060547,
1501
+ "eval_runtime": 1.3666,
1502
+ "eval_samples_per_second": 0.732,
1503
+ "eval_steps_per_second": 0.732,
1504
+ "step": 4400
1505
+ },
1506
+ {
1507
+ "epoch": 0.82,
1508
+ "learning_rate": 0.0002198725484130421,
1509
+ "loss": 0.955,
1510
+ "step": 4420
1511
+ },
1512
+ {
1513
+ "epoch": 0.82,
1514
+ "learning_rate": 0.0002195013301986017,
1515
+ "loss": 0.9653,
1516
+ "step": 4440
1517
+ },
1518
+ {
1519
+ "epoch": 0.82,
1520
+ "learning_rate": 0.00021913011198416133,
1521
+ "loss": 0.9601,
1522
+ "step": 4460
1523
+ },
1524
+ {
1525
+ "epoch": 0.83,
1526
+ "learning_rate": 0.00021875889376972096,
1527
+ "loss": 0.9589,
1528
+ "step": 4480
1529
+ },
1530
+ {
1531
+ "epoch": 0.83,
1532
+ "learning_rate": 0.00021838767555528056,
1533
+ "loss": 0.9448,
1534
+ "step": 4500
1535
+ },
1536
+ {
1537
+ "epoch": 0.83,
1538
+ "learning_rate": 0.00021801645734084016,
1539
+ "loss": 0.9538,
1540
+ "step": 4520
1541
+ },
1542
+ {
1543
+ "epoch": 0.84,
1544
+ "learning_rate": 0.00021764523912639976,
1545
+ "loss": 0.958,
1546
+ "step": 4540
1547
+ },
1548
+ {
1549
+ "epoch": 0.84,
1550
+ "learning_rate": 0.0002172740209119594,
1551
+ "loss": 0.9466,
1552
+ "step": 4560
1553
+ },
1554
+ {
1555
+ "epoch": 0.84,
1556
+ "learning_rate": 0.000216902802697519,
1557
+ "loss": 0.9544,
1558
+ "step": 4580
1559
+ },
1560
+ {
1561
+ "epoch": 0.85,
1562
+ "learning_rate": 0.0002165315844830786,
1563
+ "loss": 0.9456,
1564
+ "step": 4600
1565
+ },
1566
+ {
1567
+ "epoch": 0.85,
1568
+ "eval_loss": 0.8733828663825989,
1569
+ "eval_runtime": 1.3225,
1570
+ "eval_samples_per_second": 0.756,
1571
+ "eval_steps_per_second": 0.756,
1572
+ "step": 4600
1573
+ },
1574
+ {
1575
+ "epoch": 0.85,
1576
+ "learning_rate": 0.00021616036626863824,
1577
+ "loss": 0.956,
1578
+ "step": 4620
1579
+ },
1580
+ {
1581
+ "epoch": 0.86,
1582
+ "learning_rate": 0.0002157891480541978,
1583
+ "loss": 0.9614,
1584
+ "step": 4640
1585
+ },
1586
+ {
1587
+ "epoch": 0.86,
1588
+ "learning_rate": 0.00021541792983975744,
1589
+ "loss": 0.9568,
1590
+ "step": 4660
1591
+ },
1592
+ {
1593
+ "epoch": 0.86,
1594
+ "learning_rate": 0.00021504671162531706,
1595
+ "loss": 0.9606,
1596
+ "step": 4680
1597
+ },
1598
+ {
1599
+ "epoch": 0.87,
1600
+ "learning_rate": 0.00021467549341087666,
1601
+ "loss": 0.9375,
1602
+ "step": 4700
1603
+ },
1604
+ {
1605
+ "epoch": 0.87,
1606
+ "learning_rate": 0.0002143042751964363,
1607
+ "loss": 0.938,
1608
+ "step": 4720
1609
+ },
1610
+ {
1611
+ "epoch": 0.87,
1612
+ "learning_rate": 0.00021393305698199592,
1613
+ "loss": 0.9647,
1614
+ "step": 4740
1615
+ },
1616
+ {
1617
+ "epoch": 0.88,
1618
+ "learning_rate": 0.0002135618387675555,
1619
+ "loss": 0.96,
1620
+ "step": 4760
1621
+ },
1622
+ {
1623
+ "epoch": 0.88,
1624
+ "learning_rate": 0.00021319062055311512,
1625
+ "loss": 0.9458,
1626
+ "step": 4780
1627
+ },
1628
+ {
1629
+ "epoch": 0.89,
1630
+ "learning_rate": 0.00021281940233867472,
1631
+ "loss": 0.9492,
1632
+ "step": 4800
1633
+ },
1634
+ {
1635
+ "epoch": 0.89,
1636
+ "eval_loss": 0.869820773601532,
1637
+ "eval_runtime": 1.3761,
1638
+ "eval_samples_per_second": 0.727,
1639
+ "eval_steps_per_second": 0.727,
1640
+ "step": 4800
1641
+ },
1642
+ {
1643
+ "epoch": 0.89,
1644
+ "learning_rate": 0.00021244818412423434,
1645
+ "loss": 0.9536,
1646
+ "step": 4820
1647
+ },
1648
+ {
1649
+ "epoch": 0.89,
1650
+ "learning_rate": 0.00021207696590979397,
1651
+ "loss": 0.9522,
1652
+ "step": 4840
1653
+ },
1654
+ {
1655
+ "epoch": 0.9,
1656
+ "learning_rate": 0.00021170574769535357,
1657
+ "loss": 0.9501,
1658
+ "step": 4860
1659
+ },
1660
+ {
1661
+ "epoch": 0.9,
1662
+ "learning_rate": 0.00021133452948091317,
1663
+ "loss": 0.958,
1664
+ "step": 4880
1665
+ },
1666
+ {
1667
+ "epoch": 0.9,
1668
+ "learning_rate": 0.00021096331126647277,
1669
+ "loss": 0.9585,
1670
+ "step": 4900
1671
+ },
1672
+ {
1673
+ "epoch": 0.91,
1674
+ "learning_rate": 0.0002105920930520324,
1675
+ "loss": 0.9529,
1676
+ "step": 4920
1677
+ },
1678
+ {
1679
+ "epoch": 0.91,
1680
+ "learning_rate": 0.00021022087483759202,
1681
+ "loss": 0.9463,
1682
+ "step": 4940
1683
+ },
1684
+ {
1685
+ "epoch": 0.91,
1686
+ "learning_rate": 0.00020984965662315162,
1687
+ "loss": 0.9558,
1688
+ "step": 4960
1689
+ },
1690
+ {
1691
+ "epoch": 0.92,
1692
+ "learning_rate": 0.00020947843840871125,
1693
+ "loss": 0.9476,
1694
+ "step": 4980
1695
+ },
1696
+ {
1697
+ "epoch": 0.92,
1698
+ "learning_rate": 0.00020910722019427087,
1699
+ "loss": 0.9423,
1700
+ "step": 5000
1701
+ },
1702
+ {
1703
+ "epoch": 0.92,
1704
+ "eval_loss": 0.8774290084838867,
1705
+ "eval_runtime": 1.3657,
1706
+ "eval_samples_per_second": 0.732,
1707
+ "eval_steps_per_second": 0.732,
1708
+ "step": 5000
1709
+ },
1710
+ {
1711
+ "epoch": 0.93,
1712
+ "learning_rate": 0.00020873600197983045,
1713
+ "loss": 0.9463,
1714
+ "step": 5020
1715
+ },
1716
+ {
1717
+ "epoch": 0.93,
1718
+ "learning_rate": 0.00020836478376539007,
1719
+ "loss": 0.9389,
1720
+ "step": 5040
1721
+ },
1722
+ {
1723
+ "epoch": 0.93,
1724
+ "learning_rate": 0.00020799356555094967,
1725
+ "loss": 0.9443,
1726
+ "step": 5060
1727
+ },
1728
+ {
1729
+ "epoch": 0.94,
1730
+ "learning_rate": 0.0002076223473365093,
1731
+ "loss": 0.9412,
1732
+ "step": 5080
1733
+ },
1734
+ {
1735
+ "epoch": 0.94,
1736
+ "learning_rate": 0.00020725112912206892,
1737
+ "loss": 0.9417,
1738
+ "step": 5100
1739
+ },
1740
+ {
1741
+ "epoch": 0.94,
1742
+ "learning_rate": 0.00020687991090762852,
1743
+ "loss": 0.9485,
1744
+ "step": 5120
1745
+ },
1746
+ {
1747
+ "epoch": 0.95,
1748
+ "learning_rate": 0.00020650869269318812,
1749
+ "loss": 0.9479,
1750
+ "step": 5140
1751
+ },
1752
+ {
1753
+ "epoch": 0.95,
1754
+ "learning_rate": 0.00020613747447874772,
1755
+ "loss": 0.9541,
1756
+ "step": 5160
1757
+ },
1758
+ {
1759
+ "epoch": 0.96,
1760
+ "learning_rate": 0.00020576625626430735,
1761
+ "loss": 0.9472,
1762
+ "step": 5180
1763
+ },
1764
+ {
1765
+ "epoch": 0.96,
1766
+ "learning_rate": 0.00020539503804986698,
1767
+ "loss": 0.9382,
1768
+ "step": 5200
1769
+ },
1770
+ {
1771
+ "epoch": 0.96,
1772
+ "eval_loss": 0.9030184745788574,
1773
+ "eval_runtime": 1.4724,
1774
+ "eval_samples_per_second": 0.679,
1775
+ "eval_steps_per_second": 0.679,
1776
+ "step": 5200
1777
+ },
1778
+ {
1779
+ "epoch": 0.96,
1780
+ "learning_rate": 0.00020502381983542658,
1781
+ "loss": 0.9323,
1782
+ "step": 5220
1783
+ },
1784
+ {
1785
+ "epoch": 0.97,
1786
+ "learning_rate": 0.0002046526016209862,
1787
+ "loss": 0.9435,
1788
+ "step": 5240
1789
+ },
1790
+ {
1791
+ "epoch": 0.97,
1792
+ "learning_rate": 0.00020428138340654577,
1793
+ "loss": 0.9417,
1794
+ "step": 5260
1795
+ },
1796
+ {
1797
+ "epoch": 0.97,
1798
+ "learning_rate": 0.0002039101651921054,
1799
+ "loss": 0.9479,
1800
+ "step": 5280
1801
+ },
1802
+ {
1803
+ "epoch": 0.98,
1804
+ "learning_rate": 0.00020353894697766503,
1805
+ "loss": 0.9418,
1806
+ "step": 5300
1807
+ },
1808
+ {
1809
+ "epoch": 0.98,
1810
+ "learning_rate": 0.00020316772876322463,
1811
+ "loss": 0.946,
1812
+ "step": 5320
1813
+ },
1814
+ {
1815
+ "epoch": 0.98,
1816
+ "learning_rate": 0.00020279651054878425,
1817
+ "loss": 0.9475,
1818
+ "step": 5340
1819
+ },
1820
+ {
1821
+ "epoch": 0.99,
1822
+ "learning_rate": 0.00020242529233434388,
1823
+ "loss": 0.94,
1824
+ "step": 5360
1825
+ },
1826
+ {
1827
+ "epoch": 0.99,
1828
+ "learning_rate": 0.00020205407411990345,
1829
+ "loss": 0.943,
1830
+ "step": 5380
1831
+ },
1832
+ {
1833
+ "epoch": 1.0,
1834
+ "learning_rate": 0.00020168285590546308,
1835
+ "loss": 0.9358,
1836
+ "step": 5400
1837
+ },
1838
+ {
1839
+ "epoch": 1.0,
1840
+ "eval_loss": 0.8716119527816772,
1841
+ "eval_runtime": 1.3296,
1842
+ "eval_samples_per_second": 0.752,
1843
+ "eval_steps_per_second": 0.752,
1844
+ "step": 5400
1845
+ },
1846
+ {
1847
+ "epoch": 1.0,
1848
+ "learning_rate": 0.00020131163769102268,
1849
+ "loss": 0.9351,
1850
+ "step": 5420
1851
+ },
1852
+ {
1853
+ "epoch": 1.0,
1854
+ "learning_rate": 0.0002009404194765823,
1855
+ "loss": 0.9412,
1856
+ "step": 5440
1857
+ },
1858
+ {
1859
+ "epoch": 1.01,
1860
+ "learning_rate": 0.00020056920126214193,
1861
+ "loss": 0.9346,
1862
+ "step": 5460
1863
+ },
1864
+ {
1865
+ "epoch": 1.01,
1866
+ "learning_rate": 0.00020019798304770153,
1867
+ "loss": 0.9427,
1868
+ "step": 5480
1869
+ },
1870
+ {
1871
+ "epoch": 1.01,
1872
+ "learning_rate": 0.00019982676483326113,
1873
+ "loss": 0.9553,
1874
+ "step": 5500
1875
+ },
1876
+ {
1877
+ "epoch": 1.02,
1878
+ "learning_rate": 0.00019945554661882073,
1879
+ "loss": 0.9443,
1880
+ "step": 5520
1881
+ },
1882
+ {
1883
+ "epoch": 1.02,
1884
+ "learning_rate": 0.00019908432840438036,
1885
+ "loss": 0.9446,
1886
+ "step": 5540
1887
+ },
1888
+ {
1889
+ "epoch": 1.03,
1890
+ "learning_rate": 0.00019871311018993998,
1891
+ "loss": 0.9319,
1892
+ "step": 5560
1893
+ },
1894
+ {
1895
+ "epoch": 1.03,
1896
+ "learning_rate": 0.00019834189197549958,
1897
+ "loss": 0.9333,
1898
+ "step": 5580
1899
+ },
1900
+ {
1901
+ "epoch": 1.03,
1902
+ "learning_rate": 0.0001979706737610592,
1903
+ "loss": 0.9405,
1904
+ "step": 5600
1905
+ },
1906
+ {
1907
+ "epoch": 1.03,
1908
+ "eval_loss": 0.8719614744186401,
1909
+ "eval_runtime": 1.2911,
1910
+ "eval_samples_per_second": 0.775,
1911
+ "eval_steps_per_second": 0.775,
1912
+ "step": 5600
1913
+ },
1914
+ {
1915
+ "epoch": 1.04,
1916
+ "learning_rate": 0.00019759945554661878,
1917
+ "loss": 0.9367,
1918
+ "step": 5620
1919
+ },
1920
+ {
1921
+ "epoch": 1.04,
1922
+ "learning_rate": 0.0001972282373321784,
1923
+ "loss": 0.9352,
1924
+ "step": 5640
1925
+ },
1926
+ {
1927
+ "epoch": 1.04,
1928
+ "learning_rate": 0.00019685701911773803,
1929
+ "loss": 0.9422,
1930
+ "step": 5660
1931
+ },
1932
+ {
1933
+ "epoch": 1.05,
1934
+ "learning_rate": 0.00019648580090329763,
1935
+ "loss": 0.9543,
1936
+ "step": 5680
1937
+ },
1938
+ {
1939
+ "epoch": 1.05,
1940
+ "learning_rate": 0.00019611458268885726,
1941
+ "loss": 0.9308,
1942
+ "step": 5700
1943
+ },
1944
+ {
1945
+ "epoch": 1.05,
1946
+ "learning_rate": 0.0001957433644744169,
1947
+ "loss": 0.9471,
1948
+ "step": 5720
1949
+ },
1950
+ {
1951
+ "epoch": 1.06,
1952
+ "learning_rate": 0.00019537214625997646,
1953
+ "loss": 0.933,
1954
+ "step": 5740
1955
+ },
1956
+ {
1957
+ "epoch": 1.06,
1958
+ "learning_rate": 0.00019500092804553609,
1959
+ "loss": 0.9475,
1960
+ "step": 5760
1961
+ },
1962
+ {
1963
+ "epoch": 1.07,
1964
+ "learning_rate": 0.00019462970983109569,
1965
+ "loss": 0.9368,
1966
+ "step": 5780
1967
+ },
1968
+ {
1969
+ "epoch": 1.07,
1970
+ "learning_rate": 0.0001942584916166553,
1971
+ "loss": 0.945,
1972
+ "step": 5800
1973
+ },
1974
+ {
1975
+ "epoch": 1.07,
1976
+ "eval_loss": 0.9158815145492554,
1977
+ "eval_runtime": 1.3387,
1978
+ "eval_samples_per_second": 0.747,
1979
+ "eval_steps_per_second": 0.747,
1980
+ "step": 5800
1981
+ },
1982
+ {
1983
+ "epoch": 1.07,
1984
+ "learning_rate": 0.00019388727340221494,
1985
+ "loss": 0.9407,
1986
+ "step": 5820
1987
+ },
1988
+ {
1989
+ "epoch": 1.08,
1990
+ "learning_rate": 0.00019351605518777454,
1991
+ "loss": 0.9345,
1992
+ "step": 5840
1993
+ },
1994
+ {
1995
+ "epoch": 1.08,
1996
+ "learning_rate": 0.00019314483697333414,
1997
+ "loss": 0.9355,
1998
+ "step": 5860
1999
+ },
2000
+ {
2001
+ "epoch": 1.08,
2002
+ "learning_rate": 0.00019277361875889374,
2003
+ "loss": 0.9368,
2004
+ "step": 5880
2005
+ },
2006
+ {
2007
+ "epoch": 1.09,
2008
+ "learning_rate": 0.00019240240054445336,
2009
+ "loss": 0.9339,
2010
+ "step": 5900
2011
+ },
2012
+ {
2013
+ "epoch": 1.09,
2014
+ "learning_rate": 0.000192031182330013,
2015
+ "loss": 0.9315,
2016
+ "step": 5920
2017
+ },
2018
+ {
2019
+ "epoch": 1.1,
2020
+ "learning_rate": 0.0001916599641155726,
2021
+ "loss": 0.9382,
2022
+ "step": 5940
2023
+ },
2024
+ {
2025
+ "epoch": 1.1,
2026
+ "learning_rate": 0.00019128874590113222,
2027
+ "loss": 0.9391,
2028
+ "step": 5960
2029
+ },
2030
+ {
2031
+ "epoch": 1.1,
2032
+ "learning_rate": 0.0001909175276866918,
2033
+ "loss": 0.9324,
2034
+ "step": 5980
2035
+ },
2036
+ {
2037
+ "epoch": 1.11,
2038
+ "learning_rate": 0.00019054630947225142,
2039
+ "loss": 0.9346,
2040
+ "step": 6000
2041
+ },
2042
+ {
2043
+ "epoch": 1.11,
2044
+ "eval_loss": 0.8685809969902039,
2045
+ "eval_runtime": 1.2797,
2046
+ "eval_samples_per_second": 0.781,
2047
+ "eval_steps_per_second": 0.781,
2048
+ "step": 6000
2049
+ },
2050
+ {
2051
+ "epoch": 1.11,
2052
+ "learning_rate": 0.00019017509125781104,
2053
+ "loss": 0.9302,
2054
+ "step": 6020
2055
+ },
2056
+ {
2057
+ "epoch": 1.11,
2058
+ "learning_rate": 0.00018980387304337064,
2059
+ "loss": 0.9375,
2060
+ "step": 6040
2061
+ },
2062
+ {
2063
+ "epoch": 1.12,
2064
+ "learning_rate": 0.00018943265482893027,
2065
+ "loss": 0.9265,
2066
+ "step": 6060
2067
+ },
2068
+ {
2069
+ "epoch": 1.12,
2070
+ "learning_rate": 0.00018906143661448987,
2071
+ "loss": 0.9421,
2072
+ "step": 6080
2073
+ },
2074
+ {
2075
+ "epoch": 1.13,
2076
+ "learning_rate": 0.00018869021840004947,
2077
+ "loss": 0.9382,
2078
+ "step": 6100
2079
+ },
2080
+ {
2081
+ "epoch": 1.13,
2082
+ "learning_rate": 0.0001883190001856091,
2083
+ "loss": 0.9376,
2084
+ "step": 6120
2085
+ },
2086
+ {
2087
+ "epoch": 1.13,
2088
+ "learning_rate": 0.0001879477819711687,
2089
+ "loss": 0.9315,
2090
+ "step": 6140
2091
+ },
2092
+ {
2093
+ "epoch": 1.14,
2094
+ "learning_rate": 0.00018757656375672832,
2095
+ "loss": 0.9326,
2096
+ "step": 6160
2097
+ },
2098
+ {
2099
+ "epoch": 1.14,
2100
+ "learning_rate": 0.00018720534554228792,
2101
+ "loss": 0.9283,
2102
+ "step": 6180
2103
+ },
2104
+ {
2105
+ "epoch": 1.14,
2106
+ "learning_rate": 0.00018683412732784755,
2107
+ "loss": 0.9252,
2108
+ "step": 6200
2109
+ },
2110
+ {
2111
+ "epoch": 1.14,
2112
+ "eval_loss": 0.8812684416770935,
2113
+ "eval_runtime": 1.2442,
2114
+ "eval_samples_per_second": 0.804,
2115
+ "eval_steps_per_second": 0.804,
2116
+ "step": 6200
2117
+ },
2118
+ {
2119
+ "epoch": 1.15,
2120
+ "learning_rate": 0.00018646290911340715,
2121
+ "loss": 0.929,
2122
+ "step": 6220
2123
+ },
2124
+ {
2125
+ "epoch": 1.15,
2126
+ "learning_rate": 0.00018609169089896674,
2127
+ "loss": 0.9303,
2128
+ "step": 6240
2129
+ },
2130
+ {
2131
+ "epoch": 1.15,
2132
+ "learning_rate": 0.00018572047268452637,
2133
+ "loss": 0.9288,
2134
+ "step": 6260
2135
+ },
2136
+ {
2137
+ "epoch": 1.16,
2138
+ "learning_rate": 0.00018534925447008597,
2139
+ "loss": 0.9308,
2140
+ "step": 6280
2141
+ },
2142
+ {
2143
+ "epoch": 1.16,
2144
+ "learning_rate": 0.0001849780362556456,
2145
+ "loss": 0.9247,
2146
+ "step": 6300
2147
+ },
2148
+ {
2149
+ "epoch": 1.17,
2150
+ "learning_rate": 0.00018460681804120522,
2151
+ "loss": 0.939,
2152
+ "step": 6320
2153
+ },
2154
+ {
2155
+ "epoch": 1.17,
2156
+ "learning_rate": 0.0001842355998267648,
2157
+ "loss": 0.927,
2158
+ "step": 6340
2159
+ },
2160
+ {
2161
+ "epoch": 1.17,
2162
+ "learning_rate": 0.00018386438161232442,
2163
+ "loss": 0.9227,
2164
+ "step": 6360
2165
+ },
2166
+ {
2167
+ "epoch": 1.18,
2168
+ "learning_rate": 0.00018349316339788402,
2169
+ "loss": 0.9275,
2170
+ "step": 6380
2171
+ },
2172
+ {
2173
+ "epoch": 1.18,
2174
+ "learning_rate": 0.00018312194518344365,
2175
+ "loss": 0.9377,
2176
+ "step": 6400
2177
+ },
2178
+ {
2179
+ "epoch": 1.18,
2180
+ "eval_loss": 0.9376851916313171,
2181
+ "eval_runtime": 1.3094,
2182
+ "eval_samples_per_second": 0.764,
2183
+ "eval_steps_per_second": 0.764,
2184
+ "step": 6400
2185
+ },
2186
+ {
2187
+ "epoch": 1.18,
2188
+ "learning_rate": 0.00018275072696900328,
2189
+ "loss": 0.9386,
2190
+ "step": 6420
2191
+ },
2192
+ {
2193
+ "epoch": 1.19,
2194
+ "learning_rate": 0.00018237950875456287,
2195
+ "loss": 0.9389,
2196
+ "step": 6440
2197
+ },
2198
+ {
2199
+ "epoch": 1.19,
2200
+ "learning_rate": 0.00018200829054012247,
2201
+ "loss": 0.9323,
2202
+ "step": 6460
2203
+ },
2204
+ {
2205
+ "epoch": 1.2,
2206
+ "learning_rate": 0.00018163707232568207,
2207
+ "loss": 0.9304,
2208
+ "step": 6480
2209
+ },
2210
+ {
2211
+ "epoch": 1.2,
2212
+ "learning_rate": 0.0001812658541112417,
2213
+ "loss": 0.9382,
2214
+ "step": 6500
2215
+ },
2216
+ {
2217
+ "epoch": 1.2,
2218
+ "learning_rate": 0.00018089463589680133,
2219
+ "loss": 0.9344,
2220
+ "step": 6520
2221
+ },
2222
+ {
2223
+ "epoch": 1.21,
2224
+ "learning_rate": 0.00018052341768236093,
2225
+ "loss": 0.9382,
2226
+ "step": 6540
2227
+ },
2228
+ {
2229
+ "epoch": 1.21,
2230
+ "learning_rate": 0.00018015219946792055,
2231
+ "loss": 0.9298,
2232
+ "step": 6560
2233
+ },
2234
+ {
2235
+ "epoch": 1.21,
2236
+ "learning_rate": 0.00017978098125348013,
2237
+ "loss": 0.9246,
2238
+ "step": 6580
2239
+ },
2240
+ {
2241
+ "epoch": 1.22,
2242
+ "learning_rate": 0.00017940976303903975,
2243
+ "loss": 0.9349,
2244
+ "step": 6600
2245
+ },
2246
+ {
2247
+ "epoch": 1.22,
2248
+ "eval_loss": 0.8311364650726318,
2249
+ "eval_runtime": 1.2953,
2250
+ "eval_samples_per_second": 0.772,
2251
+ "eval_steps_per_second": 0.772,
2252
+ "step": 6600
2253
+ },
2254
+ {
2255
+ "epoch": 1.22,
2256
+ "learning_rate": 0.00017903854482459938,
2257
+ "loss": 0.9309,
2258
+ "step": 6620
2259
+ },
2260
+ {
2261
+ "epoch": 1.22,
2262
+ "learning_rate": 0.00017866732661015898,
2263
+ "loss": 0.9311,
2264
+ "step": 6640
2265
+ },
2266
+ {
2267
+ "epoch": 1.23,
2268
+ "learning_rate": 0.0001782961083957186,
2269
+ "loss": 0.938,
2270
+ "step": 6660
2271
+ },
2272
+ {
2273
+ "epoch": 1.23,
2274
+ "learning_rate": 0.00017792489018127823,
2275
+ "loss": 0.9324,
2276
+ "step": 6680
2277
+ },
2278
+ {
2279
+ "epoch": 1.24,
2280
+ "learning_rate": 0.00017755367196683783,
2281
+ "loss": 0.9128,
2282
+ "step": 6700
2283
+ },
2284
+ {
2285
+ "epoch": 1.24,
2286
+ "learning_rate": 0.00017718245375239743,
2287
+ "loss": 0.9407,
2288
+ "step": 6720
2289
+ },
2290
+ {
2291
+ "epoch": 1.24,
2292
+ "learning_rate": 0.00017681123553795703,
2293
+ "loss": 0.9286,
2294
+ "step": 6740
2295
+ },
2296
+ {
2297
+ "epoch": 1.25,
2298
+ "learning_rate": 0.00017644001732351666,
2299
+ "loss": 0.9298,
2300
+ "step": 6760
2301
+ },
2302
+ {
2303
+ "epoch": 1.25,
2304
+ "learning_rate": 0.00017606879910907628,
2305
+ "loss": 0.9329,
2306
+ "step": 6780
2307
+ },
2308
+ {
2309
+ "epoch": 1.25,
2310
+ "learning_rate": 0.00017569758089463588,
2311
+ "loss": 0.9331,
2312
+ "step": 6800
2313
+ },
2314
+ {
2315
+ "epoch": 1.25,
2316
+ "eval_loss": 0.8751674890518188,
2317
+ "eval_runtime": 1.3065,
2318
+ "eval_samples_per_second": 0.765,
2319
+ "eval_steps_per_second": 0.765,
2320
+ "step": 6800
2321
+ },
2322
+ {
2323
+ "epoch": 1.26,
2324
+ "learning_rate": 0.0001753263626801955,
2325
+ "loss": 0.9264,
2326
+ "step": 6820
2327
+ },
2328
+ {
2329
+ "epoch": 1.26,
2330
+ "learning_rate": 0.00017495514446575508,
2331
+ "loss": 0.9345,
2332
+ "step": 6840
2333
+ },
2334
+ {
2335
+ "epoch": 1.27,
2336
+ "learning_rate": 0.0001745839262513147,
2337
+ "loss": 0.925,
2338
+ "step": 6860
2339
+ },
2340
+ {
2341
+ "epoch": 1.27,
2342
+ "learning_rate": 0.00017421270803687433,
2343
+ "loss": 0.9227,
2344
+ "step": 6880
2345
+ },
2346
+ {
2347
+ "epoch": 1.27,
2348
+ "learning_rate": 0.00017384148982243393,
2349
+ "loss": 0.9237,
2350
+ "step": 6900
2351
+ },
2352
+ {
2353
+ "epoch": 1.28,
2354
+ "learning_rate": 0.00017347027160799356,
2355
+ "loss": 0.9365,
2356
+ "step": 6920
2357
+ },
2358
+ {
2359
+ "epoch": 1.28,
2360
+ "learning_rate": 0.0001730990533935532,
2361
+ "loss": 0.9356,
2362
+ "step": 6940
2363
+ },
2364
+ {
2365
+ "epoch": 1.28,
2366
+ "learning_rate": 0.00017272783517911276,
2367
+ "loss": 0.9242,
2368
+ "step": 6960
2369
+ },
2370
+ {
2371
+ "epoch": 1.29,
2372
+ "learning_rate": 0.00017235661696467239,
2373
+ "loss": 0.9297,
2374
+ "step": 6980
2375
+ },
2376
+ {
2377
+ "epoch": 1.29,
2378
+ "learning_rate": 0.00017198539875023199,
2379
+ "loss": 0.9232,
2380
+ "step": 7000
2381
+ },
2382
+ {
2383
+ "epoch": 1.29,
2384
+ "eval_loss": 0.8359224796295166,
2385
+ "eval_runtime": 1.2767,
2386
+ "eval_samples_per_second": 0.783,
2387
+ "eval_steps_per_second": 0.783,
2388
+ "step": 7000
2389
+ },
2390
+ {
2391
+ "epoch": 1.29,
2392
+ "learning_rate": 0.0001716141805357916,
2393
+ "loss": 0.9324,
2394
+ "step": 7020
2395
+ },
2396
+ {
2397
+ "epoch": 1.3,
2398
+ "learning_rate": 0.00017124296232135124,
2399
+ "loss": 0.9222,
2400
+ "step": 7040
2401
+ },
2402
+ {
2403
+ "epoch": 1.3,
2404
+ "learning_rate": 0.00017087174410691084,
2405
+ "loss": 0.9214,
2406
+ "step": 7060
2407
+ },
2408
+ {
2409
+ "epoch": 1.31,
2410
+ "learning_rate": 0.00017050052589247044,
2411
+ "loss": 0.9214,
2412
+ "step": 7080
2413
+ },
2414
+ {
2415
+ "epoch": 1.31,
2416
+ "learning_rate": 0.00017012930767803004,
2417
+ "loss": 0.931,
2418
+ "step": 7100
2419
+ },
2420
+ {
2421
+ "epoch": 1.31,
2422
+ "learning_rate": 0.00016975808946358966,
2423
+ "loss": 0.9218,
2424
+ "step": 7120
2425
+ },
2426
+ {
2427
+ "epoch": 1.32,
2428
+ "learning_rate": 0.0001693868712491493,
2429
+ "loss": 0.9136,
2430
+ "step": 7140
2431
+ },
2432
+ {
2433
+ "epoch": 1.32,
2434
+ "learning_rate": 0.0001690156530347089,
2435
+ "loss": 0.9284,
2436
+ "step": 7160
2437
+ },
2438
+ {
2439
+ "epoch": 1.32,
2440
+ "learning_rate": 0.00016864443482026852,
2441
+ "loss": 0.921,
2442
+ "step": 7180
2443
+ },
2444
+ {
2445
+ "epoch": 1.33,
2446
+ "learning_rate": 0.0001682732166058281,
2447
+ "loss": 0.9245,
2448
+ "step": 7200
2449
+ },
2450
+ {
2451
+ "epoch": 1.33,
2452
+ "eval_loss": 0.8561541438102722,
2453
+ "eval_runtime": 1.3583,
2454
+ "eval_samples_per_second": 0.736,
2455
+ "eval_steps_per_second": 0.736,
2456
+ "step": 7200
2457
+ },
2458
+ {
2459
+ "epoch": 1.33,
2460
+ "learning_rate": 0.00016790199839138771,
2461
+ "loss": 0.9259,
2462
+ "step": 7220
2463
+ },
2464
+ {
2465
+ "epoch": 1.34,
2466
+ "learning_rate": 0.00016753078017694734,
2467
+ "loss": 0.9238,
2468
+ "step": 7240
2469
+ },
2470
+ {
2471
+ "epoch": 1.34,
2472
+ "learning_rate": 0.00016715956196250694,
2473
+ "loss": 0.9237,
2474
+ "step": 7260
2475
+ },
2476
+ {
2477
+ "epoch": 1.34,
2478
+ "learning_rate": 0.00016678834374806657,
2479
+ "loss": 0.9111,
2480
+ "step": 7280
2481
+ },
2482
+ {
2483
+ "epoch": 1.35,
2484
+ "learning_rate": 0.0001664171255336262,
2485
+ "loss": 0.9257,
2486
+ "step": 7300
2487
+ },
2488
+ {
2489
+ "epoch": 1.35,
2490
+ "learning_rate": 0.00016604590731918577,
2491
+ "loss": 0.9128,
2492
+ "step": 7320
2493
+ },
2494
+ {
2495
+ "epoch": 1.35,
2496
+ "learning_rate": 0.0001656746891047454,
2497
+ "loss": 0.9219,
2498
+ "step": 7340
2499
+ },
2500
+ {
2501
+ "epoch": 1.36,
2502
+ "learning_rate": 0.000165303470890305,
2503
+ "loss": 0.9163,
2504
+ "step": 7360
2505
+ },
2506
+ {
2507
+ "epoch": 1.36,
2508
+ "learning_rate": 0.00016493225267586462,
2509
+ "loss": 0.9195,
2510
+ "step": 7380
2511
+ },
2512
+ {
2513
+ "epoch": 1.36,
2514
+ "learning_rate": 0.00016456103446142425,
2515
+ "loss": 0.9283,
2516
+ "step": 7400
2517
+ },
2518
+ {
2519
+ "epoch": 1.36,
2520
+ "eval_loss": 0.8775261044502258,
2521
+ "eval_runtime": 1.3808,
2522
+ "eval_samples_per_second": 0.724,
2523
+ "eval_steps_per_second": 0.724,
2524
+ "step": 7400
2525
+ },
2526
+ {
2527
+ "epoch": 1.37,
2528
+ "learning_rate": 0.00016418981624698385,
2529
+ "loss": 0.9177,
2530
+ "step": 7420
2531
+ },
2532
+ {
2533
+ "epoch": 1.37,
2534
+ "learning_rate": 0.00016381859803254344,
2535
+ "loss": 0.9113,
2536
+ "step": 7440
2537
+ },
2538
+ {
2539
+ "epoch": 1.38,
2540
+ "learning_rate": 0.00016344737981810304,
2541
+ "loss": 0.9189,
2542
+ "step": 7460
2543
+ },
2544
+ {
2545
+ "epoch": 1.38,
2546
+ "learning_rate": 0.00016307616160366267,
2547
+ "loss": 0.9096,
2548
+ "step": 7480
2549
+ },
2550
+ {
2551
+ "epoch": 1.38,
2552
+ "learning_rate": 0.0001627049433892223,
2553
+ "loss": 0.9191,
2554
+ "step": 7500
2555
+ },
2556
+ {
2557
+ "epoch": 1.39,
2558
+ "learning_rate": 0.0001623337251747819,
2559
+ "loss": 0.9107,
2560
+ "step": 7520
2561
+ },
2562
+ {
2563
+ "epoch": 1.39,
2564
+ "learning_rate": 0.00016196250696034152,
2565
+ "loss": 0.9173,
2566
+ "step": 7540
2567
+ },
2568
+ {
2569
+ "epoch": 1.39,
2570
+ "learning_rate": 0.0001615912887459011,
2571
+ "loss": 0.9187,
2572
+ "step": 7560
2573
+ },
2574
+ {
2575
+ "epoch": 1.4,
2576
+ "learning_rate": 0.00016122007053146072,
2577
+ "loss": 0.9187,
2578
+ "step": 7580
2579
+ },
2580
+ {
2581
+ "epoch": 1.4,
2582
+ "learning_rate": 0.00016084885231702035,
2583
+ "loss": 0.9211,
2584
+ "step": 7600
2585
+ },
2586
+ {
2587
+ "epoch": 1.4,
2588
+ "eval_loss": 0.8760462999343872,
2589
+ "eval_runtime": 1.4851,
2590
+ "eval_samples_per_second": 0.673,
2591
+ "eval_steps_per_second": 0.673,
2592
+ "step": 7600
2593
+ },
2594
+ {
2595
+ "epoch": 1.41,
2596
+ "learning_rate": 0.00016047763410257995,
2597
+ "loss": 0.9116,
2598
+ "step": 7620
2599
+ },
2600
+ {
2601
+ "epoch": 1.41,
2602
+ "learning_rate": 0.00016010641588813957,
2603
+ "loss": 0.9117,
2604
+ "step": 7640
2605
+ },
2606
+ {
2607
+ "epoch": 1.41,
2608
+ "learning_rate": 0.0001597351976736992,
2609
+ "loss": 0.9184,
2610
+ "step": 7660
2611
+ },
2612
+ {
2613
+ "epoch": 1.42,
2614
+ "learning_rate": 0.00015936397945925877,
2615
+ "loss": 0.9122,
2616
+ "step": 7680
2617
+ },
2618
+ {
2619
+ "epoch": 1.42,
2620
+ "learning_rate": 0.0001589927612448184,
2621
+ "loss": 0.9155,
2622
+ "step": 7700
2623
+ },
2624
+ {
2625
+ "epoch": 1.42,
2626
+ "learning_rate": 0.000158621543030378,
2627
+ "loss": 0.9186,
2628
+ "step": 7720
2629
+ },
2630
+ {
2631
+ "epoch": 1.43,
2632
+ "learning_rate": 0.00015825032481593763,
2633
+ "loss": 0.9144,
2634
+ "step": 7740
2635
+ },
2636
+ {
2637
+ "epoch": 1.43,
2638
+ "learning_rate": 0.00015787910660149725,
2639
+ "loss": 0.9224,
2640
+ "step": 7760
2641
+ },
2642
+ {
2643
+ "epoch": 1.43,
2644
+ "learning_rate": 0.00015750788838705685,
2645
+ "loss": 0.9192,
2646
+ "step": 7780
2647
+ },
2648
+ {
2649
+ "epoch": 1.44,
2650
+ "learning_rate": 0.00015713667017261645,
2651
+ "loss": 0.9106,
2652
+ "step": 7800
2653
+ },
2654
+ {
2655
+ "epoch": 1.44,
2656
+ "eval_loss": 0.8746755719184875,
2657
+ "eval_runtime": 1.3658,
2658
+ "eval_samples_per_second": 0.732,
2659
+ "eval_steps_per_second": 0.732,
2660
+ "step": 7800
2661
+ },
2662
+ {
2663
+ "epoch": 1.44,
2664
+ "learning_rate": 0.00015676545195817605,
2665
+ "loss": 0.9106,
2666
+ "step": 7820
2667
+ },
2668
+ {
2669
+ "epoch": 1.45,
2670
+ "learning_rate": 0.00015639423374373568,
2671
+ "loss": 0.9024,
2672
+ "step": 7840
2673
+ },
2674
+ {
2675
+ "epoch": 1.45,
2676
+ "learning_rate": 0.0001560230155292953,
2677
+ "loss": 0.9055,
2678
+ "step": 7860
2679
+ },
2680
+ {
2681
+ "epoch": 1.45,
2682
+ "learning_rate": 0.0001556517973148549,
2683
+ "loss": 0.9106,
2684
+ "step": 7880
2685
+ },
2686
+ {
2687
+ "epoch": 1.46,
2688
+ "learning_rate": 0.00015528057910041453,
2689
+ "loss": 0.9219,
2690
+ "step": 7900
2691
+ },
2692
+ {
2693
+ "epoch": 1.46,
2694
+ "learning_rate": 0.0001549093608859741,
2695
+ "loss": 0.9015,
2696
+ "step": 7920
2697
+ },
2698
+ {
2699
+ "epoch": 1.46,
2700
+ "learning_rate": 0.00015453814267153373,
2701
+ "loss": 0.9029,
2702
+ "step": 7940
2703
+ },
2704
+ {
2705
+ "epoch": 1.47,
2706
+ "learning_rate": 0.00015416692445709336,
2707
+ "loss": 0.9095,
2708
+ "step": 7960
2709
+ },
2710
+ {
2711
+ "epoch": 1.47,
2712
+ "learning_rate": 0.00015379570624265296,
2713
+ "loss": 0.9284,
2714
+ "step": 7980
2715
+ },
2716
+ {
2717
+ "epoch": 1.48,
2718
+ "learning_rate": 0.00015342448802821258,
2719
+ "loss": 0.9047,
2720
+ "step": 8000
2721
+ },
2722
+ {
2723
+ "epoch": 1.48,
2724
+ "eval_loss": 0.8788416981697083,
2725
+ "eval_runtime": 1.3646,
2726
+ "eval_samples_per_second": 0.733,
2727
+ "eval_steps_per_second": 0.733,
2728
+ "step": 8000
2729
+ },
2730
+ {
2731
+ "epoch": 1.48,
2732
+ "learning_rate": 0.00015305326981377218,
2733
+ "loss": 0.9151,
2734
+ "step": 8020
2735
+ },
2736
+ {
2737
+ "epoch": 1.48,
2738
+ "learning_rate": 0.00015268205159933178,
2739
+ "loss": 0.9076,
2740
+ "step": 8040
2741
+ },
2742
+ {
2743
+ "epoch": 1.49,
2744
+ "learning_rate": 0.0001523108333848914,
2745
+ "loss": 0.9134,
2746
+ "step": 8060
2747
+ },
2748
+ {
2749
+ "epoch": 1.49,
2750
+ "learning_rate": 0.000151939615170451,
2751
+ "loss": 0.9136,
2752
+ "step": 8080
2753
+ },
2754
+ {
2755
+ "epoch": 1.49,
2756
+ "learning_rate": 0.00015156839695601063,
2757
+ "loss": 0.9073,
2758
+ "step": 8100
2759
+ },
2760
+ {
2761
+ "epoch": 1.5,
2762
+ "learning_rate": 0.00015119717874157023,
2763
+ "loss": 0.9054,
2764
+ "step": 8120
2765
+ },
2766
+ {
2767
+ "epoch": 1.5,
2768
+ "learning_rate": 0.00015082596052712986,
2769
+ "loss": 0.9169,
2770
+ "step": 8140
2771
+ },
2772
+ {
2773
+ "epoch": 1.5,
2774
+ "learning_rate": 0.00015045474231268946,
2775
+ "loss": 0.9249,
2776
+ "step": 8160
2777
+ },
2778
+ {
2779
+ "epoch": 1.51,
2780
+ "learning_rate": 0.00015008352409824906,
2781
+ "loss": 0.9032,
2782
+ "step": 8180
2783
+ },
2784
+ {
2785
+ "epoch": 1.51,
2786
+ "learning_rate": 0.00014971230588380869,
2787
+ "loss": 0.9089,
2788
+ "step": 8200
2789
+ },
2790
+ {
2791
+ "epoch": 1.51,
2792
+ "eval_loss": 0.8750318288803101,
2793
+ "eval_runtime": 1.361,
2794
+ "eval_samples_per_second": 0.735,
2795
+ "eval_steps_per_second": 0.735,
2796
+ "step": 8200
2797
+ },
2798
+ {
2799
+ "epoch": 1.52,
2800
+ "learning_rate": 0.00014934108766936828,
2801
+ "loss": 0.903,
2802
+ "step": 8220
2803
+ },
2804
+ {
2805
+ "epoch": 1.52,
2806
+ "learning_rate": 0.0001489698694549279,
2807
+ "loss": 0.9045,
2808
+ "step": 8240
2809
+ },
2810
+ {
2811
+ "epoch": 1.52,
2812
+ "learning_rate": 0.0001485986512404875,
2813
+ "loss": 0.9146,
2814
+ "step": 8260
2815
+ },
2816
+ {
2817
+ "epoch": 1.53,
2818
+ "learning_rate": 0.00014822743302604714,
2819
+ "loss": 0.9085,
2820
+ "step": 8280
2821
+ },
2822
+ {
2823
+ "epoch": 1.53,
2824
+ "learning_rate": 0.00014785621481160674,
2825
+ "loss": 0.9105,
2826
+ "step": 8300
2827
+ },
2828
+ {
2829
+ "epoch": 1.53,
2830
+ "learning_rate": 0.00014748499659716634,
2831
+ "loss": 0.9034,
2832
+ "step": 8320
2833
+ },
2834
+ {
2835
+ "epoch": 1.54,
2836
+ "learning_rate": 0.00014711377838272596,
2837
+ "loss": 0.906,
2838
+ "step": 8340
2839
+ },
2840
+ {
2841
+ "epoch": 1.54,
2842
+ "learning_rate": 0.0001467425601682856,
2843
+ "loss": 0.9055,
2844
+ "step": 8360
2845
+ },
2846
+ {
2847
+ "epoch": 1.55,
2848
+ "learning_rate": 0.0001463713419538452,
2849
+ "loss": 0.9013,
2850
+ "step": 8380
2851
+ },
2852
+ {
2853
+ "epoch": 1.55,
2854
+ "learning_rate": 0.0001460001237394048,
2855
+ "loss": 0.9136,
2856
+ "step": 8400
2857
+ },
2858
+ {
2859
+ "epoch": 1.55,
2860
+ "eval_loss": 0.8716868162155151,
2861
+ "eval_runtime": 1.3893,
2862
+ "eval_samples_per_second": 0.72,
2863
+ "eval_steps_per_second": 0.72,
2864
+ "step": 8400
2865
+ },
2866
+ {
2867
+ "epoch": 1.55,
2868
+ "learning_rate": 0.00014562890552496441,
2869
+ "loss": 0.8991,
2870
+ "step": 8420
2871
+ },
2872
+ {
2873
+ "epoch": 1.56,
2874
+ "learning_rate": 0.00014525768731052401,
2875
+ "loss": 0.9072,
2876
+ "step": 8440
2877
+ },
2878
+ {
2879
+ "epoch": 1.56,
2880
+ "learning_rate": 0.00014488646909608364,
2881
+ "loss": 0.9185,
2882
+ "step": 8460
2883
+ },
2884
+ {
2885
+ "epoch": 1.56,
2886
+ "learning_rate": 0.00014451525088164324,
2887
+ "loss": 0.9097,
2888
+ "step": 8480
2889
+ },
2890
+ {
2891
+ "epoch": 1.57,
2892
+ "learning_rate": 0.00014414403266720284,
2893
+ "loss": 0.9044,
2894
+ "step": 8500
2895
+ },
2896
+ {
2897
+ "epoch": 1.57,
2898
+ "learning_rate": 0.00014377281445276247,
2899
+ "loss": 0.9095,
2900
+ "step": 8520
2901
+ },
2902
+ {
2903
+ "epoch": 1.58,
2904
+ "learning_rate": 0.0001434015962383221,
2905
+ "loss": 0.9119,
2906
+ "step": 8540
2907
+ },
2908
+ {
2909
+ "epoch": 1.58,
2910
+ "learning_rate": 0.0001430303780238817,
2911
+ "loss": 0.8971,
2912
+ "step": 8560
2913
+ },
2914
+ {
2915
+ "epoch": 1.58,
2916
+ "learning_rate": 0.0001426591598094413,
2917
+ "loss": 0.9011,
2918
+ "step": 8580
2919
+ },
2920
+ {
2921
+ "epoch": 1.59,
2922
+ "learning_rate": 0.00014228794159500092,
2923
+ "loss": 0.9048,
2924
+ "step": 8600
2925
+ },
2926
+ {
2927
+ "epoch": 1.59,
2928
+ "eval_loss": 0.8536492586135864,
2929
+ "eval_runtime": 1.2569,
2930
+ "eval_samples_per_second": 0.796,
2931
+ "eval_steps_per_second": 0.796,
2932
+ "step": 8600
2933
+ },
2934
+ {
2935
+ "epoch": 1.59,
2936
+ "learning_rate": 0.00014191672338056052,
2937
+ "loss": 0.9031,
2938
+ "step": 8620
2939
+ },
2940
+ {
2941
+ "epoch": 1.59,
2942
+ "learning_rate": 0.00014154550516612014,
2943
+ "loss": 0.9078,
2944
+ "step": 8640
2945
+ },
2946
+ {
2947
+ "epoch": 1.6,
2948
+ "learning_rate": 0.00014117428695167974,
2949
+ "loss": 0.9115,
2950
+ "step": 8660
2951
+ },
2952
+ {
2953
+ "epoch": 1.6,
2954
+ "learning_rate": 0.00014080306873723934,
2955
+ "loss": 0.9018,
2956
+ "step": 8680
2957
+ },
2958
+ {
2959
+ "epoch": 1.6,
2960
+ "learning_rate": 0.00014043185052279897,
2961
+ "loss": 0.8974,
2962
+ "step": 8700
2963
+ },
2964
+ {
2965
+ "epoch": 1.61,
2966
+ "learning_rate": 0.0001400606323083586,
2967
+ "loss": 0.9084,
2968
+ "step": 8720
2969
+ },
2970
+ {
2971
+ "epoch": 1.61,
2972
+ "learning_rate": 0.0001396894140939182,
2973
+ "loss": 0.9041,
2974
+ "step": 8740
2975
+ },
2976
+ {
2977
+ "epoch": 1.62,
2978
+ "learning_rate": 0.0001393181958794778,
2979
+ "loss": 0.8967,
2980
+ "step": 8760
2981
+ },
2982
+ {
2983
+ "epoch": 1.62,
2984
+ "learning_rate": 0.00013894697766503742,
2985
+ "loss": 0.8969,
2986
+ "step": 8780
2987
+ },
2988
+ {
2989
+ "epoch": 1.62,
2990
+ "learning_rate": 0.00013857575945059702,
2991
+ "loss": 0.8993,
2992
+ "step": 8800
2993
+ },
2994
+ {
2995
+ "epoch": 1.62,
2996
+ "eval_loss": 0.8307236433029175,
2997
+ "eval_runtime": 1.2771,
2998
+ "eval_samples_per_second": 0.783,
2999
+ "eval_steps_per_second": 0.783,
3000
+ "step": 8800
3001
+ },
3002
+ {
3003
+ "epoch": 1.63,
3004
+ "learning_rate": 0.00013820454123615665,
3005
+ "loss": 0.9031,
3006
+ "step": 8820
3007
+ },
3008
+ {
3009
+ "epoch": 1.63,
3010
+ "learning_rate": 0.00013783332302171625,
3011
+ "loss": 0.9008,
3012
+ "step": 8840
3013
+ },
3014
+ {
3015
+ "epoch": 1.63,
3016
+ "learning_rate": 0.00013746210480727585,
3017
+ "loss": 0.9039,
3018
+ "step": 8860
3019
+ },
3020
+ {
3021
+ "epoch": 1.64,
3022
+ "learning_rate": 0.00013709088659283547,
3023
+ "loss": 0.9063,
3024
+ "step": 8880
3025
+ },
3026
+ {
3027
+ "epoch": 1.64,
3028
+ "learning_rate": 0.0001367196683783951,
3029
+ "loss": 0.901,
3030
+ "step": 8900
3031
+ },
3032
+ {
3033
+ "epoch": 1.65,
3034
+ "learning_rate": 0.0001363484501639547,
3035
+ "loss": 0.8973,
3036
+ "step": 8920
3037
+ },
3038
+ {
3039
+ "epoch": 1.65,
3040
+ "learning_rate": 0.0001359772319495143,
3041
+ "loss": 0.9057,
3042
+ "step": 8940
3043
+ },
3044
+ {
3045
+ "epoch": 1.65,
3046
+ "learning_rate": 0.00013560601373507393,
3047
+ "loss": 0.898,
3048
+ "step": 8960
3049
+ },
3050
+ {
3051
+ "epoch": 1.66,
3052
+ "learning_rate": 0.00013523479552063353,
3053
+ "loss": 0.9027,
3054
+ "step": 8980
3055
+ },
3056
+ {
3057
+ "epoch": 1.66,
3058
+ "learning_rate": 0.00013486357730619315,
3059
+ "loss": 0.9084,
3060
+ "step": 9000
3061
+ },
3062
+ {
3063
+ "epoch": 1.66,
3064
+ "eval_loss": 0.8812193274497986,
3065
+ "eval_runtime": 1.2199,
3066
+ "eval_samples_per_second": 0.82,
3067
+ "eval_steps_per_second": 0.82,
3068
+ "step": 9000
3069
+ },
3070
+ {
3071
+ "epoch": 1.66,
3072
+ "learning_rate": 0.00013449235909175275,
3073
+ "loss": 0.9043,
3074
+ "step": 9020
3075
+ },
3076
+ {
3077
+ "epoch": 1.67,
3078
+ "learning_rate": 0.00013412114087731235,
3079
+ "loss": 0.9021,
3080
+ "step": 9040
3081
+ },
3082
+ {
3083
+ "epoch": 1.67,
3084
+ "learning_rate": 0.00013374992266287198,
3085
+ "loss": 0.8951,
3086
+ "step": 9060
3087
+ },
3088
+ {
3089
+ "epoch": 1.67,
3090
+ "learning_rate": 0.0001333787044484316,
3091
+ "loss": 0.8865,
3092
+ "step": 9080
3093
+ },
3094
+ {
3095
+ "epoch": 1.68,
3096
+ "learning_rate": 0.0001330074862339912,
3097
+ "loss": 0.9027,
3098
+ "step": 9100
3099
+ },
3100
+ {
3101
+ "epoch": 1.68,
3102
+ "learning_rate": 0.0001326362680195508,
3103
+ "loss": 0.8976,
3104
+ "step": 9120
3105
+ },
3106
+ {
3107
+ "epoch": 1.69,
3108
+ "learning_rate": 0.00013226504980511043,
3109
+ "loss": 0.8911,
3110
+ "step": 9140
3111
+ },
3112
+ {
3113
+ "epoch": 1.69,
3114
+ "learning_rate": 0.00013189383159067003,
3115
+ "loss": 0.8973,
3116
+ "step": 9160
3117
+ },
3118
+ {
3119
+ "epoch": 1.69,
3120
+ "learning_rate": 0.00013152261337622966,
3121
+ "loss": 0.9076,
3122
+ "step": 9180
3123
+ },
3124
+ {
3125
+ "epoch": 1.7,
3126
+ "learning_rate": 0.00013115139516178925,
3127
+ "loss": 0.899,
3128
+ "step": 9200
3129
+ },
3130
+ {
3131
+ "epoch": 1.7,
3132
+ "eval_loss": 0.82787024974823,
3133
+ "eval_runtime": 1.2643,
3134
+ "eval_samples_per_second": 0.791,
3135
+ "eval_steps_per_second": 0.791,
3136
+ "step": 9200
3137
+ },
3138
+ {
3139
+ "epoch": 1.7,
3140
+ "learning_rate": 0.00013078017694734885,
3141
+ "loss": 0.9022,
3142
+ "step": 9220
3143
+ },
3144
+ {
3145
+ "epoch": 1.7,
3146
+ "learning_rate": 0.00013040895873290848,
3147
+ "loss": 0.8969,
3148
+ "step": 9240
3149
+ },
3150
+ {
3151
+ "epoch": 1.71,
3152
+ "learning_rate": 0.0001300377405184681,
3153
+ "loss": 0.895,
3154
+ "step": 9260
3155
+ },
3156
+ {
3157
+ "epoch": 1.71,
3158
+ "learning_rate": 0.0001296665223040277,
3159
+ "loss": 0.9034,
3160
+ "step": 9280
3161
+ },
3162
+ {
3163
+ "epoch": 1.72,
3164
+ "learning_rate": 0.0001292953040895873,
3165
+ "loss": 0.8892,
3166
+ "step": 9300
3167
+ },
3168
+ {
3169
+ "epoch": 1.72,
3170
+ "learning_rate": 0.00012892408587514693,
3171
+ "loss": 0.8956,
3172
+ "step": 9320
3173
+ },
3174
+ {
3175
+ "epoch": 1.72,
3176
+ "learning_rate": 0.00012855286766070656,
3177
+ "loss": 0.8952,
3178
+ "step": 9340
3179
+ },
3180
+ {
3181
+ "epoch": 1.73,
3182
+ "learning_rate": 0.00012818164944626616,
3183
+ "loss": 0.89,
3184
+ "step": 9360
3185
+ },
3186
+ {
3187
+ "epoch": 1.73,
3188
+ "learning_rate": 0.00012781043123182576,
3189
+ "loss": 0.9029,
3190
+ "step": 9380
3191
+ },
3192
+ {
3193
+ "epoch": 1.73,
3194
+ "learning_rate": 0.00012743921301738539,
3195
+ "loss": 0.8954,
3196
+ "step": 9400
3197
+ },
3198
+ {
3199
+ "epoch": 1.73,
3200
+ "eval_loss": 0.8729041218757629,
3201
+ "eval_runtime": 1.2221,
3202
+ "eval_samples_per_second": 0.818,
3203
+ "eval_steps_per_second": 0.818,
3204
+ "step": 9400
3205
+ },
3206
+ {
3207
+ "epoch": 1.74,
3208
+ "learning_rate": 0.00012706799480294498,
3209
+ "loss": 0.8805,
3210
+ "step": 9420
3211
+ },
3212
+ {
3213
+ "epoch": 1.74,
3214
+ "learning_rate": 0.0001266967765885046,
3215
+ "loss": 0.8936,
3216
+ "step": 9440
3217
+ },
3218
+ {
3219
+ "epoch": 1.74,
3220
+ "learning_rate": 0.0001263255583740642,
3221
+ "loss": 0.8893,
3222
+ "step": 9460
3223
+ },
3224
+ {
3225
+ "epoch": 1.75,
3226
+ "learning_rate": 0.0001259543401596238,
3227
+ "loss": 0.8953,
3228
+ "step": 9480
3229
+ },
3230
+ {
3231
+ "epoch": 1.75,
3232
+ "learning_rate": 0.00012558312194518344,
3233
+ "loss": 0.8943,
3234
+ "step": 9500
3235
+ },
3236
+ {
3237
+ "epoch": 1.76,
3238
+ "learning_rate": 0.00012521190373074306,
3239
+ "loss": 0.8766,
3240
+ "step": 9520
3241
+ },
3242
+ {
3243
+ "epoch": 1.76,
3244
+ "learning_rate": 0.00012484068551630266,
3245
+ "loss": 0.899,
3246
+ "step": 9540
3247
+ },
3248
+ {
3249
+ "epoch": 1.76,
3250
+ "learning_rate": 0.00012446946730186226,
3251
+ "loss": 0.8949,
3252
+ "step": 9560
3253
+ },
3254
+ {
3255
+ "epoch": 1.77,
3256
+ "learning_rate": 0.0001240982490874219,
3257
+ "loss": 0.8849,
3258
+ "step": 9580
3259
+ },
3260
+ {
3261
+ "epoch": 1.77,
3262
+ "learning_rate": 0.0001237270308729815,
3263
+ "loss": 0.894,
3264
+ "step": 9600
3265
+ },
3266
+ {
3267
+ "epoch": 1.77,
3268
+ "eval_loss": 0.8376001715660095,
3269
+ "eval_runtime": 1.3277,
3270
+ "eval_samples_per_second": 0.753,
3271
+ "eval_steps_per_second": 0.753,
3272
+ "step": 9600
3273
+ },
3274
+ {
3275
+ "epoch": 1.77,
3276
+ "learning_rate": 0.00012335581265854111,
3277
+ "loss": 0.9029,
3278
+ "step": 9620
3279
+ },
3280
+ {
3281
+ "epoch": 1.78,
3282
+ "learning_rate": 0.00012298459444410071,
3283
+ "loss": 0.898,
3284
+ "step": 9640
3285
+ },
3286
+ {
3287
+ "epoch": 1.78,
3288
+ "learning_rate": 0.00012261337622966031,
3289
+ "loss": 0.8848,
3290
+ "step": 9660
3291
+ },
3292
+ {
3293
+ "epoch": 1.79,
3294
+ "learning_rate": 0.00012224215801521994,
3295
+ "loss": 0.8932,
3296
+ "step": 9680
3297
+ },
3298
+ {
3299
+ "epoch": 1.79,
3300
+ "learning_rate": 0.00012187093980077955,
3301
+ "loss": 0.891,
3302
+ "step": 9700
3303
+ },
3304
+ {
3305
+ "epoch": 1.79,
3306
+ "learning_rate": 0.00012149972158633915,
3307
+ "loss": 0.8936,
3308
+ "step": 9720
3309
+ },
3310
+ {
3311
+ "epoch": 1.8,
3312
+ "learning_rate": 0.00012112850337189877,
3313
+ "loss": 0.8907,
3314
+ "step": 9740
3315
+ },
3316
+ {
3317
+ "epoch": 1.8,
3318
+ "learning_rate": 0.00012075728515745839,
3319
+ "loss": 0.8958,
3320
+ "step": 9760
3321
+ },
3322
+ {
3323
+ "epoch": 1.8,
3324
+ "learning_rate": 0.00012038606694301799,
3325
+ "loss": 0.8912,
3326
+ "step": 9780
3327
+ },
3328
+ {
3329
+ "epoch": 1.81,
3330
+ "learning_rate": 0.0001200148487285776,
3331
+ "loss": 0.9032,
3332
+ "step": 9800
3333
+ },
3334
+ {
3335
+ "epoch": 1.81,
3336
+ "eval_loss": 0.8445655703544617,
3337
+ "eval_runtime": 1.3427,
3338
+ "eval_samples_per_second": 0.745,
3339
+ "eval_steps_per_second": 0.745,
3340
+ "step": 9800
3341
+ },
3342
+ {
3343
+ "epoch": 1.81,
3344
+ "learning_rate": 0.00011964363051413722,
3345
+ "loss": 0.8918,
3346
+ "step": 9820
3347
+ },
3348
+ {
3349
+ "epoch": 1.81,
3350
+ "learning_rate": 0.00011927241229969682,
3351
+ "loss": 0.8865,
3352
+ "step": 9840
3353
+ },
3354
+ {
3355
+ "epoch": 1.82,
3356
+ "learning_rate": 0.00011890119408525644,
3357
+ "loss": 0.8893,
3358
+ "step": 9860
3359
+ },
3360
+ {
3361
+ "epoch": 1.82,
3362
+ "learning_rate": 0.00011852997587081606,
3363
+ "loss": 0.8863,
3364
+ "step": 9880
3365
+ },
3366
+ {
3367
+ "epoch": 1.83,
3368
+ "learning_rate": 0.00011815875765637566,
3369
+ "loss": 0.8978,
3370
+ "step": 9900
3371
+ },
3372
+ {
3373
+ "epoch": 1.83,
3374
+ "learning_rate": 0.00011778753944193527,
3375
+ "loss": 0.886,
3376
+ "step": 9920
3377
+ },
3378
+ {
3379
+ "epoch": 1.83,
3380
+ "learning_rate": 0.0001174163212274949,
3381
+ "loss": 0.8789,
3382
+ "step": 9940
3383
+ },
3384
+ {
3385
+ "epoch": 1.84,
3386
+ "learning_rate": 0.0001170451030130545,
3387
+ "loss": 0.8944,
3388
+ "step": 9960
3389
+ },
3390
+ {
3391
+ "epoch": 1.84,
3392
+ "learning_rate": 0.00011667388479861411,
3393
+ "loss": 0.8964,
3394
+ "step": 9980
3395
+ },
3396
+ {
3397
+ "epoch": 1.84,
3398
+ "learning_rate": 0.00011630266658417372,
3399
+ "loss": 0.8921,
3400
+ "step": 10000
3401
+ },
3402
+ {
3403
+ "epoch": 1.84,
3404
+ "eval_loss": 0.8549759387969971,
3405
+ "eval_runtime": 1.2407,
3406
+ "eval_samples_per_second": 0.806,
3407
+ "eval_steps_per_second": 0.806,
3408
+ "step": 10000
3409
+ },
3410
+ {
3411
+ "epoch": 1.85,
3412
+ "learning_rate": 0.00011593144836973332,
3413
+ "loss": 0.8918,
3414
+ "step": 10020
3415
+ },
3416
+ {
3417
+ "epoch": 1.85,
3418
+ "learning_rate": 0.00011556023015529295,
3419
+ "loss": 0.8966,
3420
+ "step": 10040
3421
+ },
3422
+ {
3423
+ "epoch": 1.86,
3424
+ "learning_rate": 0.00011518901194085256,
3425
+ "loss": 0.8919,
3426
+ "step": 10060
3427
+ },
3428
+ {
3429
+ "epoch": 1.86,
3430
+ "learning_rate": 0.00011481779372641216,
3431
+ "loss": 0.8963,
3432
+ "step": 10080
3433
+ },
3434
+ {
3435
+ "epoch": 1.86,
3436
+ "learning_rate": 0.00011444657551197177,
3437
+ "loss": 0.8805,
3438
+ "step": 10100
3439
+ },
3440
+ {
3441
+ "epoch": 1.87,
3442
+ "learning_rate": 0.0001140753572975314,
3443
+ "loss": 0.8837,
3444
+ "step": 10120
3445
+ },
3446
+ {
3447
+ "epoch": 1.87,
3448
+ "learning_rate": 0.000113704139083091,
3449
+ "loss": 0.8919,
3450
+ "step": 10140
3451
+ },
3452
+ {
3453
+ "epoch": 1.87,
3454
+ "learning_rate": 0.00011333292086865061,
3455
+ "loss": 0.8904,
3456
+ "step": 10160
3457
+ },
3458
+ {
3459
+ "epoch": 1.88,
3460
+ "learning_rate": 0.00011296170265421023,
3461
+ "loss": 0.8765,
3462
+ "step": 10180
3463
+ },
3464
+ {
3465
+ "epoch": 1.88,
3466
+ "learning_rate": 0.00011259048443976982,
3467
+ "loss": 0.8846,
3468
+ "step": 10200
3469
+ },
3470
+ {
3471
+ "epoch": 1.88,
3472
+ "eval_loss": 0.8317739963531494,
3473
+ "eval_runtime": 1.3125,
3474
+ "eval_samples_per_second": 0.762,
3475
+ "eval_steps_per_second": 0.762,
3476
+ "step": 10200
3477
+ },
3478
+ {
3479
+ "epoch": 1.88,
3480
+ "learning_rate": 0.00011221926622532945,
3481
+ "loss": 0.8754,
3482
+ "step": 10220
3483
+ },
3484
+ {
3485
+ "epoch": 1.89,
3486
+ "learning_rate": 0.00011184804801088906,
3487
+ "loss": 0.8824,
3488
+ "step": 10240
3489
+ },
3490
+ {
3491
+ "epoch": 1.89,
3492
+ "learning_rate": 0.00011147682979644866,
3493
+ "loss": 0.888,
3494
+ "step": 10260
3495
+ },
3496
+ {
3497
+ "epoch": 1.9,
3498
+ "learning_rate": 0.00011110561158200828,
3499
+ "loss": 0.8986,
3500
+ "step": 10280
3501
+ },
3502
+ {
3503
+ "epoch": 1.9,
3504
+ "learning_rate": 0.0001107343933675679,
3505
+ "loss": 0.8764,
3506
+ "step": 10300
3507
+ },
3508
+ {
3509
+ "epoch": 1.9,
3510
+ "learning_rate": 0.0001103631751531275,
3511
+ "loss": 0.8901,
3512
+ "step": 10320
3513
+ },
3514
+ {
3515
+ "epoch": 1.91,
3516
+ "learning_rate": 0.00010999195693868712,
3517
+ "loss": 0.8859,
3518
+ "step": 10340
3519
+ },
3520
+ {
3521
+ "epoch": 1.91,
3522
+ "learning_rate": 0.00010962073872424673,
3523
+ "loss": 0.8868,
3524
+ "step": 10360
3525
+ },
3526
+ {
3527
+ "epoch": 1.91,
3528
+ "learning_rate": 0.00010924952050980633,
3529
+ "loss": 0.8827,
3530
+ "step": 10380
3531
+ },
3532
+ {
3533
+ "epoch": 1.92,
3534
+ "learning_rate": 0.00010887830229536595,
3535
+ "loss": 0.8911,
3536
+ "step": 10400
3537
+ },
3538
+ {
3539
+ "epoch": 1.92,
3540
+ "eval_loss": 0.8604958057403564,
3541
+ "eval_runtime": 1.2674,
3542
+ "eval_samples_per_second": 0.789,
3543
+ "eval_steps_per_second": 0.789,
3544
+ "step": 10400
3545
+ },
3546
+ {
3547
+ "epoch": 1.92,
3548
+ "learning_rate": 0.00010850708408092557,
3549
+ "loss": 0.8888,
3550
+ "step": 10420
3551
+ },
3552
+ {
3553
+ "epoch": 1.93,
3554
+ "learning_rate": 0.00010813586586648517,
3555
+ "loss": 0.8878,
3556
+ "step": 10440
3557
+ },
3558
+ {
3559
+ "epoch": 1.93,
3560
+ "learning_rate": 0.00010776464765204478,
3561
+ "loss": 0.8853,
3562
+ "step": 10460
3563
+ },
3564
+ {
3565
+ "epoch": 1.93,
3566
+ "learning_rate": 0.0001073934294376044,
3567
+ "loss": 0.8831,
3568
+ "step": 10480
3569
+ },
3570
+ {
3571
+ "epoch": 1.94,
3572
+ "learning_rate": 0.000107022211223164,
3573
+ "loss": 0.8906,
3574
+ "step": 10500
3575
+ },
3576
+ {
3577
+ "epoch": 1.94,
3578
+ "learning_rate": 0.00010665099300872362,
3579
+ "loss": 0.8869,
3580
+ "step": 10520
3581
+ },
3582
+ {
3583
+ "epoch": 1.94,
3584
+ "learning_rate": 0.00010627977479428323,
3585
+ "loss": 0.882,
3586
+ "step": 10540
3587
+ },
3588
+ {
3589
+ "epoch": 1.95,
3590
+ "learning_rate": 0.00010590855657984283,
3591
+ "loss": 0.8789,
3592
+ "step": 10560
3593
+ },
3594
+ {
3595
+ "epoch": 1.95,
3596
+ "learning_rate": 0.00010553733836540246,
3597
+ "loss": 0.8819,
3598
+ "step": 10580
3599
+ },
3600
+ {
3601
+ "epoch": 1.95,
3602
+ "learning_rate": 0.00010516612015096207,
3603
+ "loss": 0.8765,
3604
+ "step": 10600
3605
+ },
3606
+ {
3607
+ "epoch": 1.95,
3608
+ "eval_loss": 0.8246310949325562,
3609
+ "eval_runtime": 1.3918,
3610
+ "eval_samples_per_second": 0.719,
3611
+ "eval_steps_per_second": 0.719,
3612
+ "step": 10600
3613
+ },
3614
+ {
3615
+ "epoch": 1.96,
3616
+ "learning_rate": 0.00010479490193652167,
3617
+ "loss": 0.8754,
3618
+ "step": 10620
3619
+ },
3620
+ {
3621
+ "epoch": 1.96,
3622
+ "learning_rate": 0.00010442368372208128,
3623
+ "loss": 0.8876,
3624
+ "step": 10640
3625
+ },
3626
+ {
3627
+ "epoch": 1.97,
3628
+ "learning_rate": 0.0001040524655076409,
3629
+ "loss": 0.8871,
3630
+ "step": 10660
3631
+ },
3632
+ {
3633
+ "epoch": 1.97,
3634
+ "learning_rate": 0.00010368124729320051,
3635
+ "loss": 0.879,
3636
+ "step": 10680
3637
+ },
3638
+ {
3639
+ "epoch": 1.97,
3640
+ "learning_rate": 0.00010331002907876012,
3641
+ "loss": 0.8809,
3642
+ "step": 10700
3643
+ },
3644
+ {
3645
+ "epoch": 1.98,
3646
+ "learning_rate": 0.00010293881086431974,
3647
+ "loss": 0.8847,
3648
+ "step": 10720
3649
+ },
3650
+ {
3651
+ "epoch": 1.98,
3652
+ "learning_rate": 0.00010256759264987934,
3653
+ "loss": 0.8861,
3654
+ "step": 10740
3655
+ },
3656
+ {
3657
+ "epoch": 1.98,
3658
+ "learning_rate": 0.00010219637443543895,
3659
+ "loss": 0.8758,
3660
+ "step": 10760
3661
+ },
3662
+ {
3663
+ "epoch": 1.99,
3664
+ "learning_rate": 0.00010182515622099858,
3665
+ "loss": 0.8781,
3666
+ "step": 10780
3667
+ },
3668
+ {
3669
+ "epoch": 1.99,
3670
+ "learning_rate": 0.00010145393800655817,
3671
+ "loss": 0.8726,
3672
+ "step": 10800
3673
+ },
3674
+ {
3675
+ "epoch": 1.99,
3676
+ "eval_loss": 0.846712052822113,
3677
+ "eval_runtime": 1.3944,
3678
+ "eval_samples_per_second": 0.717,
3679
+ "eval_steps_per_second": 0.717,
3680
+ "step": 10800
3681
+ },
3682
+ {
3683
+ "epoch": 2.0,
3684
+ "learning_rate": 0.00010108271979211779,
3685
+ "loss": 0.8774,
3686
+ "step": 10820
3687
+ },
3688
+ {
3689
+ "epoch": 2.0,
3690
+ "learning_rate": 0.0001007115015776774,
3691
+ "loss": 0.8755,
3692
+ "step": 10840
3693
+ },
3694
+ {
3695
+ "epoch": 2.0,
3696
+ "learning_rate": 0.000100340283363237,
3697
+ "loss": 0.8814,
3698
+ "step": 10860
3699
+ },
3700
+ {
3701
+ "epoch": 2.01,
3702
+ "learning_rate": 9.996906514879663e-05,
3703
+ "loss": 0.8694,
3704
+ "step": 10880
3705
+ },
3706
+ {
3707
+ "epoch": 2.01,
3708
+ "learning_rate": 9.959784693435624e-05,
3709
+ "loss": 0.8808,
3710
+ "step": 10900
3711
+ },
3712
+ {
3713
+ "epoch": 2.01,
3714
+ "learning_rate": 9.922662871991584e-05,
3715
+ "loss": 0.8824,
3716
+ "step": 10920
3717
+ },
3718
+ {
3719
+ "epoch": 2.02,
3720
+ "learning_rate": 9.885541050547545e-05,
3721
+ "loss": 0.8669,
3722
+ "step": 10940
3723
+ },
3724
+ {
3725
+ "epoch": 2.02,
3726
+ "learning_rate": 9.848419229103508e-05,
3727
+ "loss": 0.8702,
3728
+ "step": 10960
3729
+ },
3730
+ {
3731
+ "epoch": 2.03,
3732
+ "learning_rate": 9.811297407659468e-05,
3733
+ "loss": 0.8722,
3734
+ "step": 10980
3735
+ },
3736
+ {
3737
+ "epoch": 2.03,
3738
+ "learning_rate": 9.774175586215429e-05,
3739
+ "loss": 0.8844,
3740
+ "step": 11000
3741
+ },
3742
+ {
3743
+ "epoch": 2.03,
3744
+ "eval_loss": 0.8575500845909119,
3745
+ "eval_runtime": 1.622,
3746
+ "eval_samples_per_second": 0.617,
3747
+ "eval_steps_per_second": 0.617,
3748
+ "step": 11000
3749
+ },
3750
+ {
3751
+ "epoch": 2.03,
3752
+ "learning_rate": 9.73705376477139e-05,
3753
+ "loss": 0.877,
3754
+ "step": 11020
3755
+ },
3756
+ {
3757
+ "epoch": 2.04,
3758
+ "learning_rate": 9.699931943327353e-05,
3759
+ "loss": 0.8635,
3760
+ "step": 11040
3761
+ },
3762
+ {
3763
+ "epoch": 2.04,
3764
+ "learning_rate": 9.662810121883313e-05,
3765
+ "loss": 0.8729,
3766
+ "step": 11060
3767
+ },
3768
+ {
3769
+ "epoch": 2.04,
3770
+ "learning_rate": 9.625688300439274e-05,
3771
+ "loss": 0.865,
3772
+ "step": 11080
3773
+ },
3774
+ {
3775
+ "epoch": 2.05,
3776
+ "learning_rate": 9.588566478995236e-05,
3777
+ "loss": 0.8811,
3778
+ "step": 11100
3779
+ },
3780
+ {
3781
+ "epoch": 2.05,
3782
+ "learning_rate": 9.551444657551196e-05,
3783
+ "loss": 0.8729,
3784
+ "step": 11120
3785
+ },
3786
+ {
3787
+ "epoch": 2.05,
3788
+ "learning_rate": 9.514322836107158e-05,
3789
+ "loss": 0.8865,
3790
+ "step": 11140
3791
+ },
3792
+ {
3793
+ "epoch": 2.06,
3794
+ "learning_rate": 9.47720101466312e-05,
3795
+ "loss": 0.8747,
3796
+ "step": 11160
3797
+ },
3798
+ {
3799
+ "epoch": 2.06,
3800
+ "learning_rate": 9.44007919321908e-05,
3801
+ "loss": 0.8779,
3802
+ "step": 11180
3803
+ },
3804
+ {
3805
+ "epoch": 2.07,
3806
+ "learning_rate": 9.402957371775041e-05,
3807
+ "loss": 0.8764,
3808
+ "step": 11200
3809
+ },
3810
+ {
3811
+ "epoch": 2.07,
3812
+ "eval_loss": 0.8142430186271667,
3813
+ "eval_runtime": 1.4953,
3814
+ "eval_samples_per_second": 0.669,
3815
+ "eval_steps_per_second": 0.669,
3816
+ "step": 11200
3817
+ },
3818
+ {
3819
+ "epoch": 2.07,
3820
+ "learning_rate": 9.365835550331003e-05,
3821
+ "loss": 0.8779,
3822
+ "step": 11220
3823
+ },
3824
+ {
3825
+ "epoch": 2.07,
3826
+ "learning_rate": 9.328713728886963e-05,
3827
+ "loss": 0.8702,
3828
+ "step": 11240
3829
+ },
3830
+ {
3831
+ "epoch": 2.08,
3832
+ "learning_rate": 9.291591907442925e-05,
3833
+ "loss": 0.8832,
3834
+ "step": 11260
3835
+ },
3836
+ {
3837
+ "epoch": 2.08,
3838
+ "learning_rate": 9.254470085998886e-05,
3839
+ "loss": 0.8718,
3840
+ "step": 11280
3841
+ },
3842
+ {
3843
+ "epoch": 2.08,
3844
+ "learning_rate": 9.217348264554846e-05,
3845
+ "loss": 0.8726,
3846
+ "step": 11300
3847
+ },
3848
+ {
3849
+ "epoch": 2.09,
3850
+ "learning_rate": 9.180226443110809e-05,
3851
+ "loss": 0.8677,
3852
+ "step": 11320
3853
+ },
3854
+ {
3855
+ "epoch": 2.09,
3856
+ "learning_rate": 9.14310462166677e-05,
3857
+ "loss": 0.8748,
3858
+ "step": 11340
3859
+ },
3860
+ {
3861
+ "epoch": 2.1,
3862
+ "learning_rate": 9.10598280022273e-05,
3863
+ "loss": 0.8694,
3864
+ "step": 11360
3865
+ },
3866
+ {
3867
+ "epoch": 2.1,
3868
+ "learning_rate": 9.068860978778691e-05,
3869
+ "loss": 0.8613,
3870
+ "step": 11380
3871
+ },
3872
+ {
3873
+ "epoch": 2.1,
3874
+ "learning_rate": 9.031739157334652e-05,
3875
+ "loss": 0.8687,
3876
+ "step": 11400
3877
+ },
3878
+ {
3879
+ "epoch": 2.1,
3880
+ "eval_loss": 0.827206552028656,
3881
+ "eval_runtime": 1.6259,
3882
+ "eval_samples_per_second": 0.615,
3883
+ "eval_steps_per_second": 0.615,
3884
+ "step": 11400
3885
+ },
3886
+ {
3887
+ "epoch": 2.11,
3888
+ "learning_rate": 8.994617335890614e-05,
3889
+ "loss": 0.8735,
3890
+ "step": 11420
3891
+ },
3892
+ {
3893
+ "epoch": 2.11,
3894
+ "learning_rate": 8.957495514446575e-05,
3895
+ "loss": 0.8691,
3896
+ "step": 11440
3897
+ },
3898
+ {
3899
+ "epoch": 2.11,
3900
+ "learning_rate": 8.920373693002536e-05,
3901
+ "loss": 0.8605,
3902
+ "step": 11460
3903
+ },
3904
+ {
3905
+ "epoch": 2.12,
3906
+ "learning_rate": 8.883251871558496e-05,
3907
+ "loss": 0.8715,
3908
+ "step": 11480
3909
+ },
3910
+ {
3911
+ "epoch": 2.12,
3912
+ "learning_rate": 8.846130050114459e-05,
3913
+ "loss": 0.8541,
3914
+ "step": 11500
3915
+ },
3916
+ {
3917
+ "epoch": 2.12,
3918
+ "learning_rate": 8.80900822867042e-05,
3919
+ "loss": 0.8668,
3920
+ "step": 11520
3921
+ },
3922
+ {
3923
+ "epoch": 2.13,
3924
+ "learning_rate": 8.77188640722638e-05,
3925
+ "loss": 0.8745,
3926
+ "step": 11540
3927
+ },
3928
+ {
3929
+ "epoch": 2.13,
3930
+ "learning_rate": 8.734764585782342e-05,
3931
+ "loss": 0.8781,
3932
+ "step": 11560
3933
+ },
3934
+ {
3935
+ "epoch": 2.14,
3936
+ "learning_rate": 8.697642764338303e-05,
3937
+ "loss": 0.8782,
3938
+ "step": 11580
3939
+ },
3940
+ {
3941
+ "epoch": 2.14,
3942
+ "learning_rate": 8.660520942894264e-05,
3943
+ "loss": 0.8733,
3944
+ "step": 11600
3945
+ },
3946
+ {
3947
+ "epoch": 2.14,
3948
+ "eval_loss": 0.8250707983970642,
3949
+ "eval_runtime": 1.5297,
3950
+ "eval_samples_per_second": 0.654,
3951
+ "eval_steps_per_second": 0.654,
3952
+ "step": 11600
3953
+ }
3954
+ ],
3955
+ "max_steps": 16263,
3956
+ "num_train_epochs": 3,
3957
+ "total_flos": 7.693572394031514e+18,
3958
+ "trial_name": null,
3959
+ "trial_params": null
3960
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b30e4cfba60dd090f9ba150f3ef0d8c6756a0922cf381751c8279ce3497c1e3
3
+ size 3899