Upload BLIP3o_Sana_DiT_freezed_1024_diffusion_distillation_1024_freezed_only_60k_v2_16_queries_1600M/trainer_state.json with huggingface_hub
Browse files
BLIP3o_Sana_DiT_freezed_1024_diffusion_distillation_1024_freezed_only_60k_v2_16_queries_1600M/trainer_state.json
ADDED
|
@@ -0,0 +1,2185 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 19.872964169381106,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3060,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06514657980456026,
|
| 14 |
+
"grad_norm": 360608.86507283914,
|
| 15 |
+
"learning_rate": 1.1764705882352942e-05,
|
| 16 |
+
"loss": 4.4398,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13029315960912052,
|
| 21 |
+
"grad_norm": 186341283.66978198,
|
| 22 |
+
"learning_rate": 2.4836601307189544e-05,
|
| 23 |
+
"loss": 3.4937,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.19543973941368079,
|
| 28 |
+
"grad_norm": 8265962.46325053,
|
| 29 |
+
"learning_rate": 3.790849673202614e-05,
|
| 30 |
+
"loss": 2.869,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.26058631921824105,
|
| 35 |
+
"grad_norm": 729.7166529653447,
|
| 36 |
+
"learning_rate": 5.0980392156862745e-05,
|
| 37 |
+
"loss": 2.2669,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.3257328990228013,
|
| 42 |
+
"grad_norm": 2581.7595169722435,
|
| 43 |
+
"learning_rate": 6.405228758169934e-05,
|
| 44 |
+
"loss": 2.068,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.39087947882736157,
|
| 49 |
+
"grad_norm": 1579.3443179118115,
|
| 50 |
+
"learning_rate": 7.712418300653595e-05,
|
| 51 |
+
"loss": 1.8933,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.4560260586319218,
|
| 56 |
+
"grad_norm": 2680.4556360526485,
|
| 57 |
+
"learning_rate": 9.019607843137255e-05,
|
| 58 |
+
"loss": 1.8315,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5211726384364821,
|
| 63 |
+
"grad_norm": 699.0730355827216,
|
| 64 |
+
"learning_rate": 0.00010326797385620916,
|
| 65 |
+
"loss": 1.7269,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5863192182410424,
|
| 70 |
+
"grad_norm": 891.0609285345522,
|
| 71 |
+
"learning_rate": 0.00011633986928104577,
|
| 72 |
+
"loss": 1.7051,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6514657980456026,
|
| 77 |
+
"grad_norm": 1471.2893578205953,
|
| 78 |
+
"learning_rate": 0.00012941176470588237,
|
| 79 |
+
"loss": 1.6762,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7166123778501629,
|
| 84 |
+
"grad_norm": 1097.7808433663154,
|
| 85 |
+
"learning_rate": 0.00014248366013071897,
|
| 86 |
+
"loss": 1.5908,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.7817589576547231,
|
| 91 |
+
"grad_norm": 294.5842824025576,
|
| 92 |
+
"learning_rate": 0.00015555555555555556,
|
| 93 |
+
"loss": 1.5773,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.8469055374592834,
|
| 98 |
+
"grad_norm": 735.6925187738833,
|
| 99 |
+
"learning_rate": 0.00016862745098039216,
|
| 100 |
+
"loss": 1.5294,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.9120521172638436,
|
| 105 |
+
"grad_norm": 1272.9136753624246,
|
| 106 |
+
"learning_rate": 0.00018169934640522876,
|
| 107 |
+
"loss": 1.5144,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.9771986970684039,
|
| 112 |
+
"grad_norm": 1559.4054545257081,
|
| 113 |
+
"learning_rate": 0.00019477124183006535,
|
| 114 |
+
"loss": 1.4846,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.0390879478827362,
|
| 119 |
+
"grad_norm": 2783.200821017181,
|
| 120 |
+
"learning_rate": 0.00019999790827905065,
|
| 121 |
+
"loss": 1.3942,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.1042345276872965,
|
| 126 |
+
"grad_norm": 4826.21640367058,
|
| 127 |
+
"learning_rate": 0.00019998512585840076,
|
| 128 |
+
"loss": 1.4203,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.1693811074918568,
|
| 133 |
+
"grad_norm": 2737.7979428281187,
|
| 134 |
+
"learning_rate": 0.00019996072457534358,
|
| 135 |
+
"loss": 1.267,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.234527687296417,
|
| 140 |
+
"grad_norm": 2992.8365594437128,
|
| 141 |
+
"learning_rate": 0.00019992470727970372,
|
| 142 |
+
"loss": 1.181,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.2996742671009773,
|
| 147 |
+
"grad_norm": 1061.7776423133055,
|
| 148 |
+
"learning_rate": 0.0001998770781779391,
|
| 149 |
+
"loss": 1.1085,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.3648208469055374,
|
| 154 |
+
"grad_norm": 1344.0957620642432,
|
| 155 |
+
"learning_rate": 0.00019981784283264997,
|
| 156 |
+
"loss": 1.1415,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.4299674267100977,
|
| 161 |
+
"grad_norm": 4903.235302726769,
|
| 162 |
+
"learning_rate": 0.00019974700816192902,
|
| 163 |
+
"loss": 1.1828,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.495114006514658,
|
| 168 |
+
"grad_norm": 1475.3930669547442,
|
| 169 |
+
"learning_rate": 0.00019966458243855358,
|
| 170 |
+
"loss": 1.322,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.5602605863192183,
|
| 175 |
+
"grad_norm": 160.51717374739323,
|
| 176 |
+
"learning_rate": 0.00019957057528901943,
|
| 177 |
+
"loss": 1.4437,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.6254071661237783,
|
| 182 |
+
"grad_norm": 90.6383884059925,
|
| 183 |
+
"learning_rate": 0.00019946499769241634,
|
| 184 |
+
"loss": 1.5052,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.6905537459283386,
|
| 189 |
+
"grad_norm": 77.20409880308091,
|
| 190 |
+
"learning_rate": 0.00019934786197914607,
|
| 191 |
+
"loss": 1.527,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.755700325732899,
|
| 196 |
+
"grad_norm": 584.6087543703921,
|
| 197 |
+
"learning_rate": 0.00019921918182948218,
|
| 198 |
+
"loss": 1.5687,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.8208469055374592,
|
| 203 |
+
"grad_norm": 223.9836170899833,
|
| 204 |
+
"learning_rate": 0.00019907897227197226,
|
| 205 |
+
"loss": 1.6124,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.8859934853420195,
|
| 210 |
+
"grad_norm": 411.66038448802556,
|
| 211 |
+
"learning_rate": 0.00019892724968168296,
|
| 212 |
+
"loss": 1.6502,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.9511400651465798,
|
| 217 |
+
"grad_norm": 268.46609498828803,
|
| 218 |
+
"learning_rate": 0.00019876403177828723,
|
| 219 |
+
"loss": 1.5626,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 2.013029315960912,
|
| 224 |
+
"grad_norm": 211.91436035082953,
|
| 225 |
+
"learning_rate": 0.0001985893376239951,
|
| 226 |
+
"loss": 1.4781,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 2.0781758957654723,
|
| 231 |
+
"grad_norm": 459.3562194957706,
|
| 232 |
+
"learning_rate": 0.0001984031876213273,
|
| 233 |
+
"loss": 1.4971,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 2.1433224755700326,
|
| 238 |
+
"grad_norm": 71.60839870383397,
|
| 239 |
+
"learning_rate": 0.0001982056035107324,
|
| 240 |
+
"loss": 1.4471,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 2.208469055374593,
|
| 245 |
+
"grad_norm": 113.83511308244341,
|
| 246 |
+
"learning_rate": 0.00019799660836804795,
|
| 247 |
+
"loss": 1.4092,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.273615635179153,
|
| 252 |
+
"grad_norm": 119.07762246968588,
|
| 253 |
+
"learning_rate": 0.0001977762266018052,
|
| 254 |
+
"loss": 1.4097,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 2.3387622149837135,
|
| 259 |
+
"grad_norm": 366.9860474050572,
|
| 260 |
+
"learning_rate": 0.00019754448395037857,
|
| 261 |
+
"loss": 1.3463,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 2.403908794788274,
|
| 266 |
+
"grad_norm": 283.1974690025577,
|
| 267 |
+
"learning_rate": 0.00019730140747897964,
|
| 268 |
+
"loss": 1.511,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 2.469055374592834,
|
| 273 |
+
"grad_norm": 769.662981937382,
|
| 274 |
+
"learning_rate": 0.0001970470255764962,
|
| 275 |
+
"loss": 2.4266,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 2.5342019543973944,
|
| 280 |
+
"grad_norm": 683.2593946232895,
|
| 281 |
+
"learning_rate": 0.00019678136795217677,
|
| 282 |
+
"loss": 1.6408,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 2.5993485342019547,
|
| 287 |
+
"grad_norm": 1488.9204150214603,
|
| 288 |
+
"learning_rate": 0.00019650446563216072,
|
| 289 |
+
"loss": 1.6902,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.6644951140065145,
|
| 294 |
+
"grad_norm": 17553.046080527718,
|
| 295 |
+
"learning_rate": 0.00019621635095585488,
|
| 296 |
+
"loss": 1.8008,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.729641693811075,
|
| 301 |
+
"grad_norm": 680178.7395430527,
|
| 302 |
+
"learning_rate": 0.0001959170575721565,
|
| 303 |
+
"loss": 2.3105,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.794788273615635,
|
| 308 |
+
"grad_norm": 10707.052697183222,
|
| 309 |
+
"learning_rate": 0.00019560662043552354,
|
| 310 |
+
"loss": 2.0898,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 2.8599348534201954,
|
| 315 |
+
"grad_norm": 3141.0285276514596,
|
| 316 |
+
"learning_rate": 0.0001952850758018922,
|
| 317 |
+
"loss": 2.9009,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.9250814332247557,
|
| 322 |
+
"grad_norm": 64923.53250048505,
|
| 323 |
+
"learning_rate": 0.00019495246122444264,
|
| 324 |
+
"loss": 5.0433,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.990228013029316,
|
| 329 |
+
"grad_norm": 49498.925873212786,
|
| 330 |
+
"learning_rate": 0.00019460881554921315,
|
| 331 |
+
"loss": 8.007,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.0521172638436482,
|
| 336 |
+
"grad_norm": 89264.87657655087,
|
| 337 |
+
"learning_rate": 0.00019425417891056338,
|
| 338 |
+
"loss": 4.0258,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 3.1172638436482085,
|
| 343 |
+
"grad_norm": 51715.483595451275,
|
| 344 |
+
"learning_rate": 0.00019388859272648695,
|
| 345 |
+
"loss": 4.2459,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 3.182410423452769,
|
| 350 |
+
"grad_norm": 47436.77709895252,
|
| 351 |
+
"learning_rate": 0.00019351209969377427,
|
| 352 |
+
"loss": 3.6299,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 3.247557003257329,
|
| 357 |
+
"grad_norm": 13459.17073854497,
|
| 358 |
+
"learning_rate": 0.00019312474378302606,
|
| 359 |
+
"loss": 3.8708,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 3.3127035830618894,
|
| 364 |
+
"grad_norm": 13115.661342301686,
|
| 365 |
+
"learning_rate": 0.0001927265702335179,
|
| 366 |
+
"loss": 4.8392,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 3.3778501628664497,
|
| 371 |
+
"grad_norm": 26676.116432336054,
|
| 372 |
+
"learning_rate": 0.00019231762554791697,
|
| 373 |
+
"loss": 6.3413,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.44299674267101,
|
| 378 |
+
"grad_norm": 65251.90743199031,
|
| 379 |
+
"learning_rate": 0.00019189795748685064,
|
| 380 |
+
"loss": 6.6099,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 3.5081433224755703,
|
| 385 |
+
"grad_norm": 26107.692790828612,
|
| 386 |
+
"learning_rate": 0.00019146761506332882,
|
| 387 |
+
"loss": 6.2382,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 3.5732899022801305,
|
| 392 |
+
"grad_norm": 5260.782049808003,
|
| 393 |
+
"learning_rate": 0.00019102664853701967,
|
| 394 |
+
"loss": 6.4178,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.6384364820846904,
|
| 399 |
+
"grad_norm": 45516.44424060653,
|
| 400 |
+
"learning_rate": 0.00019057510940837958,
|
| 401 |
+
"loss": 9.9453,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.7035830618892507,
|
| 406 |
+
"grad_norm": 30212.366369805975,
|
| 407 |
+
"learning_rate": 0.00019011305041263872,
|
| 408 |
+
"loss": 10.4226,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 3.768729641693811,
|
| 413 |
+
"grad_norm": 19991.812193183567,
|
| 414 |
+
"learning_rate": 0.00018964052551364197,
|
| 415 |
+
"loss": 7.7251,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.8338762214983713,
|
| 420 |
+
"grad_norm": 42609.684620191125,
|
| 421 |
+
"learning_rate": 0.00018915758989754646,
|
| 422 |
+
"loss": 7.0589,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.8990228013029316,
|
| 427 |
+
"grad_norm": 41904.90845956626,
|
| 428 |
+
"learning_rate": 0.00018866429996637626,
|
| 429 |
+
"loss": 6.7334,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 3.964169381107492,
|
| 434 |
+
"grad_norm": 65566.33279016556,
|
| 435 |
+
"learning_rate": 0.00018816071333143557,
|
| 436 |
+
"loss": 6.7104,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 4.026058631921824,
|
| 441 |
+
"grad_norm": 255691.8689503633,
|
| 442 |
+
"learning_rate": 0.00018764688880657987,
|
| 443 |
+
"loss": 6.0155,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 4.091205211726384,
|
| 448 |
+
"grad_norm": 54132.19692028039,
|
| 449 |
+
"learning_rate": 0.00018712288640134733,
|
| 450 |
+
"loss": 6.0403,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 4.156351791530945,
|
| 455 |
+
"grad_norm": 48495.14687968065,
|
| 456 |
+
"learning_rate": 0.00018658876731395015,
|
| 457 |
+
"loss": 5.9146,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 4.221498371335505,
|
| 462 |
+
"grad_norm": 99341.39952270637,
|
| 463 |
+
"learning_rate": 0.00018604459392412744,
|
| 464 |
+
"loss": 5.8905,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 4.286644951140065,
|
| 469 |
+
"grad_norm": 365405.5140102734,
|
| 470 |
+
"learning_rate": 0.00018549042978585953,
|
| 471 |
+
"loss": 5.6987,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 4.351791530944626,
|
| 476 |
+
"grad_norm": 1133084.5174009823,
|
| 477 |
+
"learning_rate": 0.00018492633961994588,
|
| 478 |
+
"loss": 5.567,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 4.416938110749186,
|
| 483 |
+
"grad_norm": 1026606.0570907952,
|
| 484 |
+
"learning_rate": 0.00018435238930644606,
|
| 485 |
+
"loss": 5.5081,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 4.482084690553746,
|
| 490 |
+
"grad_norm": 1092860.8188589269,
|
| 491 |
+
"learning_rate": 0.00018376864587698583,
|
| 492 |
+
"loss": 5.3961,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 4.547231270358306,
|
| 497 |
+
"grad_norm": 820609.0281828152,
|
| 498 |
+
"learning_rate": 0.00018317517750692834,
|
| 499 |
+
"loss": 5.6312,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 4.612377850162867,
|
| 504 |
+
"grad_norm": 2221021.493482524,
|
| 505 |
+
"learning_rate": 0.0001825720535074121,
|
| 506 |
+
"loss": 5.1049,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 4.677524429967427,
|
| 511 |
+
"grad_norm": 2684760.5950317224,
|
| 512 |
+
"learning_rate": 0.00018195934431725606,
|
| 513 |
+
"loss": 4.91,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 4.742671009771987,
|
| 518 |
+
"grad_norm": 7491.765043720262,
|
| 519 |
+
"learning_rate": 0.00018133712149473295,
|
| 520 |
+
"loss": 4.9222,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 4.807817589576548,
|
| 525 |
+
"grad_norm": 300349.8165475781,
|
| 526 |
+
"learning_rate": 0.0001807054577092122,
|
| 527 |
+
"loss": 4.8383,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 4.872964169381108,
|
| 532 |
+
"grad_norm": 167685.76127581843,
|
| 533 |
+
"learning_rate": 0.0001800644267326728,
|
| 534 |
+
"loss": 4.9963,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 4.938110749185668,
|
| 539 |
+
"grad_norm": 408429.7693573758,
|
| 540 |
+
"learning_rate": 0.0001794141034310874,
|
| 541 |
+
"loss": 4.7736,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 5.0,
|
| 546 |
+
"grad_norm": 643531.8115077234,
|
| 547 |
+
"learning_rate": 0.00017875456375567882,
|
| 548 |
+
"loss": 4.0613,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 5.06514657980456,
|
| 553 |
+
"grad_norm": 139360.09034358786,
|
| 554 |
+
"learning_rate": 0.00017808588473404972,
|
| 555 |
+
"loss": 4.1228,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 5.130293159609121,
|
| 560 |
+
"grad_norm": 552603.9093684686,
|
| 561 |
+
"learning_rate": 0.00017740814446118644,
|
| 562 |
+
"loss": 4.1054,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 5.195439739413681,
|
| 567 |
+
"grad_norm": 47803.68753834838,
|
| 568 |
+
"learning_rate": 0.0001767214220903383,
|
| 569 |
+
"loss": 4.207,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 5.260586319218241,
|
| 574 |
+
"grad_norm": 534526.9523964288,
|
| 575 |
+
"learning_rate": 0.00017602579782377355,
|
| 576 |
+
"loss": 4.106,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 5.3257328990228014,
|
| 581 |
+
"grad_norm": 18180.552012122607,
|
| 582 |
+
"learning_rate": 0.00017532135290341215,
|
| 583 |
+
"loss": 3.8378,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 5.390879478827362,
|
| 588 |
+
"grad_norm": 127434.81313655287,
|
| 589 |
+
"learning_rate": 0.00017460816960133786,
|
| 590 |
+
"loss": 3.5962,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 5.456026058631922,
|
| 595 |
+
"grad_norm": 82109.10912999904,
|
| 596 |
+
"learning_rate": 0.00017388633121018948,
|
| 597 |
+
"loss": 3.5202,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 5.521172638436482,
|
| 602 |
+
"grad_norm": 26628.324243340692,
|
| 603 |
+
"learning_rate": 0.0001731559220334333,
|
| 604 |
+
"loss": 3.4476,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 5.586319218241043,
|
| 609 |
+
"grad_norm": 30172.421244564586,
|
| 610 |
+
"learning_rate": 0.00017241702737551704,
|
| 611 |
+
"loss": 3.3484,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 5.651465798045603,
|
| 616 |
+
"grad_norm": 89986.67442904277,
|
| 617 |
+
"learning_rate": 0.00017166973353190736,
|
| 618 |
+
"loss": 3.2529,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 5.716612377850163,
|
| 623 |
+
"grad_norm": 3649.2509810775036,
|
| 624 |
+
"learning_rate": 0.00017091412777901148,
|
| 625 |
+
"loss": 3.2057,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 5.7817589576547235,
|
| 630 |
+
"grad_norm": 4672.807059778061,
|
| 631 |
+
"learning_rate": 0.00017015029836398387,
|
| 632 |
+
"loss": 3.3986,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 5.846905537459284,
|
| 637 |
+
"grad_norm": 8238.275849020096,
|
| 638 |
+
"learning_rate": 0.00016937833449442018,
|
| 639 |
+
"loss": 3.3441,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 5.912052117263844,
|
| 644 |
+
"grad_norm": 13814.652802695186,
|
| 645 |
+
"learning_rate": 0.00016859832632793847,
|
| 646 |
+
"loss": 3.3955,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 5.977198697068404,
|
| 651 |
+
"grad_norm": 15843.655734535501,
|
| 652 |
+
"learning_rate": 0.00016781036496164986,
|
| 653 |
+
"loss": 3.2617,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 6.039087947882736,
|
| 658 |
+
"grad_norm": 5435.79510233699,
|
| 659 |
+
"learning_rate": 0.00016701454242151924,
|
| 660 |
+
"loss": 3.0216,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 6.1042345276872965,
|
| 665 |
+
"grad_norm": 8237.198800915166,
|
| 666 |
+
"learning_rate": 0.00016621095165161748,
|
| 667 |
+
"loss": 3.11,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 6.169381107491857,
|
| 672 |
+
"grad_norm": 10013.23246581901,
|
| 673 |
+
"learning_rate": 0.00016539968650326666,
|
| 674 |
+
"loss": 3.4611,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 6.234527687296417,
|
| 679 |
+
"grad_norm": 5540.320409136581,
|
| 680 |
+
"learning_rate": 0.00016458084172407914,
|
| 681 |
+
"loss": 3.2994,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 6.299674267100977,
|
| 686 |
+
"grad_norm": 16338.465194412813,
|
| 687 |
+
"learning_rate": 0.00016375451294689186,
|
| 688 |
+
"loss": 3.0775,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 6.364820846905538,
|
| 693 |
+
"grad_norm": 5639.958082276554,
|
| 694 |
+
"learning_rate": 0.00016292079667859743,
|
| 695 |
+
"loss": 2.974,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 6.429967426710098,
|
| 700 |
+
"grad_norm": 7948.806788711637,
|
| 701 |
+
"learning_rate": 0.00016207979028887333,
|
| 702 |
+
"loss": 2.8887,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 6.495114006514658,
|
| 707 |
+
"grad_norm": 8445.954428508405,
|
| 708 |
+
"learning_rate": 0.00016123159199880974,
|
| 709 |
+
"loss": 2.9355,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 6.5602605863192185,
|
| 714 |
+
"grad_norm": 8047.279585000703,
|
| 715 |
+
"learning_rate": 0.0001603763008694386,
|
| 716 |
+
"loss": 2.8092,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 6.625407166123779,
|
| 721 |
+
"grad_norm": 4571.549254094008,
|
| 722 |
+
"learning_rate": 0.0001595140167901641,
|
| 723 |
+
"loss": 2.874,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 6.690553745928339,
|
| 728 |
+
"grad_norm": 6990.386387450156,
|
| 729 |
+
"learning_rate": 0.00015864484046709674,
|
| 730 |
+
"loss": 2.8258,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 6.755700325732899,
|
| 735 |
+
"grad_norm": 10934.42401012609,
|
| 736 |
+
"learning_rate": 0.0001577688734112917,
|
| 737 |
+
"loss": 2.8676,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 6.82084690553746,
|
| 742 |
+
"grad_norm": 4770.988463103596,
|
| 743 |
+
"learning_rate": 0.0001568862179268935,
|
| 744 |
+
"loss": 2.9155,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 6.88599348534202,
|
| 749 |
+
"grad_norm": 8569.716482858448,
|
| 750 |
+
"learning_rate": 0.00015599697709918795,
|
| 751 |
+
"loss": 2.9029,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 6.95114006514658,
|
| 756 |
+
"grad_norm": 12037.093569327675,
|
| 757 |
+
"learning_rate": 0.0001551012547825626,
|
| 758 |
+
"loss": 2.7585,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 7.013029315960912,
|
| 763 |
+
"grad_norm": 123460.51811051978,
|
| 764 |
+
"learning_rate": 0.00015419915558837787,
|
| 765 |
+
"loss": 2.5385,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 7.078175895765472,
|
| 770 |
+
"grad_norm": 5195.06198589977,
|
| 771 |
+
"learning_rate": 0.00015329078487274925,
|
| 772 |
+
"loss": 2.7717,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 7.143322475570033,
|
| 777 |
+
"grad_norm": 752.1760900809323,
|
| 778 |
+
"learning_rate": 0.00015237624872424302,
|
| 779 |
+
"loss": 2.9416,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 7.208469055374593,
|
| 784 |
+
"grad_norm": 4491.061754834767,
|
| 785 |
+
"learning_rate": 0.00015145565395148586,
|
| 786 |
+
"loss": 2.8601,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 7.273615635179153,
|
| 791 |
+
"grad_norm": 816.8308540588761,
|
| 792 |
+
"learning_rate": 0.00015052910807069093,
|
| 793 |
+
"loss": 2.7076,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 7.3387622149837135,
|
| 798 |
+
"grad_norm": 1892.4705052365514,
|
| 799 |
+
"learning_rate": 0.00014959671929310102,
|
| 800 |
+
"loss": 2.8635,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 7.403908794788274,
|
| 805 |
+
"grad_norm": 1146.1506999359708,
|
| 806 |
+
"learning_rate": 0.00014865859651235042,
|
| 807 |
+
"loss": 3.0513,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 7.469055374592834,
|
| 812 |
+
"grad_norm": 2152.4545702128808,
|
| 813 |
+
"learning_rate": 0.00014771484929174735,
|
| 814 |
+
"loss": 3.3064,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 7.534201954397394,
|
| 819 |
+
"grad_norm": 796.6506977258991,
|
| 820 |
+
"learning_rate": 0.00014676558785147817,
|
| 821 |
+
"loss": 3.6776,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 7.599348534201955,
|
| 826 |
+
"grad_norm": 977.7148755282698,
|
| 827 |
+
"learning_rate": 0.00014581092305573446,
|
| 828 |
+
"loss": 3.6312,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 7.664495114006515,
|
| 833 |
+
"grad_norm": 1153.4547190060368,
|
| 834 |
+
"learning_rate": 0.00014485096639976557,
|
| 835 |
+
"loss": 4.1871,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 7.729641693811075,
|
| 840 |
+
"grad_norm": 1225.7726918509466,
|
| 841 |
+
"learning_rate": 0.00014388582999685682,
|
| 842 |
+
"loss": 4.0179,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 7.7947882736156355,
|
| 847 |
+
"grad_norm": 7924.5161323731445,
|
| 848 |
+
"learning_rate": 0.00014291562656523592,
|
| 849 |
+
"loss": 4.1581,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 7.859934853420196,
|
| 854 |
+
"grad_norm": 1227.8649922609834,
|
| 855 |
+
"learning_rate": 0.00014194046941490857,
|
| 856 |
+
"loss": 4.0944,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 7.925081433224756,
|
| 861 |
+
"grad_norm": 1116.0165164635077,
|
| 862 |
+
"learning_rate": 0.00014096047243442506,
|
| 863 |
+
"loss": 3.6464,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 7.990228013029316,
|
| 868 |
+
"grad_norm": 2612.9048314424645,
|
| 869 |
+
"learning_rate": 0.00013997575007757904,
|
| 870 |
+
"loss": 3.3371,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 8.052117263843648,
|
| 875 |
+
"grad_norm": 1097.6352872581833,
|
| 876 |
+
"learning_rate": 0.0001389864173500408,
|
| 877 |
+
"loss": 2.5671,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 8.117263843648209,
|
| 882 |
+
"grad_norm": 17489.044109497136,
|
| 883 |
+
"learning_rate": 0.00013799258979592537,
|
| 884 |
+
"loss": 2.5536,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 8.182410423452769,
|
| 889 |
+
"grad_norm": 108.71889860233144,
|
| 890 |
+
"learning_rate": 0.00013699438348429855,
|
| 891 |
+
"loss": 2.1741,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 8.247557003257329,
|
| 896 |
+
"grad_norm": 121.8614635887603,
|
| 897 |
+
"learning_rate": 0.0001359919149956208,
|
| 898 |
+
"loss": 2.0012,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 8.31270358306189,
|
| 903 |
+
"grad_norm": 90.7982839740096,
|
| 904 |
+
"learning_rate": 0.00013498530140813213,
|
| 905 |
+
"loss": 1.8773,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 8.37785016286645,
|
| 910 |
+
"grad_norm": 114.02461055718354,
|
| 911 |
+
"learning_rate": 0.0001339746602841783,
|
| 912 |
+
"loss": 1.8259,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 8.44299674267101,
|
| 917 |
+
"grad_norm": 50.644212949664,
|
| 918 |
+
"learning_rate": 0.00013296010965648092,
|
| 919 |
+
"loss": 1.7727,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 8.50814332247557,
|
| 924 |
+
"grad_norm": 75.12434167340028,
|
| 925 |
+
"learning_rate": 0.00013194176801435229,
|
| 926 |
+
"loss": 1.7434,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 8.57328990228013,
|
| 931 |
+
"grad_norm": 323.12805776081217,
|
| 932 |
+
"learning_rate": 0.00013091975428985712,
|
| 933 |
+
"loss": 1.6879,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 8.63843648208469,
|
| 938 |
+
"grad_norm": 44.34314085754787,
|
| 939 |
+
"learning_rate": 0.0001298941878439224,
|
| 940 |
+
"loss": 1.7317,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 8.703583061889251,
|
| 945 |
+
"grad_norm": 34.66401422616777,
|
| 946 |
+
"learning_rate": 0.00012886518845239722,
|
| 947 |
+
"loss": 1.7195,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 8.768729641693811,
|
| 952 |
+
"grad_norm": 16.4420136606676,
|
| 953 |
+
"learning_rate": 0.00012783287629206425,
|
| 954 |
+
"loss": 2.0396,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 8.833876221498372,
|
| 959 |
+
"grad_norm": 45.40704029721984,
|
| 960 |
+
"learning_rate": 0.00012679737192660418,
|
| 961 |
+
"loss": 2.0091,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 8.899022801302932,
|
| 966 |
+
"grad_norm": 6.429338872225894,
|
| 967 |
+
"learning_rate": 0.00012575879629251528,
|
| 968 |
+
"loss": 1.805,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 8.964169381107492,
|
| 973 |
+
"grad_norm": 3.5149539301232893,
|
| 974 |
+
"learning_rate": 0.00012471727068498904,
|
| 975 |
+
"loss": 1.6648,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 9.026058631921824,
|
| 980 |
+
"grad_norm": 6.295661720080446,
|
| 981 |
+
"learning_rate": 0.00012367291674374414,
|
| 982 |
+
"loss": 1.5264,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 9.091205211726384,
|
| 987 |
+
"grad_norm": 9.285329989011162,
|
| 988 |
+
"learning_rate": 0.00012262585643882027,
|
| 989 |
+
"loss": 1.5247,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 9.156351791530945,
|
| 994 |
+
"grad_norm": 25.069752054281583,
|
| 995 |
+
"learning_rate": 0.00012157621205633324,
|
| 996 |
+
"loss": 1.4634,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 9.221498371335505,
|
| 1001 |
+
"grad_norm": 35.13884590389422,
|
| 1002 |
+
"learning_rate": 0.00012052410618419296,
|
| 1003 |
+
"loss": 1.4801,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 9.286644951140065,
|
| 1008 |
+
"grad_norm": 3.5847961078697637,
|
| 1009 |
+
"learning_rate": 0.00011946966169778675,
|
| 1010 |
+
"loss": 1.5129,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 9.351791530944626,
|
| 1015 |
+
"grad_norm": 3.131424741610215,
|
| 1016 |
+
"learning_rate": 0.00011841300174562848,
|
| 1017 |
+
"loss": 1.3712,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 9.416938110749186,
|
| 1022 |
+
"grad_norm": 1.6795447164578916,
|
| 1023 |
+
"learning_rate": 0.00011735424973497615,
|
| 1024 |
+
"loss": 1.3106,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 9.482084690553746,
|
| 1029 |
+
"grad_norm": 2.8978809862796466,
|
| 1030 |
+
"learning_rate": 0.00011629352931741922,
|
| 1031 |
+
"loss": 1.2279,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 9.547231270358306,
|
| 1036 |
+
"grad_norm": 2.4257444318537402,
|
| 1037 |
+
"learning_rate": 0.00011523096437443719,
|
| 1038 |
+
"loss": 1.2547,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 9.612377850162867,
|
| 1043 |
+
"grad_norm": 2.067385872609714,
|
| 1044 |
+
"learning_rate": 0.00011416667900293163,
|
| 1045 |
+
"loss": 1.1879,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 9.677524429967427,
|
| 1050 |
+
"grad_norm": 3.37525338387685,
|
| 1051 |
+
"learning_rate": 0.00011310079750073288,
|
| 1052 |
+
"loss": 1.1027,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 9.742671009771987,
|
| 1057 |
+
"grad_norm": 1.3905025295542923,
|
| 1058 |
+
"learning_rate": 0.00011203344435208328,
|
| 1059 |
+
"loss": 1.0502,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 9.807817589576548,
|
| 1064 |
+
"grad_norm": 1.5645974543168704,
|
| 1065 |
+
"learning_rate": 0.0001109647442130987,
|
| 1066 |
+
"loss": 0.9847,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 9.872964169381108,
|
| 1071 |
+
"grad_norm": 5.493655656991333,
|
| 1072 |
+
"learning_rate": 0.00010989482189720984,
|
| 1073 |
+
"loss": 0.9334,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 9.938110749185668,
|
| 1078 |
+
"grad_norm": 0.9162795335385928,
|
| 1079 |
+
"learning_rate": 0.00010882380236058551,
|
| 1080 |
+
"loss": 0.9468,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 10.0,
|
| 1085 |
+
"grad_norm": 0.5839481455808426,
|
| 1086 |
+
"learning_rate": 0.0001077518106875388,
|
| 1087 |
+
"loss": 0.87,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 10.06514657980456,
|
| 1092 |
+
"grad_norm": 1.292014566062447,
|
| 1093 |
+
"learning_rate": 0.00010667897207591858,
|
| 1094 |
+
"loss": 0.9127,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 10.13029315960912,
|
| 1099 |
+
"grad_norm": 0.6920328719990964,
|
| 1100 |
+
"learning_rate": 0.00010560541182248757,
|
| 1101 |
+
"loss": 0.8961,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 10.19543973941368,
|
| 1106 |
+
"grad_norm": 0.9746662032078112,
|
| 1107 |
+
"learning_rate": 0.00010453125530828911,
|
| 1108 |
+
"loss": 0.8533,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 10.260586319218241,
|
| 1113 |
+
"grad_norm": 0.535332759720958,
|
| 1114 |
+
"learning_rate": 0.00010345662798400366,
|
| 1115 |
+
"loss": 0.8482,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 10.325732899022801,
|
| 1120 |
+
"grad_norm": 0.9207258473761247,
|
| 1121 |
+
"learning_rate": 0.00010238165535529769,
|
| 1122 |
+
"loss": 0.8592,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 10.390879478827362,
|
| 1127 |
+
"grad_norm": 0.4141514961953131,
|
| 1128 |
+
"learning_rate": 0.00010130646296816575,
|
| 1129 |
+
"loss": 0.8315,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 10.456026058631922,
|
| 1134 |
+
"grad_norm": 0.5290865267207626,
|
| 1135 |
+
"learning_rate": 0.00010023117639426783,
|
| 1136 |
+
"loss": 0.8287,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 10.521172638436482,
|
| 1141 |
+
"grad_norm": 0.457787741565417,
|
| 1142 |
+
"learning_rate": 9.915592121626411e-05,
|
| 1143 |
+
"loss": 0.8086,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 10.586319218241043,
|
| 1148 |
+
"grad_norm": 0.6310066669916005,
|
| 1149 |
+
"learning_rate": 9.808082301314807e-05,
|
| 1150 |
+
"loss": 0.8262,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 10.651465798045603,
|
| 1155 |
+
"grad_norm": 0.9842794557627913,
|
| 1156 |
+
"learning_rate": 9.700600734557991e-05,
|
| 1157 |
+
"loss": 0.8296,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 10.716612377850163,
|
| 1162 |
+
"grad_norm": 0.44754790431009867,
|
| 1163 |
+
"learning_rate": 9.593159974122282e-05,
|
| 1164 |
+
"loss": 0.8202,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 10.781758957654723,
|
| 1169 |
+
"grad_norm": 0.8753720051026775,
|
| 1170 |
+
"learning_rate": 9.485772568008201e-05,
|
| 1171 |
+
"loss": 0.8033,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 10.846905537459284,
|
| 1176 |
+
"grad_norm": 0.4060607138625304,
|
| 1177 |
+
"learning_rate": 9.378451057985046e-05,
|
| 1178 |
+
"loss": 0.8036,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 10.912052117263844,
|
| 1183 |
+
"grad_norm": 0.7413068695333563,
|
| 1184 |
+
"learning_rate": 9.27120797812609e-05,
|
| 1185 |
+
"loss": 0.8061,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 10.977198697068404,
|
| 1190 |
+
"grad_norm": 0.8808113137643704,
|
| 1191 |
+
"learning_rate": 9.164055853344773e-05,
|
| 1192 |
+
"loss": 0.8013,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 11.039087947882736,
|
| 1197 |
+
"grad_norm": 0.8269012104989563,
|
| 1198 |
+
"learning_rate": 9.057007197931878e-05,
|
| 1199 |
+
"loss": 0.7559,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 11.104234527687296,
|
| 1204 |
+
"grad_norm": 0.3509842614832935,
|
| 1205 |
+
"learning_rate": 8.950074514094029e-05,
|
| 1206 |
+
"loss": 0.8012,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 11.169381107491857,
|
| 1211 |
+
"grad_norm": 0.5643092458169484,
|
| 1212 |
+
"learning_rate": 8.843270290493505e-05,
|
| 1213 |
+
"loss": 0.7964,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 11.234527687296417,
|
| 1218 |
+
"grad_norm": 0.5454129605231703,
|
| 1219 |
+
"learning_rate": 8.73660700078975e-05,
|
| 1220 |
+
"loss": 0.8016,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 11.299674267100977,
|
| 1225 |
+
"grad_norm": 0.9029110960474988,
|
| 1226 |
+
"learning_rate": 8.630097102182532e-05,
|
| 1227 |
+
"loss": 0.8014,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 11.364820846905538,
|
| 1232 |
+
"grad_norm": 1.915017461526828,
|
| 1233 |
+
"learning_rate": 8.523753033957079e-05,
|
| 1234 |
+
"loss": 0.8004,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 11.429967426710098,
|
| 1239 |
+
"grad_norm": 0.624129144838355,
|
| 1240 |
+
"learning_rate": 8.417587216031311e-05,
|
| 1241 |
+
"loss": 0.8218,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 11.495114006514658,
|
| 1246 |
+
"grad_norm": 0.5902780870131191,
|
| 1247 |
+
"learning_rate": 8.311612047505286e-05,
|
| 1248 |
+
"loss": 0.8132,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 11.560260586319218,
|
| 1253 |
+
"grad_norm": 0.3545895480698459,
|
| 1254 |
+
"learning_rate": 8.205839905213149e-05,
|
| 1255 |
+
"loss": 0.7979,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 11.625407166123779,
|
| 1260 |
+
"grad_norm": 1.03763337939546,
|
| 1261 |
+
"learning_rate": 8.100283142277604e-05,
|
| 1262 |
+
"loss": 0.8051,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 11.690553745928339,
|
| 1267 |
+
"grad_norm": 0.5653887108382556,
|
| 1268 |
+
"learning_rate": 7.994954086667225e-05,
|
| 1269 |
+
"loss": 0.7774,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 11.7557003257329,
|
| 1274 |
+
"grad_norm": 0.2561014240557056,
|
| 1275 |
+
"learning_rate": 7.889865039756652e-05,
|
| 1276 |
+
"loss": 0.7895,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 11.82084690553746,
|
| 1281 |
+
"grad_norm": 0.3178103010456209,
|
| 1282 |
+
"learning_rate": 7.785028274889934e-05,
|
| 1283 |
+
"loss": 0.7794,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 11.88599348534202,
|
| 1288 |
+
"grad_norm": 0.5074253182959466,
|
| 1289 |
+
"learning_rate": 7.680456035947092e-05,
|
| 1290 |
+
"loss": 0.7781,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 11.95114006514658,
|
| 1295 |
+
"grad_norm": 0.3833459768679815,
|
| 1296 |
+
"learning_rate": 7.576160535914205e-05,
|
| 1297 |
+
"loss": 0.769,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 12.013029315960912,
|
| 1302 |
+
"grad_norm": 0.3836069440662029,
|
| 1303 |
+
"learning_rate": 7.472153955457007e-05,
|
| 1304 |
+
"loss": 0.7391,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 12.078175895765472,
|
| 1309 |
+
"grad_norm": 0.46645299339808305,
|
| 1310 |
+
"learning_rate": 7.368448441498348e-05,
|
| 1311 |
+
"loss": 0.7937,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 12.143322475570033,
|
| 1316 |
+
"grad_norm": 1.0910169170119404,
|
| 1317 |
+
"learning_rate": 7.265056105799523e-05,
|
| 1318 |
+
"loss": 0.778,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 12.208469055374593,
|
| 1323 |
+
"grad_norm": 3.652502613874099,
|
| 1324 |
+
"learning_rate": 7.161989023545773e-05,
|
| 1325 |
+
"loss": 0.7747,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 12.273615635179153,
|
| 1330 |
+
"grad_norm": 0.7890215278102028,
|
| 1331 |
+
"learning_rate": 7.059259231935984e-05,
|
| 1332 |
+
"loss": 0.7821,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 12.338762214983714,
|
| 1337 |
+
"grad_norm": 0.4552578053123255,
|
| 1338 |
+
"learning_rate": 6.956878728776907e-05,
|
| 1339 |
+
"loss": 0.7756,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 12.403908794788274,
|
| 1344 |
+
"grad_norm": 0.3791285043103609,
|
| 1345 |
+
"learning_rate": 6.854859471081908e-05,
|
| 1346 |
+
"loss": 0.7837,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 12.469055374592834,
|
| 1351 |
+
"grad_norm": 0.286532747668435,
|
| 1352 |
+
"learning_rate": 6.753213373674506e-05,
|
| 1353 |
+
"loss": 0.7601,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 12.534201954397394,
|
| 1358 |
+
"grad_norm": 0.28339934580974285,
|
| 1359 |
+
"learning_rate": 6.651952307796865e-05,
|
| 1360 |
+
"loss": 0.7805,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 12.599348534201955,
|
| 1365 |
+
"grad_norm": 0.48674093008241676,
|
| 1366 |
+
"learning_rate": 6.551088099723317e-05,
|
| 1367 |
+
"loss": 0.7636,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 12.664495114006515,
|
| 1372 |
+
"grad_norm": 0.4397598095537873,
|
| 1373 |
+
"learning_rate": 6.450632529379222e-05,
|
| 1374 |
+
"loss": 0.7683,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 12.729641693811075,
|
| 1379 |
+
"grad_norm": 0.35534540398829384,
|
| 1380 |
+
"learning_rate": 6.350597328965134e-05,
|
| 1381 |
+
"loss": 0.767,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 12.794788273615636,
|
| 1386 |
+
"grad_norm": 0.4017407288733161,
|
| 1387 |
+
"learning_rate": 6.250994181586645e-05,
|
| 1388 |
+
"loss": 0.778,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 12.859934853420196,
|
| 1393 |
+
"grad_norm": 0.2855817495240616,
|
| 1394 |
+
"learning_rate": 6.151834719889881e-05,
|
| 1395 |
+
"loss": 0.7668,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 12.925081433224756,
|
| 1400 |
+
"grad_norm": 0.4583358029208476,
|
| 1401 |
+
"learning_rate": 6.053130524702956e-05,
|
| 1402 |
+
"loss": 0.7757,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 12.990228013029316,
|
| 1407 |
+
"grad_norm": 0.637743389096925,
|
| 1408 |
+
"learning_rate": 5.954893123683423e-05,
|
| 1409 |
+
"loss": 0.7692,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 13.052117263843648,
|
| 1414 |
+
"grad_norm": 0.4872280397403082,
|
| 1415 |
+
"learning_rate": 5.857133989971976e-05,
|
| 1416 |
+
"loss": 0.7086,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 13.117263843648209,
|
| 1421 |
+
"grad_norm": 0.7382889654864129,
|
| 1422 |
+
"learning_rate": 5.759864540852496e-05,
|
| 1423 |
+
"loss": 0.7729,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 13.182410423452769,
|
| 1428 |
+
"grad_norm": 1.1679070563901006,
|
| 1429 |
+
"learning_rate": 5.6630961364186194e-05,
|
| 1430 |
+
"loss": 0.7802,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 13.247557003257329,
|
| 1435 |
+
"grad_norm": 1.0300430714340285,
|
| 1436 |
+
"learning_rate": 5.566840078247007e-05,
|
| 1437 |
+
"loss": 0.7682,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 13.31270358306189,
|
| 1442 |
+
"grad_norm": 2.335343336253081,
|
| 1443 |
+
"learning_rate": 5.4711076080774295e-05,
|
| 1444 |
+
"loss": 0.7602,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 13.37785016286645,
|
| 1449 |
+
"grad_norm": 2.3435030206437273,
|
| 1450 |
+
"learning_rate": 5.3759099064998485e-05,
|
| 1451 |
+
"loss": 0.7733,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 13.44299674267101,
|
| 1456 |
+
"grad_norm": 5.352928922415569,
|
| 1457 |
+
"learning_rate": 5.281258091648611e-05,
|
| 1458 |
+
"loss": 0.7572,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 13.50814332247557,
|
| 1463 |
+
"grad_norm": 2.8250685699424154,
|
| 1464 |
+
"learning_rate": 5.1871632179040106e-05,
|
| 1465 |
+
"loss": 0.7633,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 13.57328990228013,
|
| 1470 |
+
"grad_norm": 4.112768874107428,
|
| 1471 |
+
"learning_rate": 5.0936362746011914e-05,
|
| 1472 |
+
"loss": 0.7559,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 13.63843648208469,
|
| 1477 |
+
"grad_norm": 4.034092026025879,
|
| 1478 |
+
"learning_rate": 5.0006881847467535e-05,
|
| 1479 |
+
"loss": 0.7567,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 13.703583061889251,
|
| 1484 |
+
"grad_norm": 3.2740509769772776,
|
| 1485 |
+
"learning_rate": 4.908329803743022e-05,
|
| 1486 |
+
"loss": 0.7604,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 13.768729641693811,
|
| 1491 |
+
"grad_norm": 1.3398784592419952,
|
| 1492 |
+
"learning_rate": 4.8165719181202686e-05,
|
| 1493 |
+
"loss": 0.7537,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 13.833876221498372,
|
| 1498 |
+
"grad_norm": 2.726267907133913,
|
| 1499 |
+
"learning_rate": 4.7254252442769324e-05,
|
| 1500 |
+
"loss": 0.7646,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 13.899022801302932,
|
| 1505 |
+
"grad_norm": 3.2217538293852854,
|
| 1506 |
+
"learning_rate": 4.634900427228067e-05,
|
| 1507 |
+
"loss": 0.7499,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 13.964169381107492,
|
| 1512 |
+
"grad_norm": 3.7243300540961504,
|
| 1513 |
+
"learning_rate": 4.5450080393621e-05,
|
| 1514 |
+
"loss": 0.7557,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 14.026058631921824,
|
| 1519 |
+
"grad_norm": 4.513289667701461,
|
| 1520 |
+
"learning_rate": 4.455758579206098e-05,
|
| 1521 |
+
"loss": 0.7137,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 14.091205211726384,
|
| 1526 |
+
"grad_norm": 2.21160147053622,
|
| 1527 |
+
"learning_rate": 4.367162470199627e-05,
|
| 1528 |
+
"loss": 0.738,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 14.156351791530945,
|
| 1533 |
+
"grad_norm": 1.7777659007762496,
|
| 1534 |
+
"learning_rate": 4.279230059477416e-05,
|
| 1535 |
+
"loss": 0.7493,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 14.221498371335505,
|
| 1540 |
+
"grad_norm": 1.8914544450552935,
|
| 1541 |
+
"learning_rate": 4.191971616660904e-05,
|
| 1542 |
+
"loss": 0.7463,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 14.286644951140065,
|
| 1547 |
+
"grad_norm": 1.7668273761342166,
|
| 1548 |
+
"learning_rate": 4.105397332658849e-05,
|
| 1549 |
+
"loss": 0.7507,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 14.351791530944626,
|
| 1554 |
+
"grad_norm": 1.0381981672431264,
|
| 1555 |
+
"learning_rate": 4.019517318477144e-05,
|
| 1556 |
+
"loss": 0.7558,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 14.416938110749186,
|
| 1561 |
+
"grad_norm": 2.14958675055386,
|
| 1562 |
+
"learning_rate": 3.934341604037951e-05,
|
| 1563 |
+
"loss": 0.7458,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 14.482084690553746,
|
| 1568 |
+
"grad_norm": 2.5762079835127203,
|
| 1569 |
+
"learning_rate": 3.849880137008303e-05,
|
| 1570 |
+
"loss": 0.7566,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 14.547231270358306,
|
| 1575 |
+
"grad_norm": 3.474638671972531,
|
| 1576 |
+
"learning_rate": 3.7661427816383106e-05,
|
| 1577 |
+
"loss": 0.7514,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 14.612377850162867,
|
| 1582 |
+
"grad_norm": 3.060408194609065,
|
| 1583 |
+
"learning_rate": 3.683139317609128e-05,
|
| 1584 |
+
"loss": 0.7323,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 14.677524429967427,
|
| 1589 |
+
"grad_norm": 2.12967575995364,
|
| 1590 |
+
"learning_rate": 3.6008794388907754e-05,
|
| 1591 |
+
"loss": 0.7467,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 14.742671009771987,
|
| 1596 |
+
"grad_norm": 2.94138450728384,
|
| 1597 |
+
"learning_rate": 3.519372752609995e-05,
|
| 1598 |
+
"loss": 0.7557,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 14.807817589576548,
|
| 1603 |
+
"grad_norm": 3.1242139515554608,
|
| 1604 |
+
"learning_rate": 3.438628777928213e-05,
|
| 1605 |
+
"loss": 0.7407,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 14.872964169381108,
|
| 1610 |
+
"grad_norm": 3.1286187630032853,
|
| 1611 |
+
"learning_rate": 3.358656944929822e-05,
|
| 1612 |
+
"loss": 0.7391,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 14.938110749185668,
|
| 1617 |
+
"grad_norm": 3.21626407861878,
|
| 1618 |
+
"learning_rate": 3.279466593520821e-05,
|
| 1619 |
+
"loss": 0.7388,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 15.0,
|
| 1624 |
+
"grad_norm": 2.667088088744998,
|
| 1625 |
+
"learning_rate": 3.2010669723380174e-05,
|
| 1626 |
+
"loss": 0.7025,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 15.06514657980456,
|
| 1631 |
+
"grad_norm": 2.6455747181491716,
|
| 1632 |
+
"learning_rate": 3.123467237668881e-05,
|
| 1633 |
+
"loss": 0.7297,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 15.13029315960912,
|
| 1638 |
+
"grad_norm": 3.014602988595856,
|
| 1639 |
+
"learning_rate": 3.0466764523821823e-05,
|
| 1640 |
+
"loss": 0.7332,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 15.19543973941368,
|
| 1645 |
+
"grad_norm": 2.3597277220635244,
|
| 1646 |
+
"learning_rate": 2.970703584869529e-05,
|
| 1647 |
+
"loss": 0.7521,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 15.260586319218241,
|
| 1652 |
+
"grad_norm": 4.127523198748339,
|
| 1653 |
+
"learning_rate": 2.8955575079979703e-05,
|
| 1654 |
+
"loss": 0.7415,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 15.325732899022801,
|
| 1659 |
+
"grad_norm": 2.524034767343291,
|
| 1660 |
+
"learning_rate": 2.8212469980737084e-05,
|
| 1661 |
+
"loss": 0.7371,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 15.390879478827362,
|
| 1666 |
+
"grad_norm": 4.719745727828448,
|
| 1667 |
+
"learning_rate": 2.7477807338171323e-05,
|
| 1668 |
+
"loss": 0.7271,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 15.456026058631922,
|
| 1673 |
+
"grad_norm": 5.244868726918304,
|
| 1674 |
+
"learning_rate": 2.6751672953492246e-05,
|
| 1675 |
+
"loss": 0.7366,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 15.521172638436482,
|
| 1680 |
+
"grad_norm": 5.627379282758129,
|
| 1681 |
+
"learning_rate": 2.603415163189495e-05,
|
| 1682 |
+
"loss": 0.74,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 15.586319218241043,
|
| 1687 |
+
"grad_norm": 8.949478408325684,
|
| 1688 |
+
"learning_rate": 2.532532717265535e-05,
|
| 1689 |
+
"loss": 0.735,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 15.651465798045603,
|
| 1694 |
+
"grad_norm": 6.544025376409119,
|
| 1695 |
+
"learning_rate": 2.4625282359343252e-05,
|
| 1696 |
+
"loss": 0.7257,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 15.716612377850163,
|
| 1701 |
+
"grad_norm": 12.205912057308234,
|
| 1702 |
+
"learning_rate": 2.393409895015411e-05,
|
| 1703 |
+
"loss": 0.7315,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 15.781758957654723,
|
| 1708 |
+
"grad_norm": 5.541069988378192,
|
| 1709 |
+
"learning_rate": 2.325185766836044e-05,
|
| 1710 |
+
"loss": 0.741,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 15.846905537459284,
|
| 1715 |
+
"grad_norm": 10.365946509926852,
|
| 1716 |
+
"learning_rate": 2.2578638192884295e-05,
|
| 1717 |
+
"loss": 0.7293,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 15.912052117263844,
|
| 1722 |
+
"grad_norm": 6.954057714885383,
|
| 1723 |
+
"learning_rate": 2.191451914899132e-05,
|
| 1724 |
+
"loss": 0.7135,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 15.977198697068404,
|
| 1729 |
+
"grad_norm": 6.836607642747595,
|
| 1730 |
+
"learning_rate": 2.1259578099108383e-05,
|
| 1731 |
+
"loss": 0.7093,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 16.039087947882734,
|
| 1736 |
+
"grad_norm": 6.532142502787816,
|
| 1737 |
+
"learning_rate": 2.061389153376486e-05,
|
| 1738 |
+
"loss": 0.6879,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 16.104234527687296,
|
| 1743 |
+
"grad_norm": 7.379167777628348,
|
| 1744 |
+
"learning_rate": 1.997753486265938e-05,
|
| 1745 |
+
"loss": 0.7236,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 16.169381107491855,
|
| 1750 |
+
"grad_norm": 5.627624131311463,
|
| 1751 |
+
"learning_rate": 1.9350582405852738e-05,
|
| 1752 |
+
"loss": 0.7242,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 16.234527687296417,
|
| 1757 |
+
"grad_norm": 2.566026898760179,
|
| 1758 |
+
"learning_rate": 1.8733107385088118e-05,
|
| 1759 |
+
"loss": 0.7282,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 16.299674267100976,
|
| 1764 |
+
"grad_norm": 3.2443731023533733,
|
| 1765 |
+
"learning_rate": 1.8125181915239373e-05,
|
| 1766 |
+
"loss": 0.7205,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 16.364820846905538,
|
| 1771 |
+
"grad_norm": 1.8082851262200852,
|
| 1772 |
+
"learning_rate": 1.752687699588892e-05,
|
| 1773 |
+
"loss": 0.7239,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 16.429967426710096,
|
| 1778 |
+
"grad_norm": 3.031983369049924,
|
| 1779 |
+
"learning_rate": 1.693826250303555e-05,
|
| 1780 |
+
"loss": 0.722,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 16.495114006514658,
|
| 1785 |
+
"grad_norm": 7.290608888275902,
|
| 1786 |
+
"learning_rate": 1.635940718093364e-05,
|
| 1787 |
+
"loss": 0.7213,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 16.560260586319217,
|
| 1792 |
+
"grad_norm": 2.2112949986206463,
|
| 1793 |
+
"learning_rate": 1.579037863406466e-05,
|
| 1794 |
+
"loss": 0.7179,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 16.62540716612378,
|
| 1799 |
+
"grad_norm": 1.907485565515624,
|
| 1800 |
+
"learning_rate": 1.5231243319241426e-05,
|
| 1801 |
+
"loss": 0.7217,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 16.690553745928337,
|
| 1806 |
+
"grad_norm": 1.4831231394781539,
|
| 1807 |
+
"learning_rate": 1.4682066537846806e-05,
|
| 1808 |
+
"loss": 0.727,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 16.7557003257329,
|
| 1813 |
+
"grad_norm": 2.4017105738678657,
|
| 1814 |
+
"learning_rate": 1.4142912428207048e-05,
|
| 1815 |
+
"loss": 0.7097,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 16.820846905537458,
|
| 1820 |
+
"grad_norm": 1.2214677159578557,
|
| 1821 |
+
"learning_rate": 1.3613843958101124e-05,
|
| 1822 |
+
"loss": 0.7142,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 16.88599348534202,
|
| 1827 |
+
"grad_norm": 1.4630974450979373,
|
| 1828 |
+
"learning_rate": 1.3094922917406697e-05,
|
| 1829 |
+
"loss": 0.7166,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 16.95114006514658,
|
| 1834 |
+
"grad_norm": 1.1100744058623144,
|
| 1835 |
+
"learning_rate": 1.2586209910883754e-05,
|
| 1836 |
+
"loss": 0.7096,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 17.01302931596091,
|
| 1841 |
+
"grad_norm": 0.7228289539451794,
|
| 1842 |
+
"learning_rate": 1.2087764351096478e-05,
|
| 1843 |
+
"loss": 0.6779,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 17.078175895765472,
|
| 1848 |
+
"grad_norm": 0.7760436517259812,
|
| 1849 |
+
"learning_rate": 1.1599644451474582e-05,
|
| 1850 |
+
"loss": 0.717,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 17.14332247557003,
|
| 1855 |
+
"grad_norm": 0.8704424517484689,
|
| 1856 |
+
"learning_rate": 1.1121907219514365e-05,
|
| 1857 |
+
"loss": 0.7093,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 17.208469055374593,
|
| 1862 |
+
"grad_norm": 0.9618205532410652,
|
| 1863 |
+
"learning_rate": 1.0654608450121079e-05,
|
| 1864 |
+
"loss": 0.7207,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 17.27361563517915,
|
| 1869 |
+
"grad_norm": 1.5516164758839461,
|
| 1870 |
+
"learning_rate": 1.0197802719092398e-05,
|
| 1871 |
+
"loss": 0.7071,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 17.338762214983714,
|
| 1876 |
+
"grad_norm": 1.5058404804541428,
|
| 1877 |
+
"learning_rate": 9.751543376744696e-06,
|
| 1878 |
+
"loss": 0.7226,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 17.403908794788272,
|
| 1883 |
+
"grad_norm": 0.8200647473487829,
|
| 1884 |
+
"learning_rate": 9.315882541682124e-06,
|
| 1885 |
+
"loss": 0.7181,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 17.469055374592834,
|
| 1890 |
+
"grad_norm": 1.4904450854071307,
|
| 1891 |
+
"learning_rate": 8.890871094709834e-06,
|
| 1892 |
+
"loss": 0.6974,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 17.534201954397393,
|
| 1897 |
+
"grad_norm": 1.2260823155777854,
|
| 1898 |
+
"learning_rate": 8.476558672891394e-06,
|
| 1899 |
+
"loss": 0.7083,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 17.599348534201955,
|
| 1904 |
+
"grad_norm": 0.9817270489066919,
|
| 1905 |
+
"learning_rate": 8.072993663751926e-06,
|
| 1906 |
+
"loss": 0.7075,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 17.664495114006513,
|
| 1911 |
+
"grad_norm": 1.4879300157917899,
|
| 1912 |
+
"learning_rate": 7.680223199626776e-06,
|
| 1913 |
+
"loss": 0.6973,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 17.729641693811075,
|
| 1918 |
+
"grad_norm": 0.9033769549461519,
|
| 1919 |
+
"learning_rate": 7.298293152156954e-06,
|
| 1920 |
+
"loss": 0.6992,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 17.794788273615634,
|
| 1925 |
+
"grad_norm": 0.8407800924507985,
|
| 1926 |
+
"learning_rate": 6.927248126931845e-06,
|
| 1927 |
+
"loss": 0.7001,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 17.859934853420196,
|
| 1932 |
+
"grad_norm": 1.1797502166968346,
|
| 1933 |
+
"learning_rate": 6.567131458279583e-06,
|
| 1934 |
+
"loss": 0.7024,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 17.925081433224754,
|
| 1939 |
+
"grad_norm": 0.8163306907085345,
|
| 1940 |
+
"learning_rate": 6.217985204206271e-06,
|
| 1941 |
+
"loss": 0.6887,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 17.990228013029316,
|
| 1946 |
+
"grad_norm": 0.7084964296167927,
|
| 1947 |
+
"learning_rate": 5.879850141483774e-06,
|
| 1948 |
+
"loss": 0.7045,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 18.05211726384365,
|
| 1953 |
+
"grad_norm": 1.0977073568072095,
|
| 1954 |
+
"learning_rate": 5.552765760887597e-06,
|
| 1955 |
+
"loss": 0.6598,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 18.117263843648207,
|
| 1960 |
+
"grad_norm": 0.6301620030583781,
|
| 1961 |
+
"learning_rate": 5.236770262584656e-06,
|
| 1962 |
+
"loss": 0.6996,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 18.18241042345277,
|
| 1967 |
+
"grad_norm": 0.7687636324981993,
|
| 1968 |
+
"learning_rate": 4.931900551671912e-06,
|
| 1969 |
+
"loss": 0.6897,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 18.247557003257327,
|
| 1974 |
+
"grad_norm": 0.62708546479139,
|
| 1975 |
+
"learning_rate": 4.638192233866213e-06,
|
| 1976 |
+
"loss": 0.6719,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 18.31270358306189,
|
| 1981 |
+
"grad_norm": 0.8725798711853024,
|
| 1982 |
+
"learning_rate": 4.355679611345943e-06,
|
| 1983 |
+
"loss": 0.6956,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 18.377850162866448,
|
| 1988 |
+
"grad_norm": 0.994881624196576,
|
| 1989 |
+
"learning_rate": 4.084395678744831e-06,
|
| 1990 |
+
"loss": 0.6954,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 18.44299674267101,
|
| 1995 |
+
"grad_norm": 1.533230204403063,
|
| 1996 |
+
"learning_rate": 3.824372119298525e-06,
|
| 1997 |
+
"loss": 0.6901,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 18.50814332247557,
|
| 2002 |
+
"grad_norm": 0.8204728533453552,
|
| 2003 |
+
"learning_rate": 3.5756393011443216e-06,
|
| 2004 |
+
"loss": 0.6934,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 18.57328990228013,
|
| 2009 |
+
"grad_norm": 1.2898266855528728,
|
| 2010 |
+
"learning_rate": 3.3382262737744265e-06,
|
| 2011 |
+
"loss": 0.6873,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 18.63843648208469,
|
| 2016 |
+
"grad_norm": 1.0536146862019349,
|
| 2017 |
+
"learning_rate": 3.112160764643329e-06,
|
| 2018 |
+
"loss": 0.6742,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 18.70358306188925,
|
| 2023 |
+
"grad_norm": 0.7526257737230756,
|
| 2024 |
+
"learning_rate": 2.8974691759294716e-06,
|
| 2025 |
+
"loss": 0.6943,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 18.76872964169381,
|
| 2030 |
+
"grad_norm": 0.6092087200887528,
|
| 2031 |
+
"learning_rate": 2.6941765814517727e-06,
|
| 2032 |
+
"loss": 0.6799,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 18.83387622149837,
|
| 2037 |
+
"grad_norm": 0.8917616389127406,
|
| 2038 |
+
"learning_rate": 2.502306723741213e-06,
|
| 2039 |
+
"loss": 0.683,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 18.89902280130293,
|
| 2044 |
+
"grad_norm": 0.7505448190720103,
|
| 2045 |
+
"learning_rate": 2.3218820112679638e-06,
|
| 2046 |
+
"loss": 0.6946,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 18.964169381107492,
|
| 2051 |
+
"grad_norm": 1.1035154305208421,
|
| 2052 |
+
"learning_rate": 2.152923515824342e-06,
|
| 2053 |
+
"loss": 0.6865,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 19.026058631921824,
|
| 2058 |
+
"grad_norm": 0.6583696250705815,
|
| 2059 |
+
"learning_rate": 1.9954509700637774e-06,
|
| 2060 |
+
"loss": 0.6651,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 19.091205211726383,
|
| 2065 |
+
"grad_norm": 0.5679618492990128,
|
| 2066 |
+
"learning_rate": 1.8494827651962902e-06,
|
| 2067 |
+
"loss": 0.6768,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 19.156351791530945,
|
| 2072 |
+
"grad_norm": 0.8108819287427305,
|
| 2073 |
+
"learning_rate": 1.7150359488405348e-06,
|
| 2074 |
+
"loss": 0.6935,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 19.221498371335503,
|
| 2079 |
+
"grad_norm": 0.6126197571813816,
|
| 2080 |
+
"learning_rate": 1.5921262230328377e-06,
|
| 2081 |
+
"loss": 0.6859,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 19.286644951140065,
|
| 2086 |
+
"grad_norm": 0.7187833312338922,
|
| 2087 |
+
"learning_rate": 1.4807679423933651e-06,
|
| 2088 |
+
"loss": 0.6963,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 19.351791530944624,
|
| 2093 |
+
"grad_norm": 0.5459264689232999,
|
| 2094 |
+
"learning_rate": 1.3809741124495957e-06,
|
| 2095 |
+
"loss": 0.6725,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 19.416938110749186,
|
| 2100 |
+
"grad_norm": 0.8847256765678589,
|
| 2101 |
+
"learning_rate": 1.2927563881174957e-06,
|
| 2102 |
+
"loss": 0.6803,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 19.482084690553744,
|
| 2107 |
+
"grad_norm": 0.8116732582437327,
|
| 2108 |
+
"learning_rate": 1.2161250723402427e-06,
|
| 2109 |
+
"loss": 0.6817,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 19.547231270358306,
|
| 2114 |
+
"grad_norm": 1.2980935401569431,
|
| 2115 |
+
"learning_rate": 1.1510891148850406e-06,
|
| 2116 |
+
"loss": 0.6866,
|
| 2117 |
+
"step": 3010
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 19.612377850162865,
|
| 2121 |
+
"grad_norm": 1.6154894036939988,
|
| 2122 |
+
"learning_rate": 1.097656111297777e-06,
|
| 2123 |
+
"loss": 0.6609,
|
| 2124 |
+
"step": 3020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 19.677524429967427,
|
| 2128 |
+
"grad_norm": 0.7892564418205148,
|
| 2129 |
+
"learning_rate": 1.0558323020160294e-06,
|
| 2130 |
+
"loss": 0.6644,
|
| 2131 |
+
"step": 3030
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 19.742671009771986,
|
| 2135 |
+
"grad_norm": 1.0743865740819893,
|
| 2136 |
+
"learning_rate": 1.025622571640201e-06,
|
| 2137 |
+
"loss": 0.6834,
|
| 2138 |
+
"step": 3040
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 19.807817589576548,
|
| 2142 |
+
"grad_norm": 1.0123399238773383,
|
| 2143 |
+
"learning_rate": 1.0070304483630465e-06,
|
| 2144 |
+
"loss": 0.6744,
|
| 2145 |
+
"step": 3050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 19.872964169381106,
|
| 2149 |
+
"grad_norm": 0.5968908325725324,
|
| 2150 |
+
"learning_rate": 1.0000581035576305e-06,
|
| 2151 |
+
"loss": 0.6816,
|
| 2152 |
+
"step": 3060
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 19.872964169381106,
|
| 2156 |
+
"step": 3060,
|
| 2157 |
+
"total_flos": 2755104923451392.0,
|
| 2158 |
+
"train_loss": 1.9120042853885226,
|
| 2159 |
+
"train_runtime": 46649.9639,
|
| 2160 |
+
"train_samples_per_second": 25.234,
|
| 2161 |
+
"train_steps_per_second": 0.066
|
| 2162 |
+
}
|
| 2163 |
+
],
|
| 2164 |
+
"logging_steps": 10,
|
| 2165 |
+
"max_steps": 3060,
|
| 2166 |
+
"num_input_tokens_seen": 0,
|
| 2167 |
+
"num_train_epochs": 20,
|
| 2168 |
+
"save_steps": 5000,
|
| 2169 |
+
"stateful_callbacks": {
|
| 2170 |
+
"TrainerControl": {
|
| 2171 |
+
"args": {
|
| 2172 |
+
"should_epoch_stop": false,
|
| 2173 |
+
"should_evaluate": false,
|
| 2174 |
+
"should_log": false,
|
| 2175 |
+
"should_save": true,
|
| 2176 |
+
"should_training_stop": true
|
| 2177 |
+
},
|
| 2178 |
+
"attributes": {}
|
| 2179 |
+
}
|
| 2180 |
+
},
|
| 2181 |
+
"total_flos": 2755104923451392.0,
|
| 2182 |
+
"train_batch_size": 24,
|
| 2183 |
+
"trial_name": null,
|
| 2184 |
+
"trial_params": null
|
| 2185 |
+
}
|