diff --git a/.gitattributes b/.gitattributes index 663be8f22c41aafec1025d408d15c2745daf359b..ae3da6fd548d2201ef2c5d909a9abbd9d51d2c7b 100644 --- a/.gitattributes +++ b/.gitattributes @@ -236,3 +236,5 @@ babylm_shuffle_local3_10M_seed0/runs/checkpoint-1200/tokenizer.json filter=lfs d babylm_shuffle_local3_10M_seed0/runs/checkpoint-1100/tokenizer.json filter=lfs diff=lfs merge=lfs -text babylm_shuffle_local3_10M_seed0/runs/checkpoint-1700/tokenizer.json filter=lfs diff=lfs merge=lfs -text babylm_shuffle_local3_10M_seed0/runs/checkpoint-2200/tokenizer.json filter=lfs diff=lfs merge=lfs -text +babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/tokenizer.json filter=lfs diff=lfs merge=lfs -text +babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-100/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/latest new file mode 100644 index 0000000000000000000000000000000000000000..22cd5c3402316b70299aed2025d7943595f5d495 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/latest @@ -0,0 +1 @@ +global_step1100 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..71065759a213ed5141771044da99429fe8001bde --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/trainer_state.json @@ -0,0 +1,8613 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.2054794520547945, + "eval_steps": 10, + "global_step": 1100, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6624906063079834, + "learning_rate": 4.77051177904143e-06, + "loss": 1.704, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 3.0075268745422363, + "learning_rate": 4.768480909829407e-06, + "loss": 1.6322, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.6070828437805176, + "learning_rate": 4.766450040617385e-06, + "loss": 1.6871, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 3.044222831726074, + "learning_rate": 4.764419171405362e-06, + "loss": 1.6652, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.773167610168457, + "learning_rate": 4.762388302193339e-06, + "loss": 1.7063, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.880176305770874, + "learning_rate": 4.760357432981316e-06, + "loss": 1.6775, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.6402764320373535, + "learning_rate": 4.758326563769294e-06, + "loss": 1.6694, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.9245846271514893, + "learning_rate": 4.756295694557271e-06, + "loss": 1.6688, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.623044967651367, + "learning_rate": 4.754264825345248e-06, + "loss": 1.6819, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.739896059036255, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.6401, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.8707164525985718, + "eval_runtime": 36.9513, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.603599786758423, + "learning_rate": 4.750203086921203e-06, + "loss": 1.6412, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.5571675300598145, + "learning_rate": 4.74817221770918e-06, + "loss": 1.5779, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.87091326713562, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.6878, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.8140673637390137, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7023, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.6550583839416504, + "learning_rate": 4.742079610073111e-06, + "loss": 1.5995, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.6259925365448, + "learning_rate": 4.740048740861089e-06, + "loss": 1.6194, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.971848726272583, + "learning_rate": 4.738017871649066e-06, + "loss": 1.6634, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6344778537750244, + "learning_rate": 4.735987002437044e-06, + "loss": 1.6502, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 2.9362456798553467, + "learning_rate": 4.73395613322502e-06, + "loss": 1.6948, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.8098835945129395, + "learning_rate": 4.731925264012998e-06, + "loss": 1.7161, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.8724393844604492, + "eval_runtime": 37.0876, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.933009386062622, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7014, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.8216991424560547, + "learning_rate": 4.727863525588953e-06, + "loss": 1.6192, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.7028493881225586, + "learning_rate": 4.725832656376929e-06, + "loss": 1.607, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.816818952560425, + "learning_rate": 4.723801787164907e-06, + "loss": 1.6183, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.7707865238189697, + "learning_rate": 4.721770917952885e-06, + "loss": 1.6795, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.779562473297119, + "learning_rate": 4.719740048740861e-06, + "loss": 1.63, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.907338857650757, + "learning_rate": 4.717709179528839e-06, + "loss": 1.644, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 3.073558807373047, + "learning_rate": 4.715678310316816e-06, + "loss": 1.6821, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.7556493282318115, + "learning_rate": 4.713647441104794e-06, + "loss": 1.6501, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6718344688415527, + "learning_rate": 4.71161657189277e-06, + "loss": 1.6225, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.854687213897705, + "eval_runtime": 36.9773, + "eval_samples_per_second": 27.044, + "eval_steps_per_second": 1.136, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.763335943222046, + "learning_rate": 4.709585702680748e-06, + "loss": 1.6311, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.7669479846954346, + "learning_rate": 4.707554833468725e-06, + "loss": 1.6809, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.711562395095825, + "learning_rate": 4.705523964256703e-06, + "loss": 1.741, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.799954652786255, + "learning_rate": 4.703493095044679e-06, + "loss": 1.6409, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.7278690338134766, + "learning_rate": 4.701462225832657e-06, + "loss": 1.6325, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.6802427768707275, + "learning_rate": 4.699431356620634e-06, + "loss": 1.5988, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.8826160430908203, + "learning_rate": 4.697400487408612e-06, + "loss": 1.7119, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.784343719482422, + "learning_rate": 4.695369618196588e-06, + "loss": 1.6855, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.683337688446045, + "learning_rate": 4.693338748984566e-06, + "loss": 1.6588, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.6646857261657715, + "learning_rate": 4.691307879772543e-06, + "loss": 1.623, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.8511513471603394, + "eval_runtime": 36.9852, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.8181183338165283, + "learning_rate": 4.68927701056052e-06, + "loss": 1.6171, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.8270857334136963, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.678, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.7685446739196777, + "learning_rate": 4.685215272136475e-06, + "loss": 1.6575, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.7003183364868164, + "learning_rate": 4.683184402924452e-06, + "loss": 1.6405, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.668750286102295, + "learning_rate": 4.681153533712429e-06, + "loss": 1.6541, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 3.065800905227661, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7288, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.615601062774658, + "learning_rate": 4.677091795288384e-06, + "loss": 1.6306, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.5948891639709473, + "learning_rate": 4.675060926076361e-06, + "loss": 1.6018, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.7550079822540283, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7009, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.868687868118286, + "learning_rate": 4.670999187652316e-06, + "loss": 1.6034, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.8473671674728394, + "eval_runtime": 37.0015, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.705962657928467, + "learning_rate": 4.668968318440293e-06, + "loss": 1.6084, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.7901813983917236, + "learning_rate": 4.66693744922827e-06, + "loss": 1.6171, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.688774824142456, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.6774, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.734617233276367, + "learning_rate": 4.662875710804225e-06, + "loss": 1.5796, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.742017984390259, + "learning_rate": 4.660844841592202e-06, + "loss": 1.6671, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.6438794136047363, + "learning_rate": 4.658813972380179e-06, + "loss": 1.5835, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6441051959991455, + "learning_rate": 4.656783103168157e-06, + "loss": 1.6909, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.824204206466675, + "learning_rate": 4.654752233956134e-06, + "loss": 1.5758, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.556638479232788, + "learning_rate": 4.652721364744111e-06, + "loss": 1.6504, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.897812843322754, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.6406, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.8537043333053589, + "eval_runtime": 37.257, + "eval_samples_per_second": 26.841, + "eval_steps_per_second": 1.127, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.6594088077545166, + "learning_rate": 4.648659626320066e-06, + "loss": 1.6966, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.79524302482605, + "learning_rate": 4.646628757108043e-06, + "loss": 1.6645, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.80291748046875, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.6281, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.7723848819732666, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.5959, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 3.0625650882720947, + "learning_rate": 4.640536149471974e-06, + "loss": 1.6313, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.775726079940796, + "learning_rate": 4.638505280259952e-06, + "loss": 1.6563, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.837758779525757, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.6727, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 3.041355609893799, + "learning_rate": 4.634443541835907e-06, + "loss": 1.6282, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.5775389671325684, + "learning_rate": 4.632412672623883e-06, + "loss": 1.651, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 3.026287794113159, + "learning_rate": 4.630381803411861e-06, + "loss": 1.6458, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.8461565971374512, + "eval_runtime": 37.3881, + "eval_samples_per_second": 26.747, + "eval_steps_per_second": 1.123, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.6781764030456543, + "learning_rate": 4.628350934199838e-06, + "loss": 1.5232, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.763350486755371, + "learning_rate": 4.626320064987816e-06, + "loss": 1.6186, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.7063372135162354, + "learning_rate": 4.624289195775792e-06, + "loss": 1.6784, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.824563980102539, + "learning_rate": 4.62225832656377e-06, + "loss": 1.6219, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.5335569381713867, + "learning_rate": 4.620227457351747e-06, + "loss": 1.6152, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.773644208908081, + "learning_rate": 4.618196588139724e-06, + "loss": 1.6179, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.9841721057891846, + "learning_rate": 4.616165718927701e-06, + "loss": 1.6258, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.684171676635742, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.6066, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.8817965984344482, + "learning_rate": 4.612103980503656e-06, + "loss": 1.6697, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.9846866130828857, + "learning_rate": 4.610073111291633e-06, + "loss": 1.6541, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.8485002517700195, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.904170513153076, + "learning_rate": 4.60804224207961e-06, + "loss": 1.6362, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.684476375579834, + "learning_rate": 4.606011372867588e-06, + "loss": 1.6303, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6692216396331787, + "learning_rate": 4.603980503655565e-06, + "loss": 1.638, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.73555588722229, + "learning_rate": 4.601949634443542e-06, + "loss": 1.6125, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.729525089263916, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.565, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.5972018241882324, + "learning_rate": 4.597887896019497e-06, + "loss": 1.5856, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7859137058258057, + "learning_rate": 4.595857026807474e-06, + "loss": 1.6325, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.662587881088257, + "learning_rate": 4.593826157595451e-06, + "loss": 1.6172, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.6380577087402344, + "learning_rate": 4.591795288383429e-06, + "loss": 1.6005, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8375349044799805, + "learning_rate": 4.589764419171406e-06, + "loss": 1.6507, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.833074688911438, + "eval_runtime": 37.0695, + "eval_samples_per_second": 26.976, + "eval_steps_per_second": 1.133, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.786713123321533, + "learning_rate": 4.587733549959383e-06, + "loss": 1.6283, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.6959030628204346, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.5994, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.6100692749023438, + "learning_rate": 4.583671811535338e-06, + "loss": 1.5884, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.7281863689422607, + "learning_rate": 4.581640942323315e-06, + "loss": 1.5737, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.7519376277923584, + "learning_rate": 4.579610073111292e-06, + "loss": 1.6531, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.9314866065979004, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.6755, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.9250576496124268, + "learning_rate": 4.575548334687247e-06, + "loss": 1.6604, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.6382815837860107, + "learning_rate": 4.573517465475224e-06, + "loss": 1.6773, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.6891751289367676, + "learning_rate": 4.571486596263201e-06, + "loss": 1.5849, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.7018136978149414, + "learning_rate": 4.569455727051178e-06, + "loss": 1.6358, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.838019847869873, + "eval_runtime": 37.0496, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.671952247619629, + "learning_rate": 4.567424857839156e-06, + "loss": 1.6207, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.882539749145508, + "learning_rate": 4.565393988627132e-06, + "loss": 1.6618, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.684070587158203, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.6502, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.5295751094818115, + "learning_rate": 4.561332250203087e-06, + "loss": 1.6078, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.0268194675445557, + "learning_rate": 4.559301380991065e-06, + "loss": 1.6312, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6895201206207275, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.5848, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.869950771331787, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.5951, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.909505605697632, + "learning_rate": 4.553208773354996e-06, + "loss": 1.6589, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.7604334354400635, + "learning_rate": 4.551177904142974e-06, + "loss": 1.5767, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.8792080879211426, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.6864, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.836847186088562, + "eval_runtime": 36.9142, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.607037305831909, + "learning_rate": 4.547116165718929e-06, + "loss": 1.5925, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.9338927268981934, + "learning_rate": 4.545085296506905e-06, + "loss": 1.6433, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.7723233699798584, + "learning_rate": 4.543054427294882e-06, + "loss": 1.5963, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.4122188091278076, + "learning_rate": 4.54102355808286e-06, + "loss": 1.5978, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.7582709789276123, + "learning_rate": 4.538992688870837e-06, + "loss": 1.581, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.6983039379119873, + "learning_rate": 4.536961819658814e-06, + "loss": 1.6253, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.6036365032196045, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.5957, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.759645700454712, + "learning_rate": 4.532900081234769e-06, + "loss": 1.6199, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.797744035720825, + "learning_rate": 4.530869212022746e-06, + "loss": 1.5854, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.752868890762329, + "learning_rate": 4.528838342810723e-06, + "loss": 1.619, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.8305871486663818, + "eval_runtime": 37.0307, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.7421889305114746, + "learning_rate": 4.526807473598701e-06, + "loss": 1.6182, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.6852707862854004, + "learning_rate": 4.524776604386678e-06, + "loss": 1.6135, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.861720323562622, + "learning_rate": 4.522745735174655e-06, + "loss": 1.6099, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.706851005554199, + "learning_rate": 4.520714865962632e-06, + "loss": 1.6597, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7992191314697266, + "learning_rate": 4.51868399675061e-06, + "loss": 1.6077, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.6248183250427246, + "learning_rate": 4.516653127538587e-06, + "loss": 1.6485, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.696040391921997, + "learning_rate": 4.514622258326564e-06, + "loss": 1.5853, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.5958261489868164, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.571, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.9715874195098877, + "learning_rate": 4.510560519902519e-06, + "loss": 1.6044, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.664834499359131, + "learning_rate": 4.508529650690496e-06, + "loss": 1.6568, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.8277279138565063, + "eval_runtime": 37.0489, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.869558811187744, + "learning_rate": 4.506498781478473e-06, + "loss": 1.5654, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.5814828872680664, + "learning_rate": 4.504467912266451e-06, + "loss": 1.5484, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.6158947944641113, + "learning_rate": 4.502437043054428e-06, + "loss": 1.6202, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.7424654960632324, + "learning_rate": 4.500406173842405e-06, + "loss": 1.6755, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.8670949935913086, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.6144, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.7914652824401855, + "learning_rate": 4.49634443541836e-06, + "loss": 1.6164, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.866380214691162, + "learning_rate": 4.494313566206336e-06, + "loss": 1.5961, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.966474771499634, + "learning_rate": 4.492282696994314e-06, + "loss": 1.6681, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.751291513442993, + "learning_rate": 4.490251827782291e-06, + "loss": 1.6338, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.58840012550354, + "learning_rate": 4.488220958570269e-06, + "loss": 1.5619, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.824889898300171, + "eval_runtime": 37.0821, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.786945104598999, + "learning_rate": 4.486190089358245e-06, + "loss": 1.5712, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4536845684051514, + "learning_rate": 4.484159220146223e-06, + "loss": 1.5487, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.780278205871582, + "learning_rate": 4.4821283509342e-06, + "loss": 1.6269, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.5427017211914062, + "learning_rate": 4.480097481722178e-06, + "loss": 1.6254, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.7540087699890137, + "learning_rate": 4.478066612510154e-06, + "loss": 1.6112, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.616352081298828, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.5637, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.686035633087158, + "learning_rate": 4.474004874086109e-06, + "loss": 1.5974, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.982365846633911, + "learning_rate": 4.471974004874087e-06, + "loss": 1.7034, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.6106085777282715, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.5482, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.7798519134521484, + "learning_rate": 4.467912266450041e-06, + "loss": 1.6475, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.820489764213562, + "eval_runtime": 37.0672, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.8634185791015625, + "learning_rate": 4.465881397238018e-06, + "loss": 1.6258, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.633626699447632, + "learning_rate": 4.463850528025995e-06, + "loss": 1.5806, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.719074249267578, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.6404, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.7545487880706787, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6404, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.6559669971466064, + "learning_rate": 4.457757920389927e-06, + "loss": 1.6104, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.9004898071289062, + "learning_rate": 4.455727051177904e-06, + "loss": 1.5513, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.6640021800994873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.5859, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.842026472091675, + "learning_rate": 4.451665312753859e-06, + "loss": 1.6241, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.784647226333618, + "learning_rate": 4.449634443541836e-06, + "loss": 1.627, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.7067618370056152, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.6276, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.8115602731704712, + "eval_runtime": 37.078, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.6582882404327393, + "learning_rate": 4.445572705117791e-06, + "loss": 1.5746, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7881102561950684, + "learning_rate": 4.443541835905768e-06, + "loss": 1.5749, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.9323439598083496, + "learning_rate": 4.441510966693745e-06, + "loss": 1.6421, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.63948655128479, + "learning_rate": 4.439480097481723e-06, + "loss": 1.5853, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.7746710777282715, + "learning_rate": 4.4374492282697e-06, + "loss": 1.6455, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.093895435333252, + "learning_rate": 4.435418359057677e-06, + "loss": 1.5311, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.526385545730591, + "learning_rate": 4.433387489845654e-06, + "loss": 1.6082, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.6604297161102295, + "learning_rate": 4.431356620633632e-06, + "loss": 1.61, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.6636247634887695, + "learning_rate": 4.429325751421609e-06, + "loss": 1.5815, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.5567514896392822, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6086, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.8170957565307617, + "eval_runtime": 37.4095, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.695563793182373, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.5836, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6009771823883057, + "learning_rate": 4.423233143785541e-06, + "loss": 1.5264, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.0490736961364746, + "learning_rate": 4.421202274573518e-06, + "loss": 1.5982, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.8609886169433594, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6265, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4415369033813477, + "learning_rate": 4.417140536149473e-06, + "loss": 1.4987, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.771657943725586, + "learning_rate": 4.415109666937449e-06, + "loss": 1.5494, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.739617347717285, + "learning_rate": 4.413078797725427e-06, + "loss": 1.6274, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.805098295211792, + "learning_rate": 4.411047928513404e-06, + "loss": 1.5686, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.7248735427856445, + "learning_rate": 4.409017059301382e-06, + "loss": 1.6027, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6653318405151367, + "learning_rate": 4.406986190089358e-06, + "loss": 1.5909, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.8076283931732178, + "eval_runtime": 37.5073, + "eval_samples_per_second": 26.661, + "eval_steps_per_second": 1.12, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.626783847808838, + "learning_rate": 4.404955320877336e-06, + "loss": 1.6205, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.6929268836975098, + "learning_rate": 4.402924451665313e-06, + "loss": 1.6146, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.6806800365448, + "learning_rate": 4.400893582453291e-06, + "loss": 1.5702, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.5407259464263916, + "learning_rate": 4.398862713241267e-06, + "loss": 1.5525, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.639122247695923, + "learning_rate": 4.396831844029245e-06, + "loss": 1.5826, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.669396162033081, + "learning_rate": 4.394800974817222e-06, + "loss": 1.5982, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.6377906799316406, + "learning_rate": 4.392770105605199e-06, + "loss": 1.5797, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6769638061523438, + "learning_rate": 4.390739236393176e-06, + "loss": 1.607, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.6347222328186035, + "learning_rate": 4.388708367181154e-06, + "loss": 1.6196, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.670680284500122, + "learning_rate": 4.386677497969131e-06, + "loss": 1.619, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.8124384880065918, + "eval_runtime": 37.4751, + "eval_samples_per_second": 26.684, + "eval_steps_per_second": 1.121, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.5886521339416504, + "learning_rate": 4.384646628757108e-06, + "loss": 1.6258, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.644566297531128, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6447, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.6432394981384277, + "learning_rate": 4.380584890333063e-06, + "loss": 1.5888, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.5602784156799316, + "learning_rate": 4.37855402112104e-06, + "loss": 1.5544, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.4874439239501953, + "learning_rate": 4.376523151909017e-06, + "loss": 1.5075, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.694748640060425, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.6498, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.582211494445801, + "learning_rate": 4.372461413484972e-06, + "loss": 1.5877, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.675713539123535, + "learning_rate": 4.370430544272949e-06, + "loss": 1.6045, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.7157161235809326, + "learning_rate": 4.368399675060926e-06, + "loss": 1.623, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.795076370239258, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6141, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.8062744140625, + "eval_runtime": 37.2435, + "eval_samples_per_second": 26.85, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6296374797821045, + "learning_rate": 4.364337936636881e-06, + "loss": 1.6087, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.362307067424858e-06, + "loss": 1.5182, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.7965915203094482, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.6025, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.807307481765747, + "learning_rate": 4.358245329000813e-06, + "loss": 1.574, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.584615468978882, + "learning_rate": 4.35621445978879e-06, + "loss": 1.5922, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.8472282886505127, + "learning_rate": 4.354183590576767e-06, + "loss": 1.5914, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.8471736907958984, + "learning_rate": 4.352152721364745e-06, + "loss": 1.5749, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.569225549697876, + "learning_rate": 4.350121852152722e-06, + "loss": 1.5591, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.688157796859741, + "learning_rate": 4.348090982940699e-06, + "loss": 1.6042, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.5865259170532227, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.6307, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.8040707111358643, + "eval_runtime": 36.9511, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.71850323677063, + "learning_rate": 4.344029244516654e-06, + "loss": 1.6043, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.7158093452453613, + "learning_rate": 4.341998375304631e-06, + "loss": 1.5907, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6879327297210693, + "learning_rate": 4.339967506092607e-06, + "loss": 1.6365, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.827277660369873, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.5928, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6709842681884766, + "learning_rate": 4.335905767668562e-06, + "loss": 1.5812, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.7071495056152344, + "learning_rate": 4.33387489845654e-06, + "loss": 1.5571, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6008732318878174, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.6048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.6238059997558594, + "learning_rate": 4.329813160032495e-06, + "loss": 1.5962, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.567065954208374, + "learning_rate": 4.327782290820471e-06, + "loss": 1.5713, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.573068857192993, + "learning_rate": 4.325751421608449e-06, + "loss": 1.536, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.8103471994400024, + "eval_runtime": 37.0956, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6245224475860596, + "learning_rate": 4.323720552396426e-06, + "loss": 1.5986, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.6243810653686523, + "learning_rate": 4.321689683184404e-06, + "loss": 1.589, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.5303313732147217, + "learning_rate": 4.31965881397238e-06, + "loss": 1.5793, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.5617806911468506, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.5704, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.7429656982421875, + "learning_rate": 4.315597075548335e-06, + "loss": 1.6174, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.6753251552581787, + "learning_rate": 4.313566206336312e-06, + "loss": 1.5879, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.5944488048553467, + "learning_rate": 4.311535337124289e-06, + "loss": 1.6082, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.9002904891967773, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.5521, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.5384247303009033, + "learning_rate": 4.307473598700244e-06, + "loss": 1.5751, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6015477180480957, + "learning_rate": 4.305442729488221e-06, + "loss": 1.5966, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8091154098510742, + "eval_runtime": 37.1075, + "eval_samples_per_second": 26.949, + "eval_steps_per_second": 1.132, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.6002886295318604, + "learning_rate": 4.303411860276198e-06, + "loss": 1.5403, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7755627632141113, + "learning_rate": 4.301380991064176e-06, + "loss": 1.6192, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.8351540565490723, + "learning_rate": 4.299350121852153e-06, + "loss": 1.6217, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.684007406234741, + "learning_rate": 4.29731925264013e-06, + "loss": 1.6242, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.8019614219665527, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6173, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.9122867584228516, + "learning_rate": 4.293257514216085e-06, + "loss": 1.5701, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.545732021331787, + "learning_rate": 4.291226645004062e-06, + "loss": 1.5494, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.9499566555023193, + "learning_rate": 4.289195775792039e-06, + "loss": 1.5848, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.8071892261505127, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6114, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.763864517211914, + "learning_rate": 4.285134037367994e-06, + "loss": 1.5768, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8057912588119507, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 3.001190423965454, + "learning_rate": 4.283103168155971e-06, + "loss": 1.5601, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.7640297412872314, + "learning_rate": 4.281072298943948e-06, + "loss": 1.5619, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.7372794151306152, + "learning_rate": 4.279041429731926e-06, + "loss": 1.562, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 3.0390751361846924, + "learning_rate": 4.277010560519903e-06, + "loss": 1.5519, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.758798599243164, + "learning_rate": 4.27497969130788e-06, + "loss": 1.527, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.96443772315979, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.5899, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.7941088676452637, + "learning_rate": 4.270917952883835e-06, + "loss": 1.617, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 3.0385100841522217, + "learning_rate": 4.268887083671812e-06, + "loss": 1.5996, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.6946637630462646, + "learning_rate": 4.266856214459789e-06, + "loss": 1.5641, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.776723861694336, + "learning_rate": 4.264825345247766e-06, + "loss": 1.6451, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.7911298274993896, + "eval_runtime": 37.0623, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.851224184036255, + "learning_rate": 4.262794476035744e-06, + "loss": 1.5302, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9045867919921875, + "learning_rate": 4.26076360682372e-06, + "loss": 1.5808, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.6568782329559326, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.5233, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.94191837310791, + "learning_rate": 4.256701868399675e-06, + "loss": 1.5912, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.726778984069824, + "learning_rate": 4.254670999187653e-06, + "loss": 1.5823, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.8704583644866943, + "learning_rate": 4.252640129975629e-06, + "loss": 1.5931, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.5721547603607178, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6128, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6883509159088135, + "learning_rate": 4.248578391551584e-06, + "loss": 1.5479, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.7186272144317627, + "learning_rate": 4.246547522339562e-06, + "loss": 1.5203, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.8326518535614014, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.5547, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.7948570251464844, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.7870523929595947, + "learning_rate": 4.242485783915517e-06, + "loss": 1.5891, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.866610288619995, + "learning_rate": 4.240454914703493e-06, + "loss": 1.5232, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.94161057472229, + "learning_rate": 4.23842404549147e-06, + "loss": 1.52, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.7331979274749756, + "learning_rate": 4.236393176279448e-06, + "loss": 1.5665, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.5375845432281494, + "learning_rate": 4.234362307067425e-06, + "loss": 1.5274, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.691927433013916, + "learning_rate": 4.232331437855402e-06, + "loss": 1.604, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7578394412994385, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.6, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.594538688659668, + "learning_rate": 4.228269699431357e-06, + "loss": 1.5928, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.5408968925476074, + "learning_rate": 4.226238830219334e-06, + "loss": 1.5908, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.689786672592163, + "learning_rate": 4.224207961007311e-06, + "loss": 1.5757, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.792494297027588, + "eval_runtime": 37.0822, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.735539197921753, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.5811, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.58634614944458, + "learning_rate": 4.220146222583266e-06, + "loss": 1.5808, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.7899138927459717, + "learning_rate": 4.218115353371243e-06, + "loss": 1.5801, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.6507744789123535, + "learning_rate": 4.21608448415922e-06, + "loss": 1.5585, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.6112217903137207, + "learning_rate": 4.214053614947198e-06, + "loss": 1.5329, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.729405164718628, + "learning_rate": 4.212022745735175e-06, + "loss": 1.6177, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.6431992053985596, + "learning_rate": 4.209991876523152e-06, + "loss": 1.5234, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.726243734359741, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.5926, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.6574180126190186, + "learning_rate": 4.205930138099107e-06, + "loss": 1.5424, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.6570956707000732, + "learning_rate": 4.203899268887084e-06, + "loss": 1.6119, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.781999945640564, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.528052806854248, + "learning_rate": 4.201868399675061e-06, + "loss": 1.598, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.6075727939605713, + "learning_rate": 4.199837530463039e-06, + "loss": 1.5886, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.588737964630127, + "learning_rate": 4.197806661251016e-06, + "loss": 1.507, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.773618221282959, + "learning_rate": 4.195775792038993e-06, + "loss": 1.5794, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.624725580215454, + "learning_rate": 4.19374492282697e-06, + "loss": 1.5495, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.662397861480713, + "learning_rate": 4.191714053614948e-06, + "loss": 1.5203, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.6617255210876465, + "learning_rate": 4.189683184402924e-06, + "loss": 1.5885, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.5784451961517334, + "learning_rate": 4.187652315190902e-06, + "loss": 1.5689, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.726011037826538, + "learning_rate": 4.185621445978879e-06, + "loss": 1.5913, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.6117324829101562, + "learning_rate": 4.183590576766857e-06, + "loss": 1.5453, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.7893352508544922, + "eval_runtime": 37.4096, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7025442123413086, + "learning_rate": 4.181559707554833e-06, + "loss": 1.5868, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.656102180480957, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6141, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.741962194442749, + "learning_rate": 4.177497969130788e-06, + "loss": 1.5516, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5887229442596436, + "learning_rate": 4.175467099918766e-06, + "loss": 1.5525, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.542985439300537, + "learning_rate": 4.173436230706742e-06, + "loss": 1.592, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.7797129154205322, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.5349, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.632887363433838, + "learning_rate": 4.169374492282697e-06, + "loss": 1.5548, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.691690444946289, + "learning_rate": 4.167343623070675e-06, + "loss": 1.5352, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.8922817707061768, + "learning_rate": 4.165312753858652e-06, + "loss": 1.5717, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.5517446994781494, + "learning_rate": 4.163281884646629e-06, + "loss": 1.5658, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.780922770500183, + "eval_runtime": 37.3755, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.718374490737915, + "learning_rate": 4.161251015434607e-06, + "loss": 1.5578, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.6249442100524902, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6018, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.7051141262054443, + "learning_rate": 4.157189277010561e-06, + "loss": 1.5182, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.677109956741333, + "learning_rate": 4.155158407798538e-06, + "loss": 1.5423, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.591059684753418, + "learning_rate": 4.153127538586516e-06, + "loss": 1.5694, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.699953556060791, + "learning_rate": 4.151096669374492e-06, + "loss": 1.5758, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.8837342262268066, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.5509, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.6566851139068604, + "learning_rate": 4.147034930950447e-06, + "loss": 1.5577, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.901057004928589, + "learning_rate": 4.145004061738425e-06, + "loss": 1.5797, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.796971321105957, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.5608, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.7797813415527344, + "eval_runtime": 36.8425, + "eval_samples_per_second": 27.143, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.605431318283081, + "learning_rate": 4.140942323314379e-06, + "loss": 1.491, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8084020614624023, + "learning_rate": 4.138911454102356e-06, + "loss": 1.5074, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.625511407852173, + "learning_rate": 4.136880584890333e-06, + "loss": 1.5404, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.6944286823272705, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.589, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.81792950630188, + "learning_rate": 4.132818846466288e-06, + "loss": 1.5646, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5813655853271484, + "learning_rate": 4.130787977254265e-06, + "loss": 1.5697, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.6733286380767822, + "learning_rate": 4.128757108042242e-06, + "loss": 1.5786, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.6972577571868896, + "learning_rate": 4.12672623883022e-06, + "loss": 1.5857, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7381093502044678, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6365, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.6162049770355225, + "learning_rate": 4.122664500406174e-06, + "loss": 1.5598, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.7869179248809814, + "eval_runtime": 36.974, + "eval_samples_per_second": 27.046, + "eval_steps_per_second": 1.136, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.7157270908355713, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.5533, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.560105085372925, + "learning_rate": 4.118602761982129e-06, + "loss": 1.5821, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.5261518955230713, + "learning_rate": 4.116571892770106e-06, + "loss": 1.5195, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7377054691314697, + "learning_rate": 4.114541023558083e-06, + "loss": 1.5066, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.632946252822876, + "learning_rate": 4.112510154346061e-06, + "loss": 1.5948, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.809671640396118, + "learning_rate": 4.110479285134038e-06, + "loss": 1.5435, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.843156099319458, + "learning_rate": 4.108448415922015e-06, + "loss": 1.5667, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.743614673614502, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5958, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.9596827030181885, + "learning_rate": 4.10438667749797e-06, + "loss": 1.5807, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.7321834564208984, + "learning_rate": 4.102355808285947e-06, + "loss": 1.4744, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.784859538078308, + "eval_runtime": 37.0146, + "eval_samples_per_second": 27.016, + "eval_steps_per_second": 1.135, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.010359764099121, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6146, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.744251251220703, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.5401, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.592690944671631, + "learning_rate": 4.096263200649879e-06, + "loss": 1.5508, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.786736011505127, + "learning_rate": 4.094232331437856e-06, + "loss": 1.5717, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.8773610591888428, + "learning_rate": 4.092201462225833e-06, + "loss": 1.5201, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.818610906600952, + "learning_rate": 4.090170593013811e-06, + "loss": 1.5282, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.9548542499542236, + "learning_rate": 4.088139723801787e-06, + "loss": 1.553, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.678842306137085, + "learning_rate": 4.086108854589765e-06, + "loss": 1.531, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.7757408618927, + "learning_rate": 4.084077985377742e-06, + "loss": 1.5371, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.6954638957977295, + "learning_rate": 4.08204711616572e-06, + "loss": 1.5462, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.781003475189209, + "eval_runtime": 37.0687, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6897518634796143, + "learning_rate": 4.080016246953696e-06, + "loss": 1.5979, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.581172466278076, + "learning_rate": 4.077985377741674e-06, + "loss": 1.5007, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.828234910964966, + "learning_rate": 4.075954508529651e-06, + "loss": 1.5857, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.6171951293945312, + "learning_rate": 4.073923639317629e-06, + "loss": 1.5937, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.6476316452026367, + "learning_rate": 4.071892770105605e-06, + "loss": 1.5864, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.7525634765625, + "learning_rate": 4.069861900893583e-06, + "loss": 1.5095, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.6881322860717773, + "learning_rate": 4.06783103168156e-06, + "loss": 1.4936, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.6104350090026855, + "learning_rate": 4.065800162469538e-06, + "loss": 1.5229, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.866706132888794, + "learning_rate": 4.063769293257514e-06, + "loss": 1.539, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.803999900817871, + "learning_rate": 4.061738424045492e-06, + "loss": 1.5792, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.7793465852737427, + "eval_runtime": 37.0749, + "eval_samples_per_second": 26.972, + "eval_steps_per_second": 1.133, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.670285940170288, + "learning_rate": 4.059707554833469e-06, + "loss": 1.5523, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.871267557144165, + "learning_rate": 4.057676685621446e-06, + "loss": 1.5623, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.6631205081939697, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6094, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.7580034732818604, + "learning_rate": 4.053614947197401e-06, + "loss": 1.5725, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.6412739753723145, + "learning_rate": 4.051584077985378e-06, + "loss": 1.4959, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.699176073074341, + "learning_rate": 4.049553208773355e-06, + "loss": 1.5586, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.622969150543213, + "learning_rate": 4.047522339561333e-06, + "loss": 1.5713, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.892338991165161, + "learning_rate": 4.04549147034931e-06, + "loss": 1.5471, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.77056622505188, + "learning_rate": 4.043460601137287e-06, + "loss": 1.5236, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.7122550010681152, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6562, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.7708193063735962, + "eval_runtime": 37.0743, + "eval_samples_per_second": 26.973, + "eval_steps_per_second": 1.133, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.702824592590332, + "learning_rate": 4.039398862713242e-06, + "loss": 1.534, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.6822307109832764, + "learning_rate": 4.037367993501219e-06, + "loss": 1.5368, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.7274515628814697, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6218, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.8114798069000244, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.532, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.8936192989349365, + "learning_rate": 4.031275385865151e-06, + "loss": 1.5124, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.7851672172546387, + "learning_rate": 4.029244516653128e-06, + "loss": 1.5596, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.815023183822632, + "learning_rate": 4.027213647441105e-06, + "loss": 1.5706, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5958921909332275, + "learning_rate": 4.025182778229083e-06, + "loss": 1.5161, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.779360771179199, + "learning_rate": 4.02315190901706e-06, + "loss": 1.5057, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.7411787509918213, + "learning_rate": 4.021121039805037e-06, + "loss": 1.5396, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.7778708934783936, + "eval_runtime": 37.0791, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.5709781646728516, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.5474, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8583388328552246, + "learning_rate": 4.017059301380992e-06, + "loss": 1.6149, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.6877572536468506, + "learning_rate": 4.015028432168969e-06, + "loss": 1.5292, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.680612087249756, + "learning_rate": 4.012997562956945e-06, + "loss": 1.596, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.653559446334839, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.4901, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.542787790298462, + "learning_rate": 4.0089358245329e-06, + "loss": 1.5295, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5476601123809814, + "learning_rate": 4.006904955320878e-06, + "loss": 1.5393, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.6367862224578857, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.5321, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.8112173080444336, + "learning_rate": 4.002843216896833e-06, + "loss": 1.5963, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.7136871814727783, + "learning_rate": 4.000812347684809e-06, + "loss": 1.5839, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.7763899564743042, + "eval_runtime": 37.2537, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.9095280170440674, + "learning_rate": 3.998781478472787e-06, + "loss": 1.5343, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.6196115016937256, + "learning_rate": 3.996750609260764e-06, + "loss": 1.5411, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6413371562957764, + "learning_rate": 3.994719740048742e-06, + "loss": 1.4862, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 3.0759036540985107, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6471, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.599147081375122, + "learning_rate": 3.990658001624696e-06, + "loss": 1.5556, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.726978302001953, + "learning_rate": 3.988627132412673e-06, + "loss": 1.5351, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.6760807037353516, + "learning_rate": 3.98659626320065e-06, + "loss": 1.5681, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.680483818054199, + "learning_rate": 3.984565393988627e-06, + "loss": 1.5225, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6649010181427, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.5325, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.638085126876831, + "learning_rate": 3.980503655564582e-06, + "loss": 1.564, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.7698904275894165, + "eval_runtime": 37.4492, + "eval_samples_per_second": 26.703, + "eval_steps_per_second": 1.122, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.647825241088867, + "learning_rate": 3.978472786352559e-06, + "loss": 1.5782, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.748222589492798, + "learning_rate": 3.976441917140536e-06, + "loss": 1.556, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.6319026947021484, + "learning_rate": 3.974411047928514e-06, + "loss": 1.5378, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.6759872436523438, + "learning_rate": 3.972380178716491e-06, + "loss": 1.6018, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.6760940551757812, + "learning_rate": 3.970349309504468e-06, + "loss": 1.5742, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.5565083026885986, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.565, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.5986931324005127, + "learning_rate": 3.966287571080423e-06, + "loss": 1.5124, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.713092803955078, + "learning_rate": 3.9642567018684e-06, + "loss": 1.523, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.6203417778015137, + "learning_rate": 3.962225832656377e-06, + "loss": 1.5601, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.604440927505493, + "learning_rate": 3.960194963444355e-06, + "loss": 1.5485, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.7633332014083862, + "eval_runtime": 37.49, + "eval_samples_per_second": 26.674, + "eval_steps_per_second": 1.12, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.5513081550598145, + "learning_rate": 3.958164094232332e-06, + "loss": 1.479, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.7990963459014893, + "learning_rate": 3.956133225020309e-06, + "loss": 1.5355, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.651113271713257, + "learning_rate": 3.954102355808286e-06, + "loss": 1.5406, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.7236671447753906, + "learning_rate": 3.952071486596264e-06, + "loss": 1.5314, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.697598695755005, + "learning_rate": 3.950040617384241e-06, + "loss": 1.5398, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.8271944522857666, + "learning_rate": 3.948009748172218e-06, + "loss": 1.54, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.733125925064087, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.5236, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.7555017471313477, + "learning_rate": 3.943948009748173e-06, + "loss": 1.5808, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.6592960357666016, + "learning_rate": 3.94191714053615e-06, + "loss": 1.5398, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.6870126724243164, + "learning_rate": 3.939886271324127e-06, + "loss": 1.5517, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.760675311088562, + "eval_runtime": 36.9372, + "eval_samples_per_second": 27.073, + "eval_steps_per_second": 1.137, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.7469022274017334, + "learning_rate": 3.937855402112104e-06, + "loss": 1.605, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.5255792140960693, + "learning_rate": 3.935824532900082e-06, + "loss": 1.5061, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.589054822921753, + "learning_rate": 3.933793663688058e-06, + "loss": 1.4971, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.7874889373779297, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.5643, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.819451093673706, + "learning_rate": 3.929731925264013e-06, + "loss": 1.5797, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.6516339778900146, + "learning_rate": 3.927701056051991e-06, + "loss": 1.5048, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.979091167449951, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.5174, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.743058204650879, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.5432, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.6864030361175537, + "learning_rate": 3.921608448415922e-06, + "loss": 1.5147, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.7751128673553467, + "learning_rate": 3.9195775792039e-06, + "loss": 1.4857, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.7609487771987915, + "eval_runtime": 37.09, + "eval_samples_per_second": 26.961, + "eval_steps_per_second": 1.132, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.626188278198242, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.5428, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.554067850112915, + "learning_rate": 3.915515840779855e-06, + "loss": 1.5153, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.8580875396728516, + "learning_rate": 3.913484971567831e-06, + "loss": 1.5562, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.7405097484588623, + "learning_rate": 3.911454102355808e-06, + "loss": 1.565, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.5650815963745117, + "learning_rate": 3.909423233143786e-06, + "loss": 1.4756, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.9405910968780518, + "learning_rate": 3.907392363931763e-06, + "loss": 1.5611, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.687448024749756, + "learning_rate": 3.90536149471974e-06, + "loss": 1.5405, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.7373857498168945, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.6055, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.875298261642456, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6215, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.713200330734253, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6493, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.7608598470687866, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.541935682296753, + "learning_rate": 3.897238017871649e-06, + "loss": 1.5257, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.582045316696167, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.5253, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.6884682178497314, + "learning_rate": 3.893176279447604e-06, + "loss": 1.5329, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.6456501483917236, + "learning_rate": 3.891145410235581e-06, + "loss": 1.5167, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.60788893699646, + "learning_rate": 3.889114541023558e-06, + "loss": 1.5945, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.708646774291992, + "learning_rate": 3.887083671811536e-06, + "loss": 1.5704, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.666922092437744, + "learning_rate": 3.885052802599513e-06, + "loss": 1.52, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.6623599529266357, + "learning_rate": 3.88302193338749e-06, + "loss": 1.4791, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.9031424522399902, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.5954, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.763303279876709, + "learning_rate": 3.878960194963445e-06, + "loss": 1.5618, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.7647819519042969, + "eval_runtime": 37.1392, + "eval_samples_per_second": 26.926, + "eval_steps_per_second": 1.131, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 3.0733320713043213, + "learning_rate": 3.876929325751422e-06, + "loss": 1.5191, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.7809090614318848, + "learning_rate": 3.874898456539399e-06, + "loss": 1.5804, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.6357381343841553, + "learning_rate": 3.872867587327377e-06, + "loss": 1.4843, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.680595874786377, + "learning_rate": 3.870836718115354e-06, + "loss": 1.5339, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.5908069610595703, + "learning_rate": 3.868805848903331e-06, + "loss": 1.5049, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.680586338043213, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.5188, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.763528823852539, + "learning_rate": 3.864744110479286e-06, + "loss": 1.5395, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.8522050380706787, + "learning_rate": 3.862713241267262e-06, + "loss": 1.562, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.806962490081787, + "learning_rate": 3.86068237205524e-06, + "loss": 1.5419, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.9716053009033203, + "learning_rate": 3.858651502843217e-06, + "loss": 1.5691, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.7608845233917236, + "eval_runtime": 37.1433, + "eval_samples_per_second": 26.923, + "eval_steps_per_second": 1.131, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.613098382949829, + "learning_rate": 3.856620633631195e-06, + "loss": 1.5297, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.5690557956695557, + "learning_rate": 3.854589764419171e-06, + "loss": 1.4964, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.8001089096069336, + "learning_rate": 3.852558895207149e-06, + "loss": 1.5253, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.6715309619903564, + "learning_rate": 3.850528025995126e-06, + "loss": 1.601, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.7044479846954346, + "learning_rate": 3.848497156783104e-06, + "loss": 1.5526, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.7368102073669434, + "learning_rate": 3.84646628757108e-06, + "loss": 1.5088, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.9358696937561035, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.5202, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.858172655105591, + "learning_rate": 3.842404549147035e-06, + "loss": 1.5378, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.6723833084106445, + "learning_rate": 3.840373679935013e-06, + "loss": 1.5134, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.748187780380249, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.4797, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.75319242477417, + "eval_runtime": 37.1302, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.4821436405181885, + "learning_rate": 3.836311941510967e-06, + "loss": 1.5382, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.5585227012634277, + "learning_rate": 3.834281072298944e-06, + "loss": 1.5373, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.927334785461426, + "learning_rate": 3.832250203086921e-06, + "loss": 1.5447, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.6096436977386475, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.5062, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.677304744720459, + "learning_rate": 3.828188464662876e-06, + "loss": 1.5671, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.9606077671051025, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6138, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.6971657276153564, + "learning_rate": 3.82412672623883e-06, + "loss": 1.5544, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.6380839347839355, + "learning_rate": 3.822095857026808e-06, + "loss": 1.4964, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.7537858486175537, + "learning_rate": 3.820064987814785e-06, + "loss": 1.5429, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.6189136505126953, + "learning_rate": 3.818034118602762e-06, + "loss": 1.4956, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.7546342611312866, + "eval_runtime": 37.0828, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5744292736053467, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.5256, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.5503833293914795, + "learning_rate": 3.813972380178717e-06, + "loss": 1.5211, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.979384422302246, + "learning_rate": 3.811941510966694e-06, + "loss": 1.55, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.867161512374878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5655, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.8006832599639893, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.4902, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.594712734222412, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.4451, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.6632349491119385, + "learning_rate": 3.803818034118603e-06, + "loss": 1.531, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.859894275665283, + "learning_rate": 3.80178716490658e-06, + "loss": 1.5018, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.6405653953552246, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.5765, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.7285192012786865, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.5537, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.7618242502212524, + "eval_runtime": 37.0898, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.6882379055023193, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.5143, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.706550121307373, + "learning_rate": 3.793663688058489e-06, + "loss": 1.4857, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.790907382965088, + "learning_rate": 3.791632818846467e-06, + "loss": 1.5982, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.631096363067627, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.5656, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.6981775760650635, + "learning_rate": 3.787571080422421e-06, + "loss": 1.567, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.612086296081543, + "learning_rate": 3.785540211210398e-06, + "loss": 1.5723, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.718630075454712, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.5565, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.6265041828155518, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.5326, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.578124523162842, + "learning_rate": 3.77944760357433e-06, + "loss": 1.533, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.832695484161377, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6133, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.7516050338745117, + "eval_runtime": 37.3173, + "eval_samples_per_second": 26.797, + "eval_steps_per_second": 1.125, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.851625442504883, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.5836, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.773354995938262e-06, + "loss": 1.5069, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.5738255977630615, + "learning_rate": 3.771324126726239e-06, + "loss": 1.5531, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.583575963973999, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.5589, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.7729389667510986, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.5901, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.6322245597839355, + "learning_rate": 3.765231519090171e-06, + "loss": 1.5423, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.6382699012756348, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.4574, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.7393112182617188, + "learning_rate": 3.761169780666125e-06, + "loss": 1.5351, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.684398651123047, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.4873, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.89401912689209, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.5358, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.7523635625839233, + "eval_runtime": 37.3763, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.629870891571045, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.5566, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.6936254501342773, + "learning_rate": 3.753046303818034e-06, + "loss": 1.5016, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.5889103412628174, + "learning_rate": 3.751015434606012e-06, + "loss": 1.5775, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.610703468322754, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.5508, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6318750381469727, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.5892, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.6650729179382324, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.5106, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.577174186706543, + "learning_rate": 3.742891957757921e-06, + "loss": 1.54, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.7274489402770996, + "learning_rate": 3.740861088545898e-06, + "loss": 1.5803, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6849985122680664, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.5393, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.811479091644287, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.5862, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.7490049600601196, + "eval_runtime": 36.9083, + "eval_samples_per_second": 27.094, + "eval_steps_per_second": 1.138, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.640002489089966, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.5149, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.6350319385528564, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5132, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.689814567565918, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.473, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.7181100845336914, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.4353, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.760629415512085, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.4859, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.9322562217712402, + "learning_rate": 3.724614134849716e-06, + "loss": 1.4733, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.280315399169922, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.5769, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 3.1343884468078613, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.4766, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.8537609577178955, + "learning_rate": 3.718521527213648e-06, + "loss": 1.5586, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.895299196243286, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5232, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.7473852634429932, + "eval_runtime": 37.006, + "eval_samples_per_second": 27.023, + "eval_steps_per_second": 1.135, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.82454776763916, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5057, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.711296558380127, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.4187, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.6281023025512695, + "learning_rate": 3.710398050365557e-06, + "loss": 1.4997, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.775632381439209, + "learning_rate": 3.708367181153534e-06, + "loss": 1.5087, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.8021788597106934, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.474, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 3.09647536277771, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5015, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.810450315475464, + "learning_rate": 3.702274573517466e-06, + "loss": 1.4681, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.65166974067688, + "learning_rate": 3.700243704305443e-06, + "loss": 1.4484, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.755690574645996, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5084, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.8223068714141846, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.5015, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.7516940832138062, + "eval_runtime": 37.0227, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.8401107788085938, + "learning_rate": 3.694151096669375e-06, + "loss": 1.4597, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 3.010011672973633, + "learning_rate": 3.692120227457352e-06, + "loss": 1.5625, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8575570583343506, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.4828, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.8527276515960693, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5074, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.81260085105896, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.4132, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.988429069519043, + "learning_rate": 3.683996750609261e-06, + "loss": 1.4803, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.964179515838623, + "learning_rate": 3.681965881397238e-06, + "loss": 1.4871, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.7151682376861572, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.5061, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.764187812805176, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.5188, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.810554265975952, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.4779, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.7422841787338257, + "eval_runtime": 37.0641, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.7640810012817383, + "learning_rate": 3.673842404549147e-06, + "loss": 1.4827, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.7683937549591064, + "learning_rate": 3.671811535337125e-06, + "loss": 1.4775, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.86434268951416, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.4752, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 3.0976617336273193, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.5099, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 3.107388496398926, + "learning_rate": 3.665718927701056e-06, + "loss": 1.5646, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 3.032921314239502, + "learning_rate": 3.663688058489034e-06, + "loss": 1.5048, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.9377830028533936, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5333, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.706542730331421, + "learning_rate": 3.659626320064988e-06, + "loss": 1.4903, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.819936513900757, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.4441, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.805360794067383, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.5122, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.747745156288147, + "eval_runtime": 37.0801, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.848580837249756, + "learning_rate": 3.65353371242892e-06, + "loss": 1.4984, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.772721290588379, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5026, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.760772943496704, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.4005, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.937507390975952, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.5131, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.923210859298706, + "learning_rate": 3.645410235580829e-06, + "loss": 1.4661, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8029167652130127, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.518, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.752716541290283, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.5109, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.7767865657806396, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.4831, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8483617305755615, + "learning_rate": 3.637286758732738e-06, + "loss": 1.4872, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.8714234828948975, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.5109, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.7455583810806274, + "eval_runtime": 37.0892, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7373011112213135, + "learning_rate": 3.633225020308692e-06, + "loss": 1.55, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.942373752593994, + "learning_rate": 3.63119415109667e-06, + "loss": 1.5079, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.8790931701660156, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.4669, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.9564976692199707, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5555, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.811699390411377, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.4457, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.926544427871704, + "learning_rate": 3.623070674248579e-06, + "loss": 1.4368, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.8694913387298584, + "learning_rate": 3.621039805036556e-06, + "loss": 1.4875, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 3.0294361114501953, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.4815, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8643198013305664, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.5031, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 3.0244719982147217, + "learning_rate": 3.614947197400488e-06, + "loss": 1.4914, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.7504756450653076, + "eval_runtime": 37.0238, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.8641672134399414, + "learning_rate": 3.612916328188465e-06, + "loss": 1.5122, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.8713624477386475, + "learning_rate": 3.610885458976442e-06, + "loss": 1.43, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.990912914276123, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.5124, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.963945150375366, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.533, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.802168846130371, + "learning_rate": 3.604792851340374e-06, + "loss": 1.4531, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 3.0556633472442627, + "learning_rate": 3.602761982128351e-06, + "loss": 1.4618, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.95744252204895, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5356, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.813145399093628, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.5134, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.8382728099823, + "learning_rate": 3.596669374492283e-06, + "loss": 1.4413, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.821100950241089, + "learning_rate": 3.59463850528026e-06, + "loss": 1.4748, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.7488586902618408, + "eval_runtime": 37.0361, + "eval_samples_per_second": 27.001, + "eval_steps_per_second": 1.134, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.930708408355713, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.5177, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.9312615394592285, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.4493, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 3.0051841735839844, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.5208, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.793515205383301, + "learning_rate": 3.586515028432169e-06, + "loss": 1.4813, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.7595956325531006, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5379, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.992871046066284, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.5153, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.7270801067352295, + "learning_rate": 3.580422420796101e-06, + "loss": 1.4392, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.797985076904297, + "learning_rate": 3.578391551584078e-06, + "loss": 1.4728, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9595260620117188, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.5158, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.8131442070007324, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.4214, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.7503563165664673, + "eval_runtime": 37.3831, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.124, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8242709636688232, + "learning_rate": 3.57229894394801e-06, + "loss": 1.5027, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.8973116874694824, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.5294, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 3.1207027435302734, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.4902, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.803910732269287, + "learning_rate": 3.566206336311942e-06, + "loss": 1.466, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.813541889190674, + "learning_rate": 3.564175467099919e-06, + "loss": 1.4957, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.9068171977996826, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.4435, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.8491995334625244, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.4432, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.8785598278045654, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.4791, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7701117992401123, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.4607, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.4821, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.7518969774246216, + "eval_runtime": 37.4808, + "eval_samples_per_second": 26.68, + "eval_steps_per_second": 1.121, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.5228, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.8513107299804688, + "learning_rate": 3.551990251827783e-06, + "loss": 1.4812, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.937448024749756, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.4538, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.8307411670684814, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.4807, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.7986767292022705, + "learning_rate": 3.545897644191714e-06, + "loss": 1.4706, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 3.1135106086730957, + "learning_rate": 3.543866774979692e-06, + "loss": 1.4574, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.7756283283233643, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5114, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.8114547729492188, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.4458, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 3.0544545650482178, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.4934, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8990671634674072, + "learning_rate": 3.535743298131601e-06, + "loss": 1.4808, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.7468736171722412, + "eval_runtime": 36.9135, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.976139545440674, + "learning_rate": 3.533712428919578e-06, + "loss": 1.5123, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.909146785736084, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.5204, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7225229740142822, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.4746, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.6145517826080322, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.5049, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.8660519123077393, + "learning_rate": 3.525588952071487e-06, + "loss": 1.5173, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.979801654815674, + "learning_rate": 3.523558082859464e-06, + "loss": 1.5146, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.7500879764556885, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.514, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.6923978328704834, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.455, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.6728951930999756, + "learning_rate": 3.517465475223396e-06, + "loss": 1.4712, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.769507884979248, + "learning_rate": 3.515434606011373e-06, + "loss": 1.444, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.7491599321365356, + "eval_runtime": 37.066, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.9768483638763428, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.5005, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.993196725845337, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.5347, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.886883497238159, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.4959, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.9152941703796387, + "learning_rate": 3.507311129163282e-06, + "loss": 1.4641, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.966872215270996, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.4232, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.9853367805480957, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.4883, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.9353625774383545, + "learning_rate": 3.501218521527214e-06, + "loss": 1.4826, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.7860453128814697, + "learning_rate": 3.499187652315191e-06, + "loss": 1.42, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.9269113540649414, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.5087, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 3.052962303161621, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5042, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.7421623468399048, + "eval_runtime": 37.103, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.7328579425811768, + "learning_rate": 3.493095044679123e-06, + "loss": 1.4375, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.876138925552368, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5559, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.7500345706939697, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.4043, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.836596965789795, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.4808, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.797060966491699, + "learning_rate": 3.484971567831032e-06, + "loss": 1.405, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.928368091583252, + "learning_rate": 3.482940698619009e-06, + "loss": 1.4765, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.923375368118286, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.5134, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.8624660968780518, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.4881, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.9593892097473145, + "learning_rate": 3.476848090982941e-06, + "loss": 1.4869, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.8019728660583496, + "learning_rate": 3.474817221770918e-06, + "loss": 1.468, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.7386831045150757, + "eval_runtime": 37.1034, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.8965771198272705, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.5604, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.8114798069000244, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.5092, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.9554526805877686, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.4869, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.981499195098877, + "learning_rate": 3.466693744922827e-06, + "loss": 1.4454, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.8626201152801514, + "learning_rate": 3.464662875710805e-06, + "loss": 1.4701, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.9752259254455566, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.5004, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 3.1553750038146973, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.4652, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.9458937644958496, + "learning_rate": 3.458570268074736e-06, + "loss": 1.51, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.9626898765563965, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.484, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.8093032836914062, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.468, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.740411639213562, + "eval_runtime": 37.1036, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.928344488143921, + "learning_rate": 3.452477660438668e-06, + "loss": 1.4474, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 3.0413129329681396, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.4759, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.7930357456207275, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.4406, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.8075618743896484, + "learning_rate": 3.4463850528026e-06, + "loss": 1.48, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 3.019402503967285, + "learning_rate": 3.444354183590577e-06, + "loss": 1.4713, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 3.009504795074463, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5124, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.9452531337738037, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.464, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.6980783939361572, + "learning_rate": 3.438261575954509e-06, + "loss": 1.4143, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 3.039795160293579, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.4712, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.285968542098999, + "learning_rate": 3.434199837530463e-06, + "loss": 1.5227, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.7402914762496948, + "eval_runtime": 37.1273, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 3.035813093185425, + "learning_rate": 3.432168968318441e-06, + "loss": 1.4772, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 3.1172401905059814, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.4389, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 3.0828707218170166, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.5098, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.922488212585449, + "learning_rate": 3.426076360682372e-06, + "loss": 1.5239, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 3.03352689743042, + "learning_rate": 3.42404549147035e-06, + "loss": 1.5296, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.933380603790283, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.4851, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7250027656555176, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.3966, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.991846799850464, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.4863, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 3.0704684257507324, + "learning_rate": 3.415922014622259e-06, + "loss": 1.5366, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 3.0659193992614746, + "learning_rate": 3.413891145410236e-06, + "loss": 1.516, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.7458266019821167, + "eval_runtime": 37.06, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.9695920944213867, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.508, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 3.004558801651001, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.5092, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.8377230167388916, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.4909, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.860020875930786, + "learning_rate": 3.405767668562145e-06, + "loss": 1.4415, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8683009147644043, + "learning_rate": 3.403736799350122e-06, + "loss": 1.4997, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.91894268989563, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.4886, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.8158814907073975, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.4295, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.9528515338897705, + "learning_rate": 3.397644191714054e-06, + "loss": 1.5422, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.994856119155884, + "learning_rate": 3.395613322502031e-06, + "loss": 1.4909, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.8976950645446777, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.4959, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.7350144386291504, + "eval_runtime": 37.0807, + "eval_samples_per_second": 26.968, + "eval_steps_per_second": 1.133, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.73504638671875, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.4454, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.816823720932007, + "learning_rate": 3.389520714865963e-06, + "loss": 1.4864, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.706554889678955, + "learning_rate": 3.38748984565394e-06, + "loss": 1.483, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.9004647731781006, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.492, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8677759170532227, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.4786, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.83447527885437, + "learning_rate": 3.381397238017872e-06, + "loss": 1.4575, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.9942550659179688, + "learning_rate": 3.379366368805849e-06, + "loss": 1.4746, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.124676465988159, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.5406, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.9151222705841064, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.4748, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.734571933746338, + "learning_rate": 3.373273761169781e-06, + "loss": 1.4773, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.7389248609542847, + "eval_runtime": 37.1519, + "eval_samples_per_second": 26.917, + "eval_steps_per_second": 1.13, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.8919990062713623, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.4698, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.873870611190796, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.4509, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 3.0072309970855713, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5024, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.784576177597046, + "learning_rate": 3.36515028432169e-06, + "loss": 1.4405, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.8316547870635986, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.4697, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.8758575916290283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.495, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.879115343093872, + "learning_rate": 3.359057676685622e-06, + "loss": 1.5248, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.9559383392333984, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5229, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.6783695220947266, + "learning_rate": 3.354995938261576e-06, + "loss": 1.4251, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.8372304439544678, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.4649, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.7335349321365356, + "eval_runtime": 37.3498, + "eval_samples_per_second": 26.774, + "eval_steps_per_second": 1.125, + "step": 1100 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.429040350298112e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1100/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/latest new file mode 100644 index 0000000000000000000000000000000000000000..873b08355ae55e303606c474ab5fba25f9e8262a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/latest @@ -0,0 +1 @@ +global_step1200 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..b9b7677fd1972a131cdfa25cc7141cdfd4c10c45 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/trainer_state.json @@ -0,0 +1,9393 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.3150684931506849, + "eval_steps": 10, + "global_step": 1200, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6624906063079834, + "learning_rate": 4.77051177904143e-06, + "loss": 1.704, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 3.0075268745422363, + "learning_rate": 4.768480909829407e-06, + "loss": 1.6322, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.6070828437805176, + "learning_rate": 4.766450040617385e-06, + "loss": 1.6871, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 3.044222831726074, + "learning_rate": 4.764419171405362e-06, + "loss": 1.6652, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.773167610168457, + "learning_rate": 4.762388302193339e-06, + "loss": 1.7063, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.880176305770874, + "learning_rate": 4.760357432981316e-06, + "loss": 1.6775, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.6402764320373535, + "learning_rate": 4.758326563769294e-06, + "loss": 1.6694, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.9245846271514893, + "learning_rate": 4.756295694557271e-06, + "loss": 1.6688, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.623044967651367, + "learning_rate": 4.754264825345248e-06, + "loss": 1.6819, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.739896059036255, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.6401, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.8707164525985718, + "eval_runtime": 36.9513, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.603599786758423, + "learning_rate": 4.750203086921203e-06, + "loss": 1.6412, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.5571675300598145, + "learning_rate": 4.74817221770918e-06, + "loss": 1.5779, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.87091326713562, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.6878, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.8140673637390137, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7023, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.6550583839416504, + "learning_rate": 4.742079610073111e-06, + "loss": 1.5995, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.6259925365448, + "learning_rate": 4.740048740861089e-06, + "loss": 1.6194, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.971848726272583, + "learning_rate": 4.738017871649066e-06, + "loss": 1.6634, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6344778537750244, + "learning_rate": 4.735987002437044e-06, + "loss": 1.6502, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 2.9362456798553467, + "learning_rate": 4.73395613322502e-06, + "loss": 1.6948, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.8098835945129395, + "learning_rate": 4.731925264012998e-06, + "loss": 1.7161, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.8724393844604492, + "eval_runtime": 37.0876, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.933009386062622, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7014, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.8216991424560547, + "learning_rate": 4.727863525588953e-06, + "loss": 1.6192, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.7028493881225586, + "learning_rate": 4.725832656376929e-06, + "loss": 1.607, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.816818952560425, + "learning_rate": 4.723801787164907e-06, + "loss": 1.6183, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.7707865238189697, + "learning_rate": 4.721770917952885e-06, + "loss": 1.6795, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.779562473297119, + "learning_rate": 4.719740048740861e-06, + "loss": 1.63, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.907338857650757, + "learning_rate": 4.717709179528839e-06, + "loss": 1.644, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 3.073558807373047, + "learning_rate": 4.715678310316816e-06, + "loss": 1.6821, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.7556493282318115, + "learning_rate": 4.713647441104794e-06, + "loss": 1.6501, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6718344688415527, + "learning_rate": 4.71161657189277e-06, + "loss": 1.6225, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.854687213897705, + "eval_runtime": 36.9773, + "eval_samples_per_second": 27.044, + "eval_steps_per_second": 1.136, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.763335943222046, + "learning_rate": 4.709585702680748e-06, + "loss": 1.6311, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.7669479846954346, + "learning_rate": 4.707554833468725e-06, + "loss": 1.6809, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.711562395095825, + "learning_rate": 4.705523964256703e-06, + "loss": 1.741, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.799954652786255, + "learning_rate": 4.703493095044679e-06, + "loss": 1.6409, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.7278690338134766, + "learning_rate": 4.701462225832657e-06, + "loss": 1.6325, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.6802427768707275, + "learning_rate": 4.699431356620634e-06, + "loss": 1.5988, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.8826160430908203, + "learning_rate": 4.697400487408612e-06, + "loss": 1.7119, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.784343719482422, + "learning_rate": 4.695369618196588e-06, + "loss": 1.6855, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.683337688446045, + "learning_rate": 4.693338748984566e-06, + "loss": 1.6588, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.6646857261657715, + "learning_rate": 4.691307879772543e-06, + "loss": 1.623, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.8511513471603394, + "eval_runtime": 36.9852, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.8181183338165283, + "learning_rate": 4.68927701056052e-06, + "loss": 1.6171, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.8270857334136963, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.678, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.7685446739196777, + "learning_rate": 4.685215272136475e-06, + "loss": 1.6575, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.7003183364868164, + "learning_rate": 4.683184402924452e-06, + "loss": 1.6405, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.668750286102295, + "learning_rate": 4.681153533712429e-06, + "loss": 1.6541, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 3.065800905227661, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7288, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.615601062774658, + "learning_rate": 4.677091795288384e-06, + "loss": 1.6306, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.5948891639709473, + "learning_rate": 4.675060926076361e-06, + "loss": 1.6018, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.7550079822540283, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7009, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.868687868118286, + "learning_rate": 4.670999187652316e-06, + "loss": 1.6034, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.8473671674728394, + "eval_runtime": 37.0015, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.705962657928467, + "learning_rate": 4.668968318440293e-06, + "loss": 1.6084, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.7901813983917236, + "learning_rate": 4.66693744922827e-06, + "loss": 1.6171, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.688774824142456, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.6774, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.734617233276367, + "learning_rate": 4.662875710804225e-06, + "loss": 1.5796, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.742017984390259, + "learning_rate": 4.660844841592202e-06, + "loss": 1.6671, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.6438794136047363, + "learning_rate": 4.658813972380179e-06, + "loss": 1.5835, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6441051959991455, + "learning_rate": 4.656783103168157e-06, + "loss": 1.6909, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.824204206466675, + "learning_rate": 4.654752233956134e-06, + "loss": 1.5758, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.556638479232788, + "learning_rate": 4.652721364744111e-06, + "loss": 1.6504, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.897812843322754, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.6406, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.8537043333053589, + "eval_runtime": 37.257, + "eval_samples_per_second": 26.841, + "eval_steps_per_second": 1.127, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.6594088077545166, + "learning_rate": 4.648659626320066e-06, + "loss": 1.6966, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.79524302482605, + "learning_rate": 4.646628757108043e-06, + "loss": 1.6645, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.80291748046875, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.6281, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.7723848819732666, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.5959, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 3.0625650882720947, + "learning_rate": 4.640536149471974e-06, + "loss": 1.6313, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.775726079940796, + "learning_rate": 4.638505280259952e-06, + "loss": 1.6563, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.837758779525757, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.6727, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 3.041355609893799, + "learning_rate": 4.634443541835907e-06, + "loss": 1.6282, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.5775389671325684, + "learning_rate": 4.632412672623883e-06, + "loss": 1.651, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 3.026287794113159, + "learning_rate": 4.630381803411861e-06, + "loss": 1.6458, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.8461565971374512, + "eval_runtime": 37.3881, + "eval_samples_per_second": 26.747, + "eval_steps_per_second": 1.123, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.6781764030456543, + "learning_rate": 4.628350934199838e-06, + "loss": 1.5232, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.763350486755371, + "learning_rate": 4.626320064987816e-06, + "loss": 1.6186, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.7063372135162354, + "learning_rate": 4.624289195775792e-06, + "loss": 1.6784, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.824563980102539, + "learning_rate": 4.62225832656377e-06, + "loss": 1.6219, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.5335569381713867, + "learning_rate": 4.620227457351747e-06, + "loss": 1.6152, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.773644208908081, + "learning_rate": 4.618196588139724e-06, + "loss": 1.6179, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.9841721057891846, + "learning_rate": 4.616165718927701e-06, + "loss": 1.6258, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.684171676635742, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.6066, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.8817965984344482, + "learning_rate": 4.612103980503656e-06, + "loss": 1.6697, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.9846866130828857, + "learning_rate": 4.610073111291633e-06, + "loss": 1.6541, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.8485002517700195, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.904170513153076, + "learning_rate": 4.60804224207961e-06, + "loss": 1.6362, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.684476375579834, + "learning_rate": 4.606011372867588e-06, + "loss": 1.6303, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6692216396331787, + "learning_rate": 4.603980503655565e-06, + "loss": 1.638, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.73555588722229, + "learning_rate": 4.601949634443542e-06, + "loss": 1.6125, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.729525089263916, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.565, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.5972018241882324, + "learning_rate": 4.597887896019497e-06, + "loss": 1.5856, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7859137058258057, + "learning_rate": 4.595857026807474e-06, + "loss": 1.6325, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.662587881088257, + "learning_rate": 4.593826157595451e-06, + "loss": 1.6172, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.6380577087402344, + "learning_rate": 4.591795288383429e-06, + "loss": 1.6005, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8375349044799805, + "learning_rate": 4.589764419171406e-06, + "loss": 1.6507, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.833074688911438, + "eval_runtime": 37.0695, + "eval_samples_per_second": 26.976, + "eval_steps_per_second": 1.133, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.786713123321533, + "learning_rate": 4.587733549959383e-06, + "loss": 1.6283, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.6959030628204346, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.5994, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.6100692749023438, + "learning_rate": 4.583671811535338e-06, + "loss": 1.5884, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.7281863689422607, + "learning_rate": 4.581640942323315e-06, + "loss": 1.5737, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.7519376277923584, + "learning_rate": 4.579610073111292e-06, + "loss": 1.6531, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.9314866065979004, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.6755, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.9250576496124268, + "learning_rate": 4.575548334687247e-06, + "loss": 1.6604, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.6382815837860107, + "learning_rate": 4.573517465475224e-06, + "loss": 1.6773, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.6891751289367676, + "learning_rate": 4.571486596263201e-06, + "loss": 1.5849, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.7018136978149414, + "learning_rate": 4.569455727051178e-06, + "loss": 1.6358, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.838019847869873, + "eval_runtime": 37.0496, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.671952247619629, + "learning_rate": 4.567424857839156e-06, + "loss": 1.6207, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.882539749145508, + "learning_rate": 4.565393988627132e-06, + "loss": 1.6618, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.684070587158203, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.6502, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.5295751094818115, + "learning_rate": 4.561332250203087e-06, + "loss": 1.6078, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.0268194675445557, + "learning_rate": 4.559301380991065e-06, + "loss": 1.6312, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6895201206207275, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.5848, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.869950771331787, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.5951, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.909505605697632, + "learning_rate": 4.553208773354996e-06, + "loss": 1.6589, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.7604334354400635, + "learning_rate": 4.551177904142974e-06, + "loss": 1.5767, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.8792080879211426, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.6864, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.836847186088562, + "eval_runtime": 36.9142, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.607037305831909, + "learning_rate": 4.547116165718929e-06, + "loss": 1.5925, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.9338927268981934, + "learning_rate": 4.545085296506905e-06, + "loss": 1.6433, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.7723233699798584, + "learning_rate": 4.543054427294882e-06, + "loss": 1.5963, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.4122188091278076, + "learning_rate": 4.54102355808286e-06, + "loss": 1.5978, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.7582709789276123, + "learning_rate": 4.538992688870837e-06, + "loss": 1.581, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.6983039379119873, + "learning_rate": 4.536961819658814e-06, + "loss": 1.6253, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.6036365032196045, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.5957, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.759645700454712, + "learning_rate": 4.532900081234769e-06, + "loss": 1.6199, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.797744035720825, + "learning_rate": 4.530869212022746e-06, + "loss": 1.5854, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.752868890762329, + "learning_rate": 4.528838342810723e-06, + "loss": 1.619, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.8305871486663818, + "eval_runtime": 37.0307, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.7421889305114746, + "learning_rate": 4.526807473598701e-06, + "loss": 1.6182, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.6852707862854004, + "learning_rate": 4.524776604386678e-06, + "loss": 1.6135, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.861720323562622, + "learning_rate": 4.522745735174655e-06, + "loss": 1.6099, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.706851005554199, + "learning_rate": 4.520714865962632e-06, + "loss": 1.6597, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7992191314697266, + "learning_rate": 4.51868399675061e-06, + "loss": 1.6077, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.6248183250427246, + "learning_rate": 4.516653127538587e-06, + "loss": 1.6485, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.696040391921997, + "learning_rate": 4.514622258326564e-06, + "loss": 1.5853, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.5958261489868164, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.571, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.9715874195098877, + "learning_rate": 4.510560519902519e-06, + "loss": 1.6044, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.664834499359131, + "learning_rate": 4.508529650690496e-06, + "loss": 1.6568, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.8277279138565063, + "eval_runtime": 37.0489, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.869558811187744, + "learning_rate": 4.506498781478473e-06, + "loss": 1.5654, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.5814828872680664, + "learning_rate": 4.504467912266451e-06, + "loss": 1.5484, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.6158947944641113, + "learning_rate": 4.502437043054428e-06, + "loss": 1.6202, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.7424654960632324, + "learning_rate": 4.500406173842405e-06, + "loss": 1.6755, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.8670949935913086, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.6144, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.7914652824401855, + "learning_rate": 4.49634443541836e-06, + "loss": 1.6164, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.866380214691162, + "learning_rate": 4.494313566206336e-06, + "loss": 1.5961, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.966474771499634, + "learning_rate": 4.492282696994314e-06, + "loss": 1.6681, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.751291513442993, + "learning_rate": 4.490251827782291e-06, + "loss": 1.6338, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.58840012550354, + "learning_rate": 4.488220958570269e-06, + "loss": 1.5619, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.824889898300171, + "eval_runtime": 37.0821, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.786945104598999, + "learning_rate": 4.486190089358245e-06, + "loss": 1.5712, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4536845684051514, + "learning_rate": 4.484159220146223e-06, + "loss": 1.5487, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.780278205871582, + "learning_rate": 4.4821283509342e-06, + "loss": 1.6269, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.5427017211914062, + "learning_rate": 4.480097481722178e-06, + "loss": 1.6254, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.7540087699890137, + "learning_rate": 4.478066612510154e-06, + "loss": 1.6112, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.616352081298828, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.5637, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.686035633087158, + "learning_rate": 4.474004874086109e-06, + "loss": 1.5974, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.982365846633911, + "learning_rate": 4.471974004874087e-06, + "loss": 1.7034, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.6106085777282715, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.5482, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.7798519134521484, + "learning_rate": 4.467912266450041e-06, + "loss": 1.6475, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.820489764213562, + "eval_runtime": 37.0672, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.8634185791015625, + "learning_rate": 4.465881397238018e-06, + "loss": 1.6258, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.633626699447632, + "learning_rate": 4.463850528025995e-06, + "loss": 1.5806, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.719074249267578, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.6404, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.7545487880706787, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6404, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.6559669971466064, + "learning_rate": 4.457757920389927e-06, + "loss": 1.6104, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.9004898071289062, + "learning_rate": 4.455727051177904e-06, + "loss": 1.5513, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.6640021800994873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.5859, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.842026472091675, + "learning_rate": 4.451665312753859e-06, + "loss": 1.6241, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.784647226333618, + "learning_rate": 4.449634443541836e-06, + "loss": 1.627, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.7067618370056152, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.6276, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.8115602731704712, + "eval_runtime": 37.078, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.6582882404327393, + "learning_rate": 4.445572705117791e-06, + "loss": 1.5746, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7881102561950684, + "learning_rate": 4.443541835905768e-06, + "loss": 1.5749, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.9323439598083496, + "learning_rate": 4.441510966693745e-06, + "loss": 1.6421, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.63948655128479, + "learning_rate": 4.439480097481723e-06, + "loss": 1.5853, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.7746710777282715, + "learning_rate": 4.4374492282697e-06, + "loss": 1.6455, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.093895435333252, + "learning_rate": 4.435418359057677e-06, + "loss": 1.5311, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.526385545730591, + "learning_rate": 4.433387489845654e-06, + "loss": 1.6082, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.6604297161102295, + "learning_rate": 4.431356620633632e-06, + "loss": 1.61, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.6636247634887695, + "learning_rate": 4.429325751421609e-06, + "loss": 1.5815, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.5567514896392822, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6086, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.8170957565307617, + "eval_runtime": 37.4095, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.695563793182373, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.5836, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6009771823883057, + "learning_rate": 4.423233143785541e-06, + "loss": 1.5264, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.0490736961364746, + "learning_rate": 4.421202274573518e-06, + "loss": 1.5982, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.8609886169433594, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6265, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4415369033813477, + "learning_rate": 4.417140536149473e-06, + "loss": 1.4987, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.771657943725586, + "learning_rate": 4.415109666937449e-06, + "loss": 1.5494, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.739617347717285, + "learning_rate": 4.413078797725427e-06, + "loss": 1.6274, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.805098295211792, + "learning_rate": 4.411047928513404e-06, + "loss": 1.5686, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.7248735427856445, + "learning_rate": 4.409017059301382e-06, + "loss": 1.6027, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6653318405151367, + "learning_rate": 4.406986190089358e-06, + "loss": 1.5909, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.8076283931732178, + "eval_runtime": 37.5073, + "eval_samples_per_second": 26.661, + "eval_steps_per_second": 1.12, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.626783847808838, + "learning_rate": 4.404955320877336e-06, + "loss": 1.6205, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.6929268836975098, + "learning_rate": 4.402924451665313e-06, + "loss": 1.6146, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.6806800365448, + "learning_rate": 4.400893582453291e-06, + "loss": 1.5702, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.5407259464263916, + "learning_rate": 4.398862713241267e-06, + "loss": 1.5525, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.639122247695923, + "learning_rate": 4.396831844029245e-06, + "loss": 1.5826, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.669396162033081, + "learning_rate": 4.394800974817222e-06, + "loss": 1.5982, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.6377906799316406, + "learning_rate": 4.392770105605199e-06, + "loss": 1.5797, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6769638061523438, + "learning_rate": 4.390739236393176e-06, + "loss": 1.607, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.6347222328186035, + "learning_rate": 4.388708367181154e-06, + "loss": 1.6196, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.670680284500122, + "learning_rate": 4.386677497969131e-06, + "loss": 1.619, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.8124384880065918, + "eval_runtime": 37.4751, + "eval_samples_per_second": 26.684, + "eval_steps_per_second": 1.121, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.5886521339416504, + "learning_rate": 4.384646628757108e-06, + "loss": 1.6258, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.644566297531128, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6447, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.6432394981384277, + "learning_rate": 4.380584890333063e-06, + "loss": 1.5888, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.5602784156799316, + "learning_rate": 4.37855402112104e-06, + "loss": 1.5544, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.4874439239501953, + "learning_rate": 4.376523151909017e-06, + "loss": 1.5075, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.694748640060425, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.6498, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.582211494445801, + "learning_rate": 4.372461413484972e-06, + "loss": 1.5877, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.675713539123535, + "learning_rate": 4.370430544272949e-06, + "loss": 1.6045, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.7157161235809326, + "learning_rate": 4.368399675060926e-06, + "loss": 1.623, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.795076370239258, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6141, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.8062744140625, + "eval_runtime": 37.2435, + "eval_samples_per_second": 26.85, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6296374797821045, + "learning_rate": 4.364337936636881e-06, + "loss": 1.6087, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.362307067424858e-06, + "loss": 1.5182, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.7965915203094482, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.6025, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.807307481765747, + "learning_rate": 4.358245329000813e-06, + "loss": 1.574, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.584615468978882, + "learning_rate": 4.35621445978879e-06, + "loss": 1.5922, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.8472282886505127, + "learning_rate": 4.354183590576767e-06, + "loss": 1.5914, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.8471736907958984, + "learning_rate": 4.352152721364745e-06, + "loss": 1.5749, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.569225549697876, + "learning_rate": 4.350121852152722e-06, + "loss": 1.5591, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.688157796859741, + "learning_rate": 4.348090982940699e-06, + "loss": 1.6042, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.5865259170532227, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.6307, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.8040707111358643, + "eval_runtime": 36.9511, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.71850323677063, + "learning_rate": 4.344029244516654e-06, + "loss": 1.6043, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.7158093452453613, + "learning_rate": 4.341998375304631e-06, + "loss": 1.5907, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6879327297210693, + "learning_rate": 4.339967506092607e-06, + "loss": 1.6365, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.827277660369873, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.5928, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6709842681884766, + "learning_rate": 4.335905767668562e-06, + "loss": 1.5812, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.7071495056152344, + "learning_rate": 4.33387489845654e-06, + "loss": 1.5571, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6008732318878174, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.6048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.6238059997558594, + "learning_rate": 4.329813160032495e-06, + "loss": 1.5962, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.567065954208374, + "learning_rate": 4.327782290820471e-06, + "loss": 1.5713, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.573068857192993, + "learning_rate": 4.325751421608449e-06, + "loss": 1.536, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.8103471994400024, + "eval_runtime": 37.0956, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6245224475860596, + "learning_rate": 4.323720552396426e-06, + "loss": 1.5986, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.6243810653686523, + "learning_rate": 4.321689683184404e-06, + "loss": 1.589, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.5303313732147217, + "learning_rate": 4.31965881397238e-06, + "loss": 1.5793, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.5617806911468506, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.5704, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.7429656982421875, + "learning_rate": 4.315597075548335e-06, + "loss": 1.6174, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.6753251552581787, + "learning_rate": 4.313566206336312e-06, + "loss": 1.5879, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.5944488048553467, + "learning_rate": 4.311535337124289e-06, + "loss": 1.6082, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.9002904891967773, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.5521, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.5384247303009033, + "learning_rate": 4.307473598700244e-06, + "loss": 1.5751, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6015477180480957, + "learning_rate": 4.305442729488221e-06, + "loss": 1.5966, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8091154098510742, + "eval_runtime": 37.1075, + "eval_samples_per_second": 26.949, + "eval_steps_per_second": 1.132, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.6002886295318604, + "learning_rate": 4.303411860276198e-06, + "loss": 1.5403, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7755627632141113, + "learning_rate": 4.301380991064176e-06, + "loss": 1.6192, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.8351540565490723, + "learning_rate": 4.299350121852153e-06, + "loss": 1.6217, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.684007406234741, + "learning_rate": 4.29731925264013e-06, + "loss": 1.6242, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.8019614219665527, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6173, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.9122867584228516, + "learning_rate": 4.293257514216085e-06, + "loss": 1.5701, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.545732021331787, + "learning_rate": 4.291226645004062e-06, + "loss": 1.5494, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.9499566555023193, + "learning_rate": 4.289195775792039e-06, + "loss": 1.5848, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.8071892261505127, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6114, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.763864517211914, + "learning_rate": 4.285134037367994e-06, + "loss": 1.5768, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8057912588119507, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 3.001190423965454, + "learning_rate": 4.283103168155971e-06, + "loss": 1.5601, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.7640297412872314, + "learning_rate": 4.281072298943948e-06, + "loss": 1.5619, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.7372794151306152, + "learning_rate": 4.279041429731926e-06, + "loss": 1.562, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 3.0390751361846924, + "learning_rate": 4.277010560519903e-06, + "loss": 1.5519, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.758798599243164, + "learning_rate": 4.27497969130788e-06, + "loss": 1.527, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.96443772315979, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.5899, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.7941088676452637, + "learning_rate": 4.270917952883835e-06, + "loss": 1.617, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 3.0385100841522217, + "learning_rate": 4.268887083671812e-06, + "loss": 1.5996, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.6946637630462646, + "learning_rate": 4.266856214459789e-06, + "loss": 1.5641, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.776723861694336, + "learning_rate": 4.264825345247766e-06, + "loss": 1.6451, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.7911298274993896, + "eval_runtime": 37.0623, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.851224184036255, + "learning_rate": 4.262794476035744e-06, + "loss": 1.5302, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9045867919921875, + "learning_rate": 4.26076360682372e-06, + "loss": 1.5808, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.6568782329559326, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.5233, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.94191837310791, + "learning_rate": 4.256701868399675e-06, + "loss": 1.5912, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.726778984069824, + "learning_rate": 4.254670999187653e-06, + "loss": 1.5823, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.8704583644866943, + "learning_rate": 4.252640129975629e-06, + "loss": 1.5931, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.5721547603607178, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6128, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6883509159088135, + "learning_rate": 4.248578391551584e-06, + "loss": 1.5479, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.7186272144317627, + "learning_rate": 4.246547522339562e-06, + "loss": 1.5203, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.8326518535614014, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.5547, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.7948570251464844, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.7870523929595947, + "learning_rate": 4.242485783915517e-06, + "loss": 1.5891, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.866610288619995, + "learning_rate": 4.240454914703493e-06, + "loss": 1.5232, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.94161057472229, + "learning_rate": 4.23842404549147e-06, + "loss": 1.52, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.7331979274749756, + "learning_rate": 4.236393176279448e-06, + "loss": 1.5665, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.5375845432281494, + "learning_rate": 4.234362307067425e-06, + "loss": 1.5274, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.691927433013916, + "learning_rate": 4.232331437855402e-06, + "loss": 1.604, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7578394412994385, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.6, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.594538688659668, + "learning_rate": 4.228269699431357e-06, + "loss": 1.5928, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.5408968925476074, + "learning_rate": 4.226238830219334e-06, + "loss": 1.5908, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.689786672592163, + "learning_rate": 4.224207961007311e-06, + "loss": 1.5757, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.792494297027588, + "eval_runtime": 37.0822, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.735539197921753, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.5811, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.58634614944458, + "learning_rate": 4.220146222583266e-06, + "loss": 1.5808, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.7899138927459717, + "learning_rate": 4.218115353371243e-06, + "loss": 1.5801, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.6507744789123535, + "learning_rate": 4.21608448415922e-06, + "loss": 1.5585, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.6112217903137207, + "learning_rate": 4.214053614947198e-06, + "loss": 1.5329, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.729405164718628, + "learning_rate": 4.212022745735175e-06, + "loss": 1.6177, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.6431992053985596, + "learning_rate": 4.209991876523152e-06, + "loss": 1.5234, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.726243734359741, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.5926, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.6574180126190186, + "learning_rate": 4.205930138099107e-06, + "loss": 1.5424, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.6570956707000732, + "learning_rate": 4.203899268887084e-06, + "loss": 1.6119, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.781999945640564, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.528052806854248, + "learning_rate": 4.201868399675061e-06, + "loss": 1.598, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.6075727939605713, + "learning_rate": 4.199837530463039e-06, + "loss": 1.5886, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.588737964630127, + "learning_rate": 4.197806661251016e-06, + "loss": 1.507, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.773618221282959, + "learning_rate": 4.195775792038993e-06, + "loss": 1.5794, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.624725580215454, + "learning_rate": 4.19374492282697e-06, + "loss": 1.5495, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.662397861480713, + "learning_rate": 4.191714053614948e-06, + "loss": 1.5203, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.6617255210876465, + "learning_rate": 4.189683184402924e-06, + "loss": 1.5885, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.5784451961517334, + "learning_rate": 4.187652315190902e-06, + "loss": 1.5689, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.726011037826538, + "learning_rate": 4.185621445978879e-06, + "loss": 1.5913, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.6117324829101562, + "learning_rate": 4.183590576766857e-06, + "loss": 1.5453, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.7893352508544922, + "eval_runtime": 37.4096, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7025442123413086, + "learning_rate": 4.181559707554833e-06, + "loss": 1.5868, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.656102180480957, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6141, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.741962194442749, + "learning_rate": 4.177497969130788e-06, + "loss": 1.5516, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5887229442596436, + "learning_rate": 4.175467099918766e-06, + "loss": 1.5525, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.542985439300537, + "learning_rate": 4.173436230706742e-06, + "loss": 1.592, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.7797129154205322, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.5349, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.632887363433838, + "learning_rate": 4.169374492282697e-06, + "loss": 1.5548, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.691690444946289, + "learning_rate": 4.167343623070675e-06, + "loss": 1.5352, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.8922817707061768, + "learning_rate": 4.165312753858652e-06, + "loss": 1.5717, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.5517446994781494, + "learning_rate": 4.163281884646629e-06, + "loss": 1.5658, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.780922770500183, + "eval_runtime": 37.3755, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.718374490737915, + "learning_rate": 4.161251015434607e-06, + "loss": 1.5578, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.6249442100524902, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6018, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.7051141262054443, + "learning_rate": 4.157189277010561e-06, + "loss": 1.5182, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.677109956741333, + "learning_rate": 4.155158407798538e-06, + "loss": 1.5423, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.591059684753418, + "learning_rate": 4.153127538586516e-06, + "loss": 1.5694, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.699953556060791, + "learning_rate": 4.151096669374492e-06, + "loss": 1.5758, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.8837342262268066, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.5509, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.6566851139068604, + "learning_rate": 4.147034930950447e-06, + "loss": 1.5577, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.901057004928589, + "learning_rate": 4.145004061738425e-06, + "loss": 1.5797, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.796971321105957, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.5608, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.7797813415527344, + "eval_runtime": 36.8425, + "eval_samples_per_second": 27.143, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.605431318283081, + "learning_rate": 4.140942323314379e-06, + "loss": 1.491, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8084020614624023, + "learning_rate": 4.138911454102356e-06, + "loss": 1.5074, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.625511407852173, + "learning_rate": 4.136880584890333e-06, + "loss": 1.5404, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.6944286823272705, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.589, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.81792950630188, + "learning_rate": 4.132818846466288e-06, + "loss": 1.5646, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5813655853271484, + "learning_rate": 4.130787977254265e-06, + "loss": 1.5697, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.6733286380767822, + "learning_rate": 4.128757108042242e-06, + "loss": 1.5786, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.6972577571868896, + "learning_rate": 4.12672623883022e-06, + "loss": 1.5857, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7381093502044678, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6365, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.6162049770355225, + "learning_rate": 4.122664500406174e-06, + "loss": 1.5598, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.7869179248809814, + "eval_runtime": 36.974, + "eval_samples_per_second": 27.046, + "eval_steps_per_second": 1.136, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.7157270908355713, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.5533, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.560105085372925, + "learning_rate": 4.118602761982129e-06, + "loss": 1.5821, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.5261518955230713, + "learning_rate": 4.116571892770106e-06, + "loss": 1.5195, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7377054691314697, + "learning_rate": 4.114541023558083e-06, + "loss": 1.5066, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.632946252822876, + "learning_rate": 4.112510154346061e-06, + "loss": 1.5948, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.809671640396118, + "learning_rate": 4.110479285134038e-06, + "loss": 1.5435, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.843156099319458, + "learning_rate": 4.108448415922015e-06, + "loss": 1.5667, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.743614673614502, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5958, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.9596827030181885, + "learning_rate": 4.10438667749797e-06, + "loss": 1.5807, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.7321834564208984, + "learning_rate": 4.102355808285947e-06, + "loss": 1.4744, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.784859538078308, + "eval_runtime": 37.0146, + "eval_samples_per_second": 27.016, + "eval_steps_per_second": 1.135, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.010359764099121, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6146, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.744251251220703, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.5401, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.592690944671631, + "learning_rate": 4.096263200649879e-06, + "loss": 1.5508, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.786736011505127, + "learning_rate": 4.094232331437856e-06, + "loss": 1.5717, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.8773610591888428, + "learning_rate": 4.092201462225833e-06, + "loss": 1.5201, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.818610906600952, + "learning_rate": 4.090170593013811e-06, + "loss": 1.5282, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.9548542499542236, + "learning_rate": 4.088139723801787e-06, + "loss": 1.553, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.678842306137085, + "learning_rate": 4.086108854589765e-06, + "loss": 1.531, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.7757408618927, + "learning_rate": 4.084077985377742e-06, + "loss": 1.5371, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.6954638957977295, + "learning_rate": 4.08204711616572e-06, + "loss": 1.5462, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.781003475189209, + "eval_runtime": 37.0687, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6897518634796143, + "learning_rate": 4.080016246953696e-06, + "loss": 1.5979, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.581172466278076, + "learning_rate": 4.077985377741674e-06, + "loss": 1.5007, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.828234910964966, + "learning_rate": 4.075954508529651e-06, + "loss": 1.5857, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.6171951293945312, + "learning_rate": 4.073923639317629e-06, + "loss": 1.5937, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.6476316452026367, + "learning_rate": 4.071892770105605e-06, + "loss": 1.5864, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.7525634765625, + "learning_rate": 4.069861900893583e-06, + "loss": 1.5095, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.6881322860717773, + "learning_rate": 4.06783103168156e-06, + "loss": 1.4936, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.6104350090026855, + "learning_rate": 4.065800162469538e-06, + "loss": 1.5229, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.866706132888794, + "learning_rate": 4.063769293257514e-06, + "loss": 1.539, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.803999900817871, + "learning_rate": 4.061738424045492e-06, + "loss": 1.5792, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.7793465852737427, + "eval_runtime": 37.0749, + "eval_samples_per_second": 26.972, + "eval_steps_per_second": 1.133, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.670285940170288, + "learning_rate": 4.059707554833469e-06, + "loss": 1.5523, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.871267557144165, + "learning_rate": 4.057676685621446e-06, + "loss": 1.5623, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.6631205081939697, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6094, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.7580034732818604, + "learning_rate": 4.053614947197401e-06, + "loss": 1.5725, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.6412739753723145, + "learning_rate": 4.051584077985378e-06, + "loss": 1.4959, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.699176073074341, + "learning_rate": 4.049553208773355e-06, + "loss": 1.5586, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.622969150543213, + "learning_rate": 4.047522339561333e-06, + "loss": 1.5713, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.892338991165161, + "learning_rate": 4.04549147034931e-06, + "loss": 1.5471, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.77056622505188, + "learning_rate": 4.043460601137287e-06, + "loss": 1.5236, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.7122550010681152, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6562, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.7708193063735962, + "eval_runtime": 37.0743, + "eval_samples_per_second": 26.973, + "eval_steps_per_second": 1.133, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.702824592590332, + "learning_rate": 4.039398862713242e-06, + "loss": 1.534, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.6822307109832764, + "learning_rate": 4.037367993501219e-06, + "loss": 1.5368, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.7274515628814697, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6218, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.8114798069000244, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.532, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.8936192989349365, + "learning_rate": 4.031275385865151e-06, + "loss": 1.5124, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.7851672172546387, + "learning_rate": 4.029244516653128e-06, + "loss": 1.5596, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.815023183822632, + "learning_rate": 4.027213647441105e-06, + "loss": 1.5706, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5958921909332275, + "learning_rate": 4.025182778229083e-06, + "loss": 1.5161, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.779360771179199, + "learning_rate": 4.02315190901706e-06, + "loss": 1.5057, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.7411787509918213, + "learning_rate": 4.021121039805037e-06, + "loss": 1.5396, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.7778708934783936, + "eval_runtime": 37.0791, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.5709781646728516, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.5474, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8583388328552246, + "learning_rate": 4.017059301380992e-06, + "loss": 1.6149, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.6877572536468506, + "learning_rate": 4.015028432168969e-06, + "loss": 1.5292, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.680612087249756, + "learning_rate": 4.012997562956945e-06, + "loss": 1.596, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.653559446334839, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.4901, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.542787790298462, + "learning_rate": 4.0089358245329e-06, + "loss": 1.5295, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5476601123809814, + "learning_rate": 4.006904955320878e-06, + "loss": 1.5393, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.6367862224578857, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.5321, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.8112173080444336, + "learning_rate": 4.002843216896833e-06, + "loss": 1.5963, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.7136871814727783, + "learning_rate": 4.000812347684809e-06, + "loss": 1.5839, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.7763899564743042, + "eval_runtime": 37.2537, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.9095280170440674, + "learning_rate": 3.998781478472787e-06, + "loss": 1.5343, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.6196115016937256, + "learning_rate": 3.996750609260764e-06, + "loss": 1.5411, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6413371562957764, + "learning_rate": 3.994719740048742e-06, + "loss": 1.4862, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 3.0759036540985107, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6471, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.599147081375122, + "learning_rate": 3.990658001624696e-06, + "loss": 1.5556, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.726978302001953, + "learning_rate": 3.988627132412673e-06, + "loss": 1.5351, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.6760807037353516, + "learning_rate": 3.98659626320065e-06, + "loss": 1.5681, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.680483818054199, + "learning_rate": 3.984565393988627e-06, + "loss": 1.5225, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6649010181427, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.5325, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.638085126876831, + "learning_rate": 3.980503655564582e-06, + "loss": 1.564, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.7698904275894165, + "eval_runtime": 37.4492, + "eval_samples_per_second": 26.703, + "eval_steps_per_second": 1.122, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.647825241088867, + "learning_rate": 3.978472786352559e-06, + "loss": 1.5782, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.748222589492798, + "learning_rate": 3.976441917140536e-06, + "loss": 1.556, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.6319026947021484, + "learning_rate": 3.974411047928514e-06, + "loss": 1.5378, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.6759872436523438, + "learning_rate": 3.972380178716491e-06, + "loss": 1.6018, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.6760940551757812, + "learning_rate": 3.970349309504468e-06, + "loss": 1.5742, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.5565083026885986, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.565, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.5986931324005127, + "learning_rate": 3.966287571080423e-06, + "loss": 1.5124, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.713092803955078, + "learning_rate": 3.9642567018684e-06, + "loss": 1.523, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.6203417778015137, + "learning_rate": 3.962225832656377e-06, + "loss": 1.5601, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.604440927505493, + "learning_rate": 3.960194963444355e-06, + "loss": 1.5485, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.7633332014083862, + "eval_runtime": 37.49, + "eval_samples_per_second": 26.674, + "eval_steps_per_second": 1.12, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.5513081550598145, + "learning_rate": 3.958164094232332e-06, + "loss": 1.479, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.7990963459014893, + "learning_rate": 3.956133225020309e-06, + "loss": 1.5355, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.651113271713257, + "learning_rate": 3.954102355808286e-06, + "loss": 1.5406, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.7236671447753906, + "learning_rate": 3.952071486596264e-06, + "loss": 1.5314, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.697598695755005, + "learning_rate": 3.950040617384241e-06, + "loss": 1.5398, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.8271944522857666, + "learning_rate": 3.948009748172218e-06, + "loss": 1.54, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.733125925064087, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.5236, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.7555017471313477, + "learning_rate": 3.943948009748173e-06, + "loss": 1.5808, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.6592960357666016, + "learning_rate": 3.94191714053615e-06, + "loss": 1.5398, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.6870126724243164, + "learning_rate": 3.939886271324127e-06, + "loss": 1.5517, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.760675311088562, + "eval_runtime": 36.9372, + "eval_samples_per_second": 27.073, + "eval_steps_per_second": 1.137, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.7469022274017334, + "learning_rate": 3.937855402112104e-06, + "loss": 1.605, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.5255792140960693, + "learning_rate": 3.935824532900082e-06, + "loss": 1.5061, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.589054822921753, + "learning_rate": 3.933793663688058e-06, + "loss": 1.4971, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.7874889373779297, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.5643, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.819451093673706, + "learning_rate": 3.929731925264013e-06, + "loss": 1.5797, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.6516339778900146, + "learning_rate": 3.927701056051991e-06, + "loss": 1.5048, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.979091167449951, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.5174, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.743058204650879, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.5432, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.6864030361175537, + "learning_rate": 3.921608448415922e-06, + "loss": 1.5147, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.7751128673553467, + "learning_rate": 3.9195775792039e-06, + "loss": 1.4857, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.7609487771987915, + "eval_runtime": 37.09, + "eval_samples_per_second": 26.961, + "eval_steps_per_second": 1.132, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.626188278198242, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.5428, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.554067850112915, + "learning_rate": 3.915515840779855e-06, + "loss": 1.5153, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.8580875396728516, + "learning_rate": 3.913484971567831e-06, + "loss": 1.5562, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.7405097484588623, + "learning_rate": 3.911454102355808e-06, + "loss": 1.565, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.5650815963745117, + "learning_rate": 3.909423233143786e-06, + "loss": 1.4756, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.9405910968780518, + "learning_rate": 3.907392363931763e-06, + "loss": 1.5611, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.687448024749756, + "learning_rate": 3.90536149471974e-06, + "loss": 1.5405, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.7373857498168945, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.6055, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.875298261642456, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6215, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.713200330734253, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6493, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.7608598470687866, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.541935682296753, + "learning_rate": 3.897238017871649e-06, + "loss": 1.5257, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.582045316696167, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.5253, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.6884682178497314, + "learning_rate": 3.893176279447604e-06, + "loss": 1.5329, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.6456501483917236, + "learning_rate": 3.891145410235581e-06, + "loss": 1.5167, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.60788893699646, + "learning_rate": 3.889114541023558e-06, + "loss": 1.5945, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.708646774291992, + "learning_rate": 3.887083671811536e-06, + "loss": 1.5704, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.666922092437744, + "learning_rate": 3.885052802599513e-06, + "loss": 1.52, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.6623599529266357, + "learning_rate": 3.88302193338749e-06, + "loss": 1.4791, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.9031424522399902, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.5954, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.763303279876709, + "learning_rate": 3.878960194963445e-06, + "loss": 1.5618, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.7647819519042969, + "eval_runtime": 37.1392, + "eval_samples_per_second": 26.926, + "eval_steps_per_second": 1.131, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 3.0733320713043213, + "learning_rate": 3.876929325751422e-06, + "loss": 1.5191, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.7809090614318848, + "learning_rate": 3.874898456539399e-06, + "loss": 1.5804, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.6357381343841553, + "learning_rate": 3.872867587327377e-06, + "loss": 1.4843, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.680595874786377, + "learning_rate": 3.870836718115354e-06, + "loss": 1.5339, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.5908069610595703, + "learning_rate": 3.868805848903331e-06, + "loss": 1.5049, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.680586338043213, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.5188, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.763528823852539, + "learning_rate": 3.864744110479286e-06, + "loss": 1.5395, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.8522050380706787, + "learning_rate": 3.862713241267262e-06, + "loss": 1.562, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.806962490081787, + "learning_rate": 3.86068237205524e-06, + "loss": 1.5419, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.9716053009033203, + "learning_rate": 3.858651502843217e-06, + "loss": 1.5691, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.7608845233917236, + "eval_runtime": 37.1433, + "eval_samples_per_second": 26.923, + "eval_steps_per_second": 1.131, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.613098382949829, + "learning_rate": 3.856620633631195e-06, + "loss": 1.5297, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.5690557956695557, + "learning_rate": 3.854589764419171e-06, + "loss": 1.4964, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.8001089096069336, + "learning_rate": 3.852558895207149e-06, + "loss": 1.5253, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.6715309619903564, + "learning_rate": 3.850528025995126e-06, + "loss": 1.601, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.7044479846954346, + "learning_rate": 3.848497156783104e-06, + "loss": 1.5526, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.7368102073669434, + "learning_rate": 3.84646628757108e-06, + "loss": 1.5088, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.9358696937561035, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.5202, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.858172655105591, + "learning_rate": 3.842404549147035e-06, + "loss": 1.5378, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.6723833084106445, + "learning_rate": 3.840373679935013e-06, + "loss": 1.5134, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.748187780380249, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.4797, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.75319242477417, + "eval_runtime": 37.1302, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.4821436405181885, + "learning_rate": 3.836311941510967e-06, + "loss": 1.5382, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.5585227012634277, + "learning_rate": 3.834281072298944e-06, + "loss": 1.5373, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.927334785461426, + "learning_rate": 3.832250203086921e-06, + "loss": 1.5447, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.6096436977386475, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.5062, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.677304744720459, + "learning_rate": 3.828188464662876e-06, + "loss": 1.5671, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.9606077671051025, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6138, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.6971657276153564, + "learning_rate": 3.82412672623883e-06, + "loss": 1.5544, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.6380839347839355, + "learning_rate": 3.822095857026808e-06, + "loss": 1.4964, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.7537858486175537, + "learning_rate": 3.820064987814785e-06, + "loss": 1.5429, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.6189136505126953, + "learning_rate": 3.818034118602762e-06, + "loss": 1.4956, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.7546342611312866, + "eval_runtime": 37.0828, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5744292736053467, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.5256, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.5503833293914795, + "learning_rate": 3.813972380178717e-06, + "loss": 1.5211, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.979384422302246, + "learning_rate": 3.811941510966694e-06, + "loss": 1.55, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.867161512374878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5655, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.8006832599639893, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.4902, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.594712734222412, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.4451, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.6632349491119385, + "learning_rate": 3.803818034118603e-06, + "loss": 1.531, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.859894275665283, + "learning_rate": 3.80178716490658e-06, + "loss": 1.5018, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.6405653953552246, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.5765, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.7285192012786865, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.5537, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.7618242502212524, + "eval_runtime": 37.0898, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.6882379055023193, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.5143, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.706550121307373, + "learning_rate": 3.793663688058489e-06, + "loss": 1.4857, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.790907382965088, + "learning_rate": 3.791632818846467e-06, + "loss": 1.5982, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.631096363067627, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.5656, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.6981775760650635, + "learning_rate": 3.787571080422421e-06, + "loss": 1.567, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.612086296081543, + "learning_rate": 3.785540211210398e-06, + "loss": 1.5723, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.718630075454712, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.5565, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.6265041828155518, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.5326, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.578124523162842, + "learning_rate": 3.77944760357433e-06, + "loss": 1.533, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.832695484161377, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6133, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.7516050338745117, + "eval_runtime": 37.3173, + "eval_samples_per_second": 26.797, + "eval_steps_per_second": 1.125, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.851625442504883, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.5836, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.773354995938262e-06, + "loss": 1.5069, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.5738255977630615, + "learning_rate": 3.771324126726239e-06, + "loss": 1.5531, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.583575963973999, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.5589, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.7729389667510986, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.5901, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.6322245597839355, + "learning_rate": 3.765231519090171e-06, + "loss": 1.5423, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.6382699012756348, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.4574, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.7393112182617188, + "learning_rate": 3.761169780666125e-06, + "loss": 1.5351, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.684398651123047, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.4873, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.89401912689209, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.5358, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.7523635625839233, + "eval_runtime": 37.3763, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.629870891571045, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.5566, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.6936254501342773, + "learning_rate": 3.753046303818034e-06, + "loss": 1.5016, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.5889103412628174, + "learning_rate": 3.751015434606012e-06, + "loss": 1.5775, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.610703468322754, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.5508, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6318750381469727, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.5892, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.6650729179382324, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.5106, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.577174186706543, + "learning_rate": 3.742891957757921e-06, + "loss": 1.54, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.7274489402770996, + "learning_rate": 3.740861088545898e-06, + "loss": 1.5803, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6849985122680664, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.5393, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.811479091644287, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.5862, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.7490049600601196, + "eval_runtime": 36.9083, + "eval_samples_per_second": 27.094, + "eval_steps_per_second": 1.138, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.640002489089966, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.5149, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.6350319385528564, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5132, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.689814567565918, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.473, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.7181100845336914, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.4353, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.760629415512085, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.4859, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.9322562217712402, + "learning_rate": 3.724614134849716e-06, + "loss": 1.4733, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.280315399169922, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.5769, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 3.1343884468078613, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.4766, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.8537609577178955, + "learning_rate": 3.718521527213648e-06, + "loss": 1.5586, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.895299196243286, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5232, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.7473852634429932, + "eval_runtime": 37.006, + "eval_samples_per_second": 27.023, + "eval_steps_per_second": 1.135, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.82454776763916, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5057, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.711296558380127, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.4187, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.6281023025512695, + "learning_rate": 3.710398050365557e-06, + "loss": 1.4997, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.775632381439209, + "learning_rate": 3.708367181153534e-06, + "loss": 1.5087, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.8021788597106934, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.474, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 3.09647536277771, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5015, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.810450315475464, + "learning_rate": 3.702274573517466e-06, + "loss": 1.4681, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.65166974067688, + "learning_rate": 3.700243704305443e-06, + "loss": 1.4484, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.755690574645996, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5084, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.8223068714141846, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.5015, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.7516940832138062, + "eval_runtime": 37.0227, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.8401107788085938, + "learning_rate": 3.694151096669375e-06, + "loss": 1.4597, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 3.010011672973633, + "learning_rate": 3.692120227457352e-06, + "loss": 1.5625, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8575570583343506, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.4828, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.8527276515960693, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5074, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.81260085105896, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.4132, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.988429069519043, + "learning_rate": 3.683996750609261e-06, + "loss": 1.4803, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.964179515838623, + "learning_rate": 3.681965881397238e-06, + "loss": 1.4871, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.7151682376861572, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.5061, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.764187812805176, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.5188, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.810554265975952, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.4779, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.7422841787338257, + "eval_runtime": 37.0641, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.7640810012817383, + "learning_rate": 3.673842404549147e-06, + "loss": 1.4827, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.7683937549591064, + "learning_rate": 3.671811535337125e-06, + "loss": 1.4775, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.86434268951416, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.4752, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 3.0976617336273193, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.5099, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 3.107388496398926, + "learning_rate": 3.665718927701056e-06, + "loss": 1.5646, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 3.032921314239502, + "learning_rate": 3.663688058489034e-06, + "loss": 1.5048, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.9377830028533936, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5333, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.706542730331421, + "learning_rate": 3.659626320064988e-06, + "loss": 1.4903, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.819936513900757, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.4441, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.805360794067383, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.5122, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.747745156288147, + "eval_runtime": 37.0801, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.848580837249756, + "learning_rate": 3.65353371242892e-06, + "loss": 1.4984, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.772721290588379, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5026, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.760772943496704, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.4005, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.937507390975952, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.5131, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.923210859298706, + "learning_rate": 3.645410235580829e-06, + "loss": 1.4661, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8029167652130127, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.518, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.752716541290283, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.5109, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.7767865657806396, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.4831, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8483617305755615, + "learning_rate": 3.637286758732738e-06, + "loss": 1.4872, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.8714234828948975, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.5109, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.7455583810806274, + "eval_runtime": 37.0892, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7373011112213135, + "learning_rate": 3.633225020308692e-06, + "loss": 1.55, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.942373752593994, + "learning_rate": 3.63119415109667e-06, + "loss": 1.5079, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.8790931701660156, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.4669, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.9564976692199707, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5555, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.811699390411377, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.4457, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.926544427871704, + "learning_rate": 3.623070674248579e-06, + "loss": 1.4368, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.8694913387298584, + "learning_rate": 3.621039805036556e-06, + "loss": 1.4875, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 3.0294361114501953, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.4815, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8643198013305664, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.5031, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 3.0244719982147217, + "learning_rate": 3.614947197400488e-06, + "loss": 1.4914, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.7504756450653076, + "eval_runtime": 37.0238, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.8641672134399414, + "learning_rate": 3.612916328188465e-06, + "loss": 1.5122, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.8713624477386475, + "learning_rate": 3.610885458976442e-06, + "loss": 1.43, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.990912914276123, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.5124, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.963945150375366, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.533, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.802168846130371, + "learning_rate": 3.604792851340374e-06, + "loss": 1.4531, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 3.0556633472442627, + "learning_rate": 3.602761982128351e-06, + "loss": 1.4618, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.95744252204895, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5356, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.813145399093628, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.5134, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.8382728099823, + "learning_rate": 3.596669374492283e-06, + "loss": 1.4413, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.821100950241089, + "learning_rate": 3.59463850528026e-06, + "loss": 1.4748, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.7488586902618408, + "eval_runtime": 37.0361, + "eval_samples_per_second": 27.001, + "eval_steps_per_second": 1.134, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.930708408355713, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.5177, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.9312615394592285, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.4493, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 3.0051841735839844, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.5208, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.793515205383301, + "learning_rate": 3.586515028432169e-06, + "loss": 1.4813, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.7595956325531006, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5379, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.992871046066284, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.5153, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.7270801067352295, + "learning_rate": 3.580422420796101e-06, + "loss": 1.4392, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.797985076904297, + "learning_rate": 3.578391551584078e-06, + "loss": 1.4728, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9595260620117188, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.5158, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.8131442070007324, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.4214, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.7503563165664673, + "eval_runtime": 37.3831, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.124, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8242709636688232, + "learning_rate": 3.57229894394801e-06, + "loss": 1.5027, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.8973116874694824, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.5294, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 3.1207027435302734, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.4902, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.803910732269287, + "learning_rate": 3.566206336311942e-06, + "loss": 1.466, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.813541889190674, + "learning_rate": 3.564175467099919e-06, + "loss": 1.4957, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.9068171977996826, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.4435, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.8491995334625244, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.4432, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.8785598278045654, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.4791, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7701117992401123, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.4607, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.4821, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.7518969774246216, + "eval_runtime": 37.4808, + "eval_samples_per_second": 26.68, + "eval_steps_per_second": 1.121, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.5228, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.8513107299804688, + "learning_rate": 3.551990251827783e-06, + "loss": 1.4812, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.937448024749756, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.4538, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.8307411670684814, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.4807, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.7986767292022705, + "learning_rate": 3.545897644191714e-06, + "loss": 1.4706, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 3.1135106086730957, + "learning_rate": 3.543866774979692e-06, + "loss": 1.4574, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.7756283283233643, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5114, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.8114547729492188, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.4458, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 3.0544545650482178, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.4934, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8990671634674072, + "learning_rate": 3.535743298131601e-06, + "loss": 1.4808, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.7468736171722412, + "eval_runtime": 36.9135, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.976139545440674, + "learning_rate": 3.533712428919578e-06, + "loss": 1.5123, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.909146785736084, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.5204, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7225229740142822, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.4746, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.6145517826080322, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.5049, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.8660519123077393, + "learning_rate": 3.525588952071487e-06, + "loss": 1.5173, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.979801654815674, + "learning_rate": 3.523558082859464e-06, + "loss": 1.5146, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.7500879764556885, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.514, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.6923978328704834, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.455, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.6728951930999756, + "learning_rate": 3.517465475223396e-06, + "loss": 1.4712, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.769507884979248, + "learning_rate": 3.515434606011373e-06, + "loss": 1.444, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.7491599321365356, + "eval_runtime": 37.066, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.9768483638763428, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.5005, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.993196725845337, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.5347, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.886883497238159, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.4959, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.9152941703796387, + "learning_rate": 3.507311129163282e-06, + "loss": 1.4641, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.966872215270996, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.4232, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.9853367805480957, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.4883, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.9353625774383545, + "learning_rate": 3.501218521527214e-06, + "loss": 1.4826, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.7860453128814697, + "learning_rate": 3.499187652315191e-06, + "loss": 1.42, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.9269113540649414, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.5087, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 3.052962303161621, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5042, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.7421623468399048, + "eval_runtime": 37.103, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.7328579425811768, + "learning_rate": 3.493095044679123e-06, + "loss": 1.4375, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.876138925552368, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5559, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.7500345706939697, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.4043, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.836596965789795, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.4808, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.797060966491699, + "learning_rate": 3.484971567831032e-06, + "loss": 1.405, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.928368091583252, + "learning_rate": 3.482940698619009e-06, + "loss": 1.4765, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.923375368118286, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.5134, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.8624660968780518, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.4881, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.9593892097473145, + "learning_rate": 3.476848090982941e-06, + "loss": 1.4869, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.8019728660583496, + "learning_rate": 3.474817221770918e-06, + "loss": 1.468, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.7386831045150757, + "eval_runtime": 37.1034, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.8965771198272705, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.5604, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.8114798069000244, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.5092, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.9554526805877686, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.4869, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.981499195098877, + "learning_rate": 3.466693744922827e-06, + "loss": 1.4454, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.8626201152801514, + "learning_rate": 3.464662875710805e-06, + "loss": 1.4701, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.9752259254455566, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.5004, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 3.1553750038146973, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.4652, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.9458937644958496, + "learning_rate": 3.458570268074736e-06, + "loss": 1.51, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.9626898765563965, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.484, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.8093032836914062, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.468, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.740411639213562, + "eval_runtime": 37.1036, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.928344488143921, + "learning_rate": 3.452477660438668e-06, + "loss": 1.4474, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 3.0413129329681396, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.4759, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.7930357456207275, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.4406, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.8075618743896484, + "learning_rate": 3.4463850528026e-06, + "loss": 1.48, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 3.019402503967285, + "learning_rate": 3.444354183590577e-06, + "loss": 1.4713, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 3.009504795074463, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5124, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.9452531337738037, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.464, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.6980783939361572, + "learning_rate": 3.438261575954509e-06, + "loss": 1.4143, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 3.039795160293579, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.4712, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.285968542098999, + "learning_rate": 3.434199837530463e-06, + "loss": 1.5227, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.7402914762496948, + "eval_runtime": 37.1273, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 3.035813093185425, + "learning_rate": 3.432168968318441e-06, + "loss": 1.4772, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 3.1172401905059814, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.4389, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 3.0828707218170166, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.5098, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.922488212585449, + "learning_rate": 3.426076360682372e-06, + "loss": 1.5239, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 3.03352689743042, + "learning_rate": 3.42404549147035e-06, + "loss": 1.5296, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.933380603790283, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.4851, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7250027656555176, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.3966, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.991846799850464, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.4863, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 3.0704684257507324, + "learning_rate": 3.415922014622259e-06, + "loss": 1.5366, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 3.0659193992614746, + "learning_rate": 3.413891145410236e-06, + "loss": 1.516, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.7458266019821167, + "eval_runtime": 37.06, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.9695920944213867, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.508, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 3.004558801651001, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.5092, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.8377230167388916, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.4909, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.860020875930786, + "learning_rate": 3.405767668562145e-06, + "loss": 1.4415, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8683009147644043, + "learning_rate": 3.403736799350122e-06, + "loss": 1.4997, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.91894268989563, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.4886, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.8158814907073975, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.4295, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.9528515338897705, + "learning_rate": 3.397644191714054e-06, + "loss": 1.5422, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.994856119155884, + "learning_rate": 3.395613322502031e-06, + "loss": 1.4909, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.8976950645446777, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.4959, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.7350144386291504, + "eval_runtime": 37.0807, + "eval_samples_per_second": 26.968, + "eval_steps_per_second": 1.133, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.73504638671875, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.4454, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.816823720932007, + "learning_rate": 3.389520714865963e-06, + "loss": 1.4864, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.706554889678955, + "learning_rate": 3.38748984565394e-06, + "loss": 1.483, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.9004647731781006, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.492, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8677759170532227, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.4786, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.83447527885437, + "learning_rate": 3.381397238017872e-06, + "loss": 1.4575, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.9942550659179688, + "learning_rate": 3.379366368805849e-06, + "loss": 1.4746, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.124676465988159, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.5406, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.9151222705841064, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.4748, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.734571933746338, + "learning_rate": 3.373273761169781e-06, + "loss": 1.4773, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.7389248609542847, + "eval_runtime": 37.1519, + "eval_samples_per_second": 26.917, + "eval_steps_per_second": 1.13, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.8919990062713623, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.4698, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.873870611190796, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.4509, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 3.0072309970855713, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5024, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.784576177597046, + "learning_rate": 3.36515028432169e-06, + "loss": 1.4405, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.8316547870635986, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.4697, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.8758575916290283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.495, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.879115343093872, + "learning_rate": 3.359057676685622e-06, + "loss": 1.5248, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.9559383392333984, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5229, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.6783695220947266, + "learning_rate": 3.354995938261576e-06, + "loss": 1.4251, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.8372304439544678, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.4649, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.7335349321365356, + "eval_runtime": 37.3498, + "eval_samples_per_second": 26.774, + "eval_steps_per_second": 1.125, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.729576826095581, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.4263, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.7717669010162354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.4467, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.8166871070861816, + "learning_rate": 3.346872461413485e-06, + "loss": 1.4937, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.845677137374878, + "learning_rate": 3.344841592201463e-06, + "loss": 1.4303, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.9527740478515625, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.4651, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.810873031616211, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.4521, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.77546763420105, + "learning_rate": 3.338748984565394e-06, + "loss": 1.4857, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.9098589420318604, + "learning_rate": 3.336718115353372e-06, + "loss": 1.4988, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.797712802886963, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.5307, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.916668176651001, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.4828, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.7372808456420898, + "eval_runtime": 36.888, + "eval_samples_per_second": 27.109, + "eval_steps_per_second": 1.139, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.8220417499542236, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.4636, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.851010322570801, + "learning_rate": 3.32859463850528e-06, + "loss": 1.4579, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.983619213104248, + "learning_rate": 3.326563769293258e-06, + "loss": 1.4801, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.750732898712158, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.4438, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.631105422973633, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.4439, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.944300889968872, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.437, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8841664791107178, + "learning_rate": 3.318440292445167e-06, + "loss": 1.4201, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.917189121246338, + "learning_rate": 3.316409423233144e-06, + "loss": 1.4489, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.112077236175537, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.4449, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.9847209453582764, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.5153, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.7323338985443115, + "eval_runtime": 36.9935, + "eval_samples_per_second": 27.032, + "eval_steps_per_second": 1.135, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.948901414871216, + "learning_rate": 3.310316815597076e-06, + "loss": 1.473, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.84930419921875, + "learning_rate": 3.308285946385053e-06, + "loss": 1.4796, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.9120237827301025, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5392, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.7088463306427, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.4206, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.7495498657226562, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.4759, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.9146509170532227, + "learning_rate": 3.300162469536962e-06, + "loss": 1.51, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.781975030899048, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.4667, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.7524054050445557, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.5036, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.7672629356384277, + "learning_rate": 3.294069861900894e-06, + "loss": 1.4827, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.875802516937256, + "learning_rate": 3.292038992688871e-06, + "loss": 1.519, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.7273142337799072, + "eval_runtime": 37.0032, + "eval_samples_per_second": 27.025, + "eval_steps_per_second": 1.135, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.8699231147766113, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.474, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.8670589923858643, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.5039, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.9946320056915283, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5226, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.767071008682251, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.4221, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8807036876678467, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.486, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.687720537185669, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.3944, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.8170154094696045, + "learning_rate": 3.277822908204712e-06, + "loss": 1.5213, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.905557632446289, + "learning_rate": 3.275792038992689e-06, + "loss": 1.4598, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.643657922744751, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.4282, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.798712968826294, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.4752, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.7295492887496948, + "eval_runtime": 37.0432, + "eval_samples_per_second": 26.996, + "eval_steps_per_second": 1.134, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.8480567932128906, + "learning_rate": 3.269699431356621e-06, + "loss": 1.4322, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.8311381340026855, + "learning_rate": 3.267668562144598e-06, + "loss": 1.4744, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.83522891998291, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.4578, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.757307767868042, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.4469, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.9117610454559326, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.5034, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.8141143321990967, + "learning_rate": 3.259545085296507e-06, + "loss": 1.473, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 3.050138473510742, + "learning_rate": 3.257514216084484e-06, + "loss": 1.528, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.660510540008545, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.4335, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.898832082748413, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.5335, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.955589532852173, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.4832, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.7293461561203003, + "eval_runtime": 37.0096, + "eval_samples_per_second": 27.02, + "eval_steps_per_second": 1.135, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.8524296283721924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.4728, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.9741404056549072, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.4805, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.738152265548706, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.4532, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.717066764831543, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.4741, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.8367488384246826, + "learning_rate": 3.241267262388302e-06, + "loss": 1.4773, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.987780809402466, + "learning_rate": 3.23923639317628e-06, + "loss": 1.4214, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.705444097518921, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.4537, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.8966188430786133, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.4552, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.9027600288391113, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.4883, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.965813159942627, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5318, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.7236050367355347, + "eval_runtime": 37.0306, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.6390798091888428, + "learning_rate": 3.229082047116166e-06, + "loss": 1.4431, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.822848081588745, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5163, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7289068698883057, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.4551, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.766545057296753, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.4648, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.8657522201538086, + "learning_rate": 3.220958570268075e-06, + "loss": 1.4825, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.7456932067871094, + "learning_rate": 3.218927701056052e-06, + "loss": 1.3909, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 3.0274853706359863, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.4461, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.990029811859131, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.4057, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 3.086756706237793, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5251, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 3.0227134227752686, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.427, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.7248852252960205, + "eval_runtime": 37.0619, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.815480947494507, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.469, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.9171998500823975, + "learning_rate": 3.206742485783916e-06, + "loss": 1.4977, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 3.087686777114868, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.4233, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.7349331378936768, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.4187, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.9189705848693848, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.4949, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 3.0483345985412598, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5041, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.7958145141601562, + "learning_rate": 3.196588139723802e-06, + "loss": 1.4009, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 3.081714153289795, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.4718, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.7897303104400635, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.4926, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.899045944213867, + "learning_rate": 3.190495532087734e-06, + "loss": 1.4524, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.7273337841033936, + "eval_runtime": 37.0682, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.979063034057617, + "learning_rate": 3.188464662875711e-06, + "loss": 1.4706, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.8633804321289062, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5273, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.8997795581817627, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.4638, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.906151294708252, + "learning_rate": 3.182372055239643e-06, + "loss": 1.4187, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 3.0212671756744385, + "learning_rate": 3.18034118602762e-06, + "loss": 1.4213, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 3.0891034603118896, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.4921, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.9302260875701904, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.4787, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.8741140365600586, + "learning_rate": 3.174248578391552e-06, + "loss": 1.4591, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 3.0567023754119873, + "learning_rate": 3.172217709179529e-06, + "loss": 1.4636, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.80245304107666, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.4523, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.7270530462265015, + "eval_runtime": 37.325, + "eval_samples_per_second": 26.792, + "eval_steps_per_second": 1.125, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.816023349761963, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.4579, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.7405545711517334, + "learning_rate": 3.166125101543461e-06, + "loss": 1.4225, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.9569272994995117, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.483, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 3.0433337688446045, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.4846, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.856781244277954, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.426, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.8880672454833984, + "learning_rate": 3.15800162469537e-06, + "loss": 1.4764, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 3.0363011360168457, + "learning_rate": 3.155970755483347e-06, + "loss": 1.4916, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 3.0171539783477783, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.4596, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 3.183234930038452, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.4936, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.802027463912964, + "learning_rate": 3.149878147847279e-06, + "loss": 1.4733, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.7253632545471191, + "eval_runtime": 37.4737, + "eval_samples_per_second": 26.685, + "eval_steps_per_second": 1.121, + "step": 1200 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.740771291234304e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1200/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/latest new file mode 100644 index 0000000000000000000000000000000000000000..d71c4c79bfb8e74842ee005fbfc27bedb903c39b --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/latest @@ -0,0 +1 @@ +global_step1300 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..61e7f48405e546beb934b2717fa81816f64df078 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/trainer_state.json @@ -0,0 +1,10173 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.4246575342465753, + "eval_steps": 10, + "global_step": 1300, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6624906063079834, + "learning_rate": 4.77051177904143e-06, + "loss": 1.704, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 3.0075268745422363, + "learning_rate": 4.768480909829407e-06, + "loss": 1.6322, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.6070828437805176, + "learning_rate": 4.766450040617385e-06, + "loss": 1.6871, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 3.044222831726074, + "learning_rate": 4.764419171405362e-06, + "loss": 1.6652, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.773167610168457, + "learning_rate": 4.762388302193339e-06, + "loss": 1.7063, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.880176305770874, + "learning_rate": 4.760357432981316e-06, + "loss": 1.6775, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.6402764320373535, + "learning_rate": 4.758326563769294e-06, + "loss": 1.6694, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.9245846271514893, + "learning_rate": 4.756295694557271e-06, + "loss": 1.6688, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.623044967651367, + "learning_rate": 4.754264825345248e-06, + "loss": 1.6819, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.739896059036255, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.6401, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.8707164525985718, + "eval_runtime": 36.9513, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.603599786758423, + "learning_rate": 4.750203086921203e-06, + "loss": 1.6412, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.5571675300598145, + "learning_rate": 4.74817221770918e-06, + "loss": 1.5779, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.87091326713562, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.6878, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.8140673637390137, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7023, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.6550583839416504, + "learning_rate": 4.742079610073111e-06, + "loss": 1.5995, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.6259925365448, + "learning_rate": 4.740048740861089e-06, + "loss": 1.6194, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.971848726272583, + "learning_rate": 4.738017871649066e-06, + "loss": 1.6634, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6344778537750244, + "learning_rate": 4.735987002437044e-06, + "loss": 1.6502, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 2.9362456798553467, + "learning_rate": 4.73395613322502e-06, + "loss": 1.6948, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.8098835945129395, + "learning_rate": 4.731925264012998e-06, + "loss": 1.7161, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.8724393844604492, + "eval_runtime": 37.0876, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.933009386062622, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7014, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.8216991424560547, + "learning_rate": 4.727863525588953e-06, + "loss": 1.6192, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.7028493881225586, + "learning_rate": 4.725832656376929e-06, + "loss": 1.607, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.816818952560425, + "learning_rate": 4.723801787164907e-06, + "loss": 1.6183, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.7707865238189697, + "learning_rate": 4.721770917952885e-06, + "loss": 1.6795, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.779562473297119, + "learning_rate": 4.719740048740861e-06, + "loss": 1.63, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.907338857650757, + "learning_rate": 4.717709179528839e-06, + "loss": 1.644, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 3.073558807373047, + "learning_rate": 4.715678310316816e-06, + "loss": 1.6821, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.7556493282318115, + "learning_rate": 4.713647441104794e-06, + "loss": 1.6501, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6718344688415527, + "learning_rate": 4.71161657189277e-06, + "loss": 1.6225, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.854687213897705, + "eval_runtime": 36.9773, + "eval_samples_per_second": 27.044, + "eval_steps_per_second": 1.136, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.763335943222046, + "learning_rate": 4.709585702680748e-06, + "loss": 1.6311, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.7669479846954346, + "learning_rate": 4.707554833468725e-06, + "loss": 1.6809, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.711562395095825, + "learning_rate": 4.705523964256703e-06, + "loss": 1.741, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.799954652786255, + "learning_rate": 4.703493095044679e-06, + "loss": 1.6409, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.7278690338134766, + "learning_rate": 4.701462225832657e-06, + "loss": 1.6325, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.6802427768707275, + "learning_rate": 4.699431356620634e-06, + "loss": 1.5988, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.8826160430908203, + "learning_rate": 4.697400487408612e-06, + "loss": 1.7119, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.784343719482422, + "learning_rate": 4.695369618196588e-06, + "loss": 1.6855, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.683337688446045, + "learning_rate": 4.693338748984566e-06, + "loss": 1.6588, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.6646857261657715, + "learning_rate": 4.691307879772543e-06, + "loss": 1.623, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.8511513471603394, + "eval_runtime": 36.9852, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.8181183338165283, + "learning_rate": 4.68927701056052e-06, + "loss": 1.6171, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.8270857334136963, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.678, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.7685446739196777, + "learning_rate": 4.685215272136475e-06, + "loss": 1.6575, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.7003183364868164, + "learning_rate": 4.683184402924452e-06, + "loss": 1.6405, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.668750286102295, + "learning_rate": 4.681153533712429e-06, + "loss": 1.6541, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 3.065800905227661, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7288, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.615601062774658, + "learning_rate": 4.677091795288384e-06, + "loss": 1.6306, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.5948891639709473, + "learning_rate": 4.675060926076361e-06, + "loss": 1.6018, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.7550079822540283, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7009, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.868687868118286, + "learning_rate": 4.670999187652316e-06, + "loss": 1.6034, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.8473671674728394, + "eval_runtime": 37.0015, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.705962657928467, + "learning_rate": 4.668968318440293e-06, + "loss": 1.6084, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.7901813983917236, + "learning_rate": 4.66693744922827e-06, + "loss": 1.6171, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.688774824142456, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.6774, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.734617233276367, + "learning_rate": 4.662875710804225e-06, + "loss": 1.5796, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.742017984390259, + "learning_rate": 4.660844841592202e-06, + "loss": 1.6671, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.6438794136047363, + "learning_rate": 4.658813972380179e-06, + "loss": 1.5835, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6441051959991455, + "learning_rate": 4.656783103168157e-06, + "loss": 1.6909, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.824204206466675, + "learning_rate": 4.654752233956134e-06, + "loss": 1.5758, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.556638479232788, + "learning_rate": 4.652721364744111e-06, + "loss": 1.6504, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.897812843322754, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.6406, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.8537043333053589, + "eval_runtime": 37.257, + "eval_samples_per_second": 26.841, + "eval_steps_per_second": 1.127, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.6594088077545166, + "learning_rate": 4.648659626320066e-06, + "loss": 1.6966, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.79524302482605, + "learning_rate": 4.646628757108043e-06, + "loss": 1.6645, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.80291748046875, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.6281, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.7723848819732666, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.5959, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 3.0625650882720947, + "learning_rate": 4.640536149471974e-06, + "loss": 1.6313, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.775726079940796, + "learning_rate": 4.638505280259952e-06, + "loss": 1.6563, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.837758779525757, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.6727, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 3.041355609893799, + "learning_rate": 4.634443541835907e-06, + "loss": 1.6282, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.5775389671325684, + "learning_rate": 4.632412672623883e-06, + "loss": 1.651, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 3.026287794113159, + "learning_rate": 4.630381803411861e-06, + "loss": 1.6458, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.8461565971374512, + "eval_runtime": 37.3881, + "eval_samples_per_second": 26.747, + "eval_steps_per_second": 1.123, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.6781764030456543, + "learning_rate": 4.628350934199838e-06, + "loss": 1.5232, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.763350486755371, + "learning_rate": 4.626320064987816e-06, + "loss": 1.6186, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.7063372135162354, + "learning_rate": 4.624289195775792e-06, + "loss": 1.6784, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.824563980102539, + "learning_rate": 4.62225832656377e-06, + "loss": 1.6219, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.5335569381713867, + "learning_rate": 4.620227457351747e-06, + "loss": 1.6152, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.773644208908081, + "learning_rate": 4.618196588139724e-06, + "loss": 1.6179, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.9841721057891846, + "learning_rate": 4.616165718927701e-06, + "loss": 1.6258, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.684171676635742, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.6066, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.8817965984344482, + "learning_rate": 4.612103980503656e-06, + "loss": 1.6697, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.9846866130828857, + "learning_rate": 4.610073111291633e-06, + "loss": 1.6541, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.8485002517700195, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.904170513153076, + "learning_rate": 4.60804224207961e-06, + "loss": 1.6362, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.684476375579834, + "learning_rate": 4.606011372867588e-06, + "loss": 1.6303, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6692216396331787, + "learning_rate": 4.603980503655565e-06, + "loss": 1.638, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.73555588722229, + "learning_rate": 4.601949634443542e-06, + "loss": 1.6125, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.729525089263916, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.565, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.5972018241882324, + "learning_rate": 4.597887896019497e-06, + "loss": 1.5856, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7859137058258057, + "learning_rate": 4.595857026807474e-06, + "loss": 1.6325, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.662587881088257, + "learning_rate": 4.593826157595451e-06, + "loss": 1.6172, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.6380577087402344, + "learning_rate": 4.591795288383429e-06, + "loss": 1.6005, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8375349044799805, + "learning_rate": 4.589764419171406e-06, + "loss": 1.6507, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.833074688911438, + "eval_runtime": 37.0695, + "eval_samples_per_second": 26.976, + "eval_steps_per_second": 1.133, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.786713123321533, + "learning_rate": 4.587733549959383e-06, + "loss": 1.6283, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.6959030628204346, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.5994, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.6100692749023438, + "learning_rate": 4.583671811535338e-06, + "loss": 1.5884, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.7281863689422607, + "learning_rate": 4.581640942323315e-06, + "loss": 1.5737, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.7519376277923584, + "learning_rate": 4.579610073111292e-06, + "loss": 1.6531, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.9314866065979004, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.6755, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.9250576496124268, + "learning_rate": 4.575548334687247e-06, + "loss": 1.6604, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.6382815837860107, + "learning_rate": 4.573517465475224e-06, + "loss": 1.6773, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.6891751289367676, + "learning_rate": 4.571486596263201e-06, + "loss": 1.5849, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.7018136978149414, + "learning_rate": 4.569455727051178e-06, + "loss": 1.6358, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.838019847869873, + "eval_runtime": 37.0496, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.671952247619629, + "learning_rate": 4.567424857839156e-06, + "loss": 1.6207, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.882539749145508, + "learning_rate": 4.565393988627132e-06, + "loss": 1.6618, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.684070587158203, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.6502, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.5295751094818115, + "learning_rate": 4.561332250203087e-06, + "loss": 1.6078, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.0268194675445557, + "learning_rate": 4.559301380991065e-06, + "loss": 1.6312, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6895201206207275, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.5848, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.869950771331787, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.5951, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.909505605697632, + "learning_rate": 4.553208773354996e-06, + "loss": 1.6589, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.7604334354400635, + "learning_rate": 4.551177904142974e-06, + "loss": 1.5767, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.8792080879211426, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.6864, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.836847186088562, + "eval_runtime": 36.9142, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.607037305831909, + "learning_rate": 4.547116165718929e-06, + "loss": 1.5925, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.9338927268981934, + "learning_rate": 4.545085296506905e-06, + "loss": 1.6433, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.7723233699798584, + "learning_rate": 4.543054427294882e-06, + "loss": 1.5963, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.4122188091278076, + "learning_rate": 4.54102355808286e-06, + "loss": 1.5978, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.7582709789276123, + "learning_rate": 4.538992688870837e-06, + "loss": 1.581, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.6983039379119873, + "learning_rate": 4.536961819658814e-06, + "loss": 1.6253, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.6036365032196045, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.5957, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.759645700454712, + "learning_rate": 4.532900081234769e-06, + "loss": 1.6199, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.797744035720825, + "learning_rate": 4.530869212022746e-06, + "loss": 1.5854, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.752868890762329, + "learning_rate": 4.528838342810723e-06, + "loss": 1.619, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.8305871486663818, + "eval_runtime": 37.0307, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.7421889305114746, + "learning_rate": 4.526807473598701e-06, + "loss": 1.6182, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.6852707862854004, + "learning_rate": 4.524776604386678e-06, + "loss": 1.6135, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.861720323562622, + "learning_rate": 4.522745735174655e-06, + "loss": 1.6099, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.706851005554199, + "learning_rate": 4.520714865962632e-06, + "loss": 1.6597, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7992191314697266, + "learning_rate": 4.51868399675061e-06, + "loss": 1.6077, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.6248183250427246, + "learning_rate": 4.516653127538587e-06, + "loss": 1.6485, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.696040391921997, + "learning_rate": 4.514622258326564e-06, + "loss": 1.5853, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.5958261489868164, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.571, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.9715874195098877, + "learning_rate": 4.510560519902519e-06, + "loss": 1.6044, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.664834499359131, + "learning_rate": 4.508529650690496e-06, + "loss": 1.6568, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.8277279138565063, + "eval_runtime": 37.0489, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.869558811187744, + "learning_rate": 4.506498781478473e-06, + "loss": 1.5654, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.5814828872680664, + "learning_rate": 4.504467912266451e-06, + "loss": 1.5484, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.6158947944641113, + "learning_rate": 4.502437043054428e-06, + "loss": 1.6202, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.7424654960632324, + "learning_rate": 4.500406173842405e-06, + "loss": 1.6755, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.8670949935913086, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.6144, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.7914652824401855, + "learning_rate": 4.49634443541836e-06, + "loss": 1.6164, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.866380214691162, + "learning_rate": 4.494313566206336e-06, + "loss": 1.5961, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.966474771499634, + "learning_rate": 4.492282696994314e-06, + "loss": 1.6681, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.751291513442993, + "learning_rate": 4.490251827782291e-06, + "loss": 1.6338, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.58840012550354, + "learning_rate": 4.488220958570269e-06, + "loss": 1.5619, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.824889898300171, + "eval_runtime": 37.0821, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.786945104598999, + "learning_rate": 4.486190089358245e-06, + "loss": 1.5712, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4536845684051514, + "learning_rate": 4.484159220146223e-06, + "loss": 1.5487, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.780278205871582, + "learning_rate": 4.4821283509342e-06, + "loss": 1.6269, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.5427017211914062, + "learning_rate": 4.480097481722178e-06, + "loss": 1.6254, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.7540087699890137, + "learning_rate": 4.478066612510154e-06, + "loss": 1.6112, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.616352081298828, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.5637, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.686035633087158, + "learning_rate": 4.474004874086109e-06, + "loss": 1.5974, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.982365846633911, + "learning_rate": 4.471974004874087e-06, + "loss": 1.7034, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.6106085777282715, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.5482, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.7798519134521484, + "learning_rate": 4.467912266450041e-06, + "loss": 1.6475, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.820489764213562, + "eval_runtime": 37.0672, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.8634185791015625, + "learning_rate": 4.465881397238018e-06, + "loss": 1.6258, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.633626699447632, + "learning_rate": 4.463850528025995e-06, + "loss": 1.5806, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.719074249267578, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.6404, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.7545487880706787, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6404, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.6559669971466064, + "learning_rate": 4.457757920389927e-06, + "loss": 1.6104, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.9004898071289062, + "learning_rate": 4.455727051177904e-06, + "loss": 1.5513, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.6640021800994873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.5859, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.842026472091675, + "learning_rate": 4.451665312753859e-06, + "loss": 1.6241, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.784647226333618, + "learning_rate": 4.449634443541836e-06, + "loss": 1.627, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.7067618370056152, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.6276, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.8115602731704712, + "eval_runtime": 37.078, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.6582882404327393, + "learning_rate": 4.445572705117791e-06, + "loss": 1.5746, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7881102561950684, + "learning_rate": 4.443541835905768e-06, + "loss": 1.5749, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.9323439598083496, + "learning_rate": 4.441510966693745e-06, + "loss": 1.6421, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.63948655128479, + "learning_rate": 4.439480097481723e-06, + "loss": 1.5853, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.7746710777282715, + "learning_rate": 4.4374492282697e-06, + "loss": 1.6455, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.093895435333252, + "learning_rate": 4.435418359057677e-06, + "loss": 1.5311, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.526385545730591, + "learning_rate": 4.433387489845654e-06, + "loss": 1.6082, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.6604297161102295, + "learning_rate": 4.431356620633632e-06, + "loss": 1.61, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.6636247634887695, + "learning_rate": 4.429325751421609e-06, + "loss": 1.5815, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.5567514896392822, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6086, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.8170957565307617, + "eval_runtime": 37.4095, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.695563793182373, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.5836, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6009771823883057, + "learning_rate": 4.423233143785541e-06, + "loss": 1.5264, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.0490736961364746, + "learning_rate": 4.421202274573518e-06, + "loss": 1.5982, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.8609886169433594, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6265, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4415369033813477, + "learning_rate": 4.417140536149473e-06, + "loss": 1.4987, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.771657943725586, + "learning_rate": 4.415109666937449e-06, + "loss": 1.5494, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.739617347717285, + "learning_rate": 4.413078797725427e-06, + "loss": 1.6274, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.805098295211792, + "learning_rate": 4.411047928513404e-06, + "loss": 1.5686, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.7248735427856445, + "learning_rate": 4.409017059301382e-06, + "loss": 1.6027, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6653318405151367, + "learning_rate": 4.406986190089358e-06, + "loss": 1.5909, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.8076283931732178, + "eval_runtime": 37.5073, + "eval_samples_per_second": 26.661, + "eval_steps_per_second": 1.12, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.626783847808838, + "learning_rate": 4.404955320877336e-06, + "loss": 1.6205, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.6929268836975098, + "learning_rate": 4.402924451665313e-06, + "loss": 1.6146, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.6806800365448, + "learning_rate": 4.400893582453291e-06, + "loss": 1.5702, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.5407259464263916, + "learning_rate": 4.398862713241267e-06, + "loss": 1.5525, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.639122247695923, + "learning_rate": 4.396831844029245e-06, + "loss": 1.5826, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.669396162033081, + "learning_rate": 4.394800974817222e-06, + "loss": 1.5982, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.6377906799316406, + "learning_rate": 4.392770105605199e-06, + "loss": 1.5797, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6769638061523438, + "learning_rate": 4.390739236393176e-06, + "loss": 1.607, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.6347222328186035, + "learning_rate": 4.388708367181154e-06, + "loss": 1.6196, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.670680284500122, + "learning_rate": 4.386677497969131e-06, + "loss": 1.619, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.8124384880065918, + "eval_runtime": 37.4751, + "eval_samples_per_second": 26.684, + "eval_steps_per_second": 1.121, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.5886521339416504, + "learning_rate": 4.384646628757108e-06, + "loss": 1.6258, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.644566297531128, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6447, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.6432394981384277, + "learning_rate": 4.380584890333063e-06, + "loss": 1.5888, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.5602784156799316, + "learning_rate": 4.37855402112104e-06, + "loss": 1.5544, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.4874439239501953, + "learning_rate": 4.376523151909017e-06, + "loss": 1.5075, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.694748640060425, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.6498, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.582211494445801, + "learning_rate": 4.372461413484972e-06, + "loss": 1.5877, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.675713539123535, + "learning_rate": 4.370430544272949e-06, + "loss": 1.6045, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.7157161235809326, + "learning_rate": 4.368399675060926e-06, + "loss": 1.623, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.795076370239258, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6141, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.8062744140625, + "eval_runtime": 37.2435, + "eval_samples_per_second": 26.85, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6296374797821045, + "learning_rate": 4.364337936636881e-06, + "loss": 1.6087, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.362307067424858e-06, + "loss": 1.5182, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.7965915203094482, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.6025, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.807307481765747, + "learning_rate": 4.358245329000813e-06, + "loss": 1.574, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.584615468978882, + "learning_rate": 4.35621445978879e-06, + "loss": 1.5922, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.8472282886505127, + "learning_rate": 4.354183590576767e-06, + "loss": 1.5914, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.8471736907958984, + "learning_rate": 4.352152721364745e-06, + "loss": 1.5749, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.569225549697876, + "learning_rate": 4.350121852152722e-06, + "loss": 1.5591, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.688157796859741, + "learning_rate": 4.348090982940699e-06, + "loss": 1.6042, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.5865259170532227, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.6307, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.8040707111358643, + "eval_runtime": 36.9511, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.71850323677063, + "learning_rate": 4.344029244516654e-06, + "loss": 1.6043, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.7158093452453613, + "learning_rate": 4.341998375304631e-06, + "loss": 1.5907, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6879327297210693, + "learning_rate": 4.339967506092607e-06, + "loss": 1.6365, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.827277660369873, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.5928, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6709842681884766, + "learning_rate": 4.335905767668562e-06, + "loss": 1.5812, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.7071495056152344, + "learning_rate": 4.33387489845654e-06, + "loss": 1.5571, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6008732318878174, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.6048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.6238059997558594, + "learning_rate": 4.329813160032495e-06, + "loss": 1.5962, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.567065954208374, + "learning_rate": 4.327782290820471e-06, + "loss": 1.5713, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.573068857192993, + "learning_rate": 4.325751421608449e-06, + "loss": 1.536, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.8103471994400024, + "eval_runtime": 37.0956, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6245224475860596, + "learning_rate": 4.323720552396426e-06, + "loss": 1.5986, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.6243810653686523, + "learning_rate": 4.321689683184404e-06, + "loss": 1.589, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.5303313732147217, + "learning_rate": 4.31965881397238e-06, + "loss": 1.5793, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.5617806911468506, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.5704, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.7429656982421875, + "learning_rate": 4.315597075548335e-06, + "loss": 1.6174, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.6753251552581787, + "learning_rate": 4.313566206336312e-06, + "loss": 1.5879, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.5944488048553467, + "learning_rate": 4.311535337124289e-06, + "loss": 1.6082, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.9002904891967773, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.5521, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.5384247303009033, + "learning_rate": 4.307473598700244e-06, + "loss": 1.5751, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6015477180480957, + "learning_rate": 4.305442729488221e-06, + "loss": 1.5966, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8091154098510742, + "eval_runtime": 37.1075, + "eval_samples_per_second": 26.949, + "eval_steps_per_second": 1.132, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.6002886295318604, + "learning_rate": 4.303411860276198e-06, + "loss": 1.5403, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7755627632141113, + "learning_rate": 4.301380991064176e-06, + "loss": 1.6192, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.8351540565490723, + "learning_rate": 4.299350121852153e-06, + "loss": 1.6217, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.684007406234741, + "learning_rate": 4.29731925264013e-06, + "loss": 1.6242, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.8019614219665527, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6173, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.9122867584228516, + "learning_rate": 4.293257514216085e-06, + "loss": 1.5701, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.545732021331787, + "learning_rate": 4.291226645004062e-06, + "loss": 1.5494, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.9499566555023193, + "learning_rate": 4.289195775792039e-06, + "loss": 1.5848, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.8071892261505127, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6114, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.763864517211914, + "learning_rate": 4.285134037367994e-06, + "loss": 1.5768, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8057912588119507, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 3.001190423965454, + "learning_rate": 4.283103168155971e-06, + "loss": 1.5601, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.7640297412872314, + "learning_rate": 4.281072298943948e-06, + "loss": 1.5619, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.7372794151306152, + "learning_rate": 4.279041429731926e-06, + "loss": 1.562, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 3.0390751361846924, + "learning_rate": 4.277010560519903e-06, + "loss": 1.5519, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.758798599243164, + "learning_rate": 4.27497969130788e-06, + "loss": 1.527, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.96443772315979, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.5899, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.7941088676452637, + "learning_rate": 4.270917952883835e-06, + "loss": 1.617, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 3.0385100841522217, + "learning_rate": 4.268887083671812e-06, + "loss": 1.5996, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.6946637630462646, + "learning_rate": 4.266856214459789e-06, + "loss": 1.5641, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.776723861694336, + "learning_rate": 4.264825345247766e-06, + "loss": 1.6451, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.7911298274993896, + "eval_runtime": 37.0623, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.851224184036255, + "learning_rate": 4.262794476035744e-06, + "loss": 1.5302, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9045867919921875, + "learning_rate": 4.26076360682372e-06, + "loss": 1.5808, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.6568782329559326, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.5233, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.94191837310791, + "learning_rate": 4.256701868399675e-06, + "loss": 1.5912, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.726778984069824, + "learning_rate": 4.254670999187653e-06, + "loss": 1.5823, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.8704583644866943, + "learning_rate": 4.252640129975629e-06, + "loss": 1.5931, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.5721547603607178, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6128, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6883509159088135, + "learning_rate": 4.248578391551584e-06, + "loss": 1.5479, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.7186272144317627, + "learning_rate": 4.246547522339562e-06, + "loss": 1.5203, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.8326518535614014, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.5547, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.7948570251464844, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.7870523929595947, + "learning_rate": 4.242485783915517e-06, + "loss": 1.5891, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.866610288619995, + "learning_rate": 4.240454914703493e-06, + "loss": 1.5232, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.94161057472229, + "learning_rate": 4.23842404549147e-06, + "loss": 1.52, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.7331979274749756, + "learning_rate": 4.236393176279448e-06, + "loss": 1.5665, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.5375845432281494, + "learning_rate": 4.234362307067425e-06, + "loss": 1.5274, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.691927433013916, + "learning_rate": 4.232331437855402e-06, + "loss": 1.604, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7578394412994385, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.6, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.594538688659668, + "learning_rate": 4.228269699431357e-06, + "loss": 1.5928, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.5408968925476074, + "learning_rate": 4.226238830219334e-06, + "loss": 1.5908, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.689786672592163, + "learning_rate": 4.224207961007311e-06, + "loss": 1.5757, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.792494297027588, + "eval_runtime": 37.0822, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.735539197921753, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.5811, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.58634614944458, + "learning_rate": 4.220146222583266e-06, + "loss": 1.5808, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.7899138927459717, + "learning_rate": 4.218115353371243e-06, + "loss": 1.5801, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.6507744789123535, + "learning_rate": 4.21608448415922e-06, + "loss": 1.5585, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.6112217903137207, + "learning_rate": 4.214053614947198e-06, + "loss": 1.5329, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.729405164718628, + "learning_rate": 4.212022745735175e-06, + "loss": 1.6177, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.6431992053985596, + "learning_rate": 4.209991876523152e-06, + "loss": 1.5234, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.726243734359741, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.5926, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.6574180126190186, + "learning_rate": 4.205930138099107e-06, + "loss": 1.5424, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.6570956707000732, + "learning_rate": 4.203899268887084e-06, + "loss": 1.6119, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.781999945640564, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.528052806854248, + "learning_rate": 4.201868399675061e-06, + "loss": 1.598, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.6075727939605713, + "learning_rate": 4.199837530463039e-06, + "loss": 1.5886, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.588737964630127, + "learning_rate": 4.197806661251016e-06, + "loss": 1.507, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.773618221282959, + "learning_rate": 4.195775792038993e-06, + "loss": 1.5794, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.624725580215454, + "learning_rate": 4.19374492282697e-06, + "loss": 1.5495, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.662397861480713, + "learning_rate": 4.191714053614948e-06, + "loss": 1.5203, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.6617255210876465, + "learning_rate": 4.189683184402924e-06, + "loss": 1.5885, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.5784451961517334, + "learning_rate": 4.187652315190902e-06, + "loss": 1.5689, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.726011037826538, + "learning_rate": 4.185621445978879e-06, + "loss": 1.5913, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.6117324829101562, + "learning_rate": 4.183590576766857e-06, + "loss": 1.5453, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.7893352508544922, + "eval_runtime": 37.4096, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7025442123413086, + "learning_rate": 4.181559707554833e-06, + "loss": 1.5868, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.656102180480957, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6141, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.741962194442749, + "learning_rate": 4.177497969130788e-06, + "loss": 1.5516, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5887229442596436, + "learning_rate": 4.175467099918766e-06, + "loss": 1.5525, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.542985439300537, + "learning_rate": 4.173436230706742e-06, + "loss": 1.592, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.7797129154205322, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.5349, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.632887363433838, + "learning_rate": 4.169374492282697e-06, + "loss": 1.5548, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.691690444946289, + "learning_rate": 4.167343623070675e-06, + "loss": 1.5352, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.8922817707061768, + "learning_rate": 4.165312753858652e-06, + "loss": 1.5717, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.5517446994781494, + "learning_rate": 4.163281884646629e-06, + "loss": 1.5658, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.780922770500183, + "eval_runtime": 37.3755, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.718374490737915, + "learning_rate": 4.161251015434607e-06, + "loss": 1.5578, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.6249442100524902, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6018, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.7051141262054443, + "learning_rate": 4.157189277010561e-06, + "loss": 1.5182, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.677109956741333, + "learning_rate": 4.155158407798538e-06, + "loss": 1.5423, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.591059684753418, + "learning_rate": 4.153127538586516e-06, + "loss": 1.5694, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.699953556060791, + "learning_rate": 4.151096669374492e-06, + "loss": 1.5758, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.8837342262268066, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.5509, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.6566851139068604, + "learning_rate": 4.147034930950447e-06, + "loss": 1.5577, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.901057004928589, + "learning_rate": 4.145004061738425e-06, + "loss": 1.5797, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.796971321105957, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.5608, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.7797813415527344, + "eval_runtime": 36.8425, + "eval_samples_per_second": 27.143, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.605431318283081, + "learning_rate": 4.140942323314379e-06, + "loss": 1.491, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8084020614624023, + "learning_rate": 4.138911454102356e-06, + "loss": 1.5074, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.625511407852173, + "learning_rate": 4.136880584890333e-06, + "loss": 1.5404, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.6944286823272705, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.589, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.81792950630188, + "learning_rate": 4.132818846466288e-06, + "loss": 1.5646, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5813655853271484, + "learning_rate": 4.130787977254265e-06, + "loss": 1.5697, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.6733286380767822, + "learning_rate": 4.128757108042242e-06, + "loss": 1.5786, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.6972577571868896, + "learning_rate": 4.12672623883022e-06, + "loss": 1.5857, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7381093502044678, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6365, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.6162049770355225, + "learning_rate": 4.122664500406174e-06, + "loss": 1.5598, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.7869179248809814, + "eval_runtime": 36.974, + "eval_samples_per_second": 27.046, + "eval_steps_per_second": 1.136, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.7157270908355713, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.5533, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.560105085372925, + "learning_rate": 4.118602761982129e-06, + "loss": 1.5821, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.5261518955230713, + "learning_rate": 4.116571892770106e-06, + "loss": 1.5195, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7377054691314697, + "learning_rate": 4.114541023558083e-06, + "loss": 1.5066, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.632946252822876, + "learning_rate": 4.112510154346061e-06, + "loss": 1.5948, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.809671640396118, + "learning_rate": 4.110479285134038e-06, + "loss": 1.5435, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.843156099319458, + "learning_rate": 4.108448415922015e-06, + "loss": 1.5667, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.743614673614502, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5958, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.9596827030181885, + "learning_rate": 4.10438667749797e-06, + "loss": 1.5807, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.7321834564208984, + "learning_rate": 4.102355808285947e-06, + "loss": 1.4744, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.784859538078308, + "eval_runtime": 37.0146, + "eval_samples_per_second": 27.016, + "eval_steps_per_second": 1.135, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.010359764099121, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6146, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.744251251220703, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.5401, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.592690944671631, + "learning_rate": 4.096263200649879e-06, + "loss": 1.5508, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.786736011505127, + "learning_rate": 4.094232331437856e-06, + "loss": 1.5717, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.8773610591888428, + "learning_rate": 4.092201462225833e-06, + "loss": 1.5201, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.818610906600952, + "learning_rate": 4.090170593013811e-06, + "loss": 1.5282, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.9548542499542236, + "learning_rate": 4.088139723801787e-06, + "loss": 1.553, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.678842306137085, + "learning_rate": 4.086108854589765e-06, + "loss": 1.531, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.7757408618927, + "learning_rate": 4.084077985377742e-06, + "loss": 1.5371, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.6954638957977295, + "learning_rate": 4.08204711616572e-06, + "loss": 1.5462, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.781003475189209, + "eval_runtime": 37.0687, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6897518634796143, + "learning_rate": 4.080016246953696e-06, + "loss": 1.5979, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.581172466278076, + "learning_rate": 4.077985377741674e-06, + "loss": 1.5007, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.828234910964966, + "learning_rate": 4.075954508529651e-06, + "loss": 1.5857, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.6171951293945312, + "learning_rate": 4.073923639317629e-06, + "loss": 1.5937, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.6476316452026367, + "learning_rate": 4.071892770105605e-06, + "loss": 1.5864, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.7525634765625, + "learning_rate": 4.069861900893583e-06, + "loss": 1.5095, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.6881322860717773, + "learning_rate": 4.06783103168156e-06, + "loss": 1.4936, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.6104350090026855, + "learning_rate": 4.065800162469538e-06, + "loss": 1.5229, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.866706132888794, + "learning_rate": 4.063769293257514e-06, + "loss": 1.539, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.803999900817871, + "learning_rate": 4.061738424045492e-06, + "loss": 1.5792, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.7793465852737427, + "eval_runtime": 37.0749, + "eval_samples_per_second": 26.972, + "eval_steps_per_second": 1.133, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.670285940170288, + "learning_rate": 4.059707554833469e-06, + "loss": 1.5523, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.871267557144165, + "learning_rate": 4.057676685621446e-06, + "loss": 1.5623, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.6631205081939697, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6094, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.7580034732818604, + "learning_rate": 4.053614947197401e-06, + "loss": 1.5725, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.6412739753723145, + "learning_rate": 4.051584077985378e-06, + "loss": 1.4959, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.699176073074341, + "learning_rate": 4.049553208773355e-06, + "loss": 1.5586, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.622969150543213, + "learning_rate": 4.047522339561333e-06, + "loss": 1.5713, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.892338991165161, + "learning_rate": 4.04549147034931e-06, + "loss": 1.5471, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.77056622505188, + "learning_rate": 4.043460601137287e-06, + "loss": 1.5236, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.7122550010681152, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6562, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.7708193063735962, + "eval_runtime": 37.0743, + "eval_samples_per_second": 26.973, + "eval_steps_per_second": 1.133, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.702824592590332, + "learning_rate": 4.039398862713242e-06, + "loss": 1.534, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.6822307109832764, + "learning_rate": 4.037367993501219e-06, + "loss": 1.5368, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.7274515628814697, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6218, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.8114798069000244, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.532, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.8936192989349365, + "learning_rate": 4.031275385865151e-06, + "loss": 1.5124, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.7851672172546387, + "learning_rate": 4.029244516653128e-06, + "loss": 1.5596, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.815023183822632, + "learning_rate": 4.027213647441105e-06, + "loss": 1.5706, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5958921909332275, + "learning_rate": 4.025182778229083e-06, + "loss": 1.5161, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.779360771179199, + "learning_rate": 4.02315190901706e-06, + "loss": 1.5057, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.7411787509918213, + "learning_rate": 4.021121039805037e-06, + "loss": 1.5396, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.7778708934783936, + "eval_runtime": 37.0791, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.5709781646728516, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.5474, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8583388328552246, + "learning_rate": 4.017059301380992e-06, + "loss": 1.6149, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.6877572536468506, + "learning_rate": 4.015028432168969e-06, + "loss": 1.5292, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.680612087249756, + "learning_rate": 4.012997562956945e-06, + "loss": 1.596, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.653559446334839, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.4901, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.542787790298462, + "learning_rate": 4.0089358245329e-06, + "loss": 1.5295, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5476601123809814, + "learning_rate": 4.006904955320878e-06, + "loss": 1.5393, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.6367862224578857, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.5321, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.8112173080444336, + "learning_rate": 4.002843216896833e-06, + "loss": 1.5963, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.7136871814727783, + "learning_rate": 4.000812347684809e-06, + "loss": 1.5839, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.7763899564743042, + "eval_runtime": 37.2537, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.9095280170440674, + "learning_rate": 3.998781478472787e-06, + "loss": 1.5343, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.6196115016937256, + "learning_rate": 3.996750609260764e-06, + "loss": 1.5411, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6413371562957764, + "learning_rate": 3.994719740048742e-06, + "loss": 1.4862, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 3.0759036540985107, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6471, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.599147081375122, + "learning_rate": 3.990658001624696e-06, + "loss": 1.5556, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.726978302001953, + "learning_rate": 3.988627132412673e-06, + "loss": 1.5351, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.6760807037353516, + "learning_rate": 3.98659626320065e-06, + "loss": 1.5681, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.680483818054199, + "learning_rate": 3.984565393988627e-06, + "loss": 1.5225, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6649010181427, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.5325, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.638085126876831, + "learning_rate": 3.980503655564582e-06, + "loss": 1.564, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.7698904275894165, + "eval_runtime": 37.4492, + "eval_samples_per_second": 26.703, + "eval_steps_per_second": 1.122, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.647825241088867, + "learning_rate": 3.978472786352559e-06, + "loss": 1.5782, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.748222589492798, + "learning_rate": 3.976441917140536e-06, + "loss": 1.556, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.6319026947021484, + "learning_rate": 3.974411047928514e-06, + "loss": 1.5378, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.6759872436523438, + "learning_rate": 3.972380178716491e-06, + "loss": 1.6018, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.6760940551757812, + "learning_rate": 3.970349309504468e-06, + "loss": 1.5742, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.5565083026885986, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.565, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.5986931324005127, + "learning_rate": 3.966287571080423e-06, + "loss": 1.5124, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.713092803955078, + "learning_rate": 3.9642567018684e-06, + "loss": 1.523, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.6203417778015137, + "learning_rate": 3.962225832656377e-06, + "loss": 1.5601, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.604440927505493, + "learning_rate": 3.960194963444355e-06, + "loss": 1.5485, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.7633332014083862, + "eval_runtime": 37.49, + "eval_samples_per_second": 26.674, + "eval_steps_per_second": 1.12, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.5513081550598145, + "learning_rate": 3.958164094232332e-06, + "loss": 1.479, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.7990963459014893, + "learning_rate": 3.956133225020309e-06, + "loss": 1.5355, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.651113271713257, + "learning_rate": 3.954102355808286e-06, + "loss": 1.5406, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.7236671447753906, + "learning_rate": 3.952071486596264e-06, + "loss": 1.5314, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.697598695755005, + "learning_rate": 3.950040617384241e-06, + "loss": 1.5398, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.8271944522857666, + "learning_rate": 3.948009748172218e-06, + "loss": 1.54, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.733125925064087, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.5236, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.7555017471313477, + "learning_rate": 3.943948009748173e-06, + "loss": 1.5808, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.6592960357666016, + "learning_rate": 3.94191714053615e-06, + "loss": 1.5398, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.6870126724243164, + "learning_rate": 3.939886271324127e-06, + "loss": 1.5517, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.760675311088562, + "eval_runtime": 36.9372, + "eval_samples_per_second": 27.073, + "eval_steps_per_second": 1.137, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.7469022274017334, + "learning_rate": 3.937855402112104e-06, + "loss": 1.605, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.5255792140960693, + "learning_rate": 3.935824532900082e-06, + "loss": 1.5061, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.589054822921753, + "learning_rate": 3.933793663688058e-06, + "loss": 1.4971, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.7874889373779297, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.5643, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.819451093673706, + "learning_rate": 3.929731925264013e-06, + "loss": 1.5797, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.6516339778900146, + "learning_rate": 3.927701056051991e-06, + "loss": 1.5048, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.979091167449951, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.5174, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.743058204650879, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.5432, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.6864030361175537, + "learning_rate": 3.921608448415922e-06, + "loss": 1.5147, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.7751128673553467, + "learning_rate": 3.9195775792039e-06, + "loss": 1.4857, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.7609487771987915, + "eval_runtime": 37.09, + "eval_samples_per_second": 26.961, + "eval_steps_per_second": 1.132, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.626188278198242, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.5428, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.554067850112915, + "learning_rate": 3.915515840779855e-06, + "loss": 1.5153, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.8580875396728516, + "learning_rate": 3.913484971567831e-06, + "loss": 1.5562, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.7405097484588623, + "learning_rate": 3.911454102355808e-06, + "loss": 1.565, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.5650815963745117, + "learning_rate": 3.909423233143786e-06, + "loss": 1.4756, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.9405910968780518, + "learning_rate": 3.907392363931763e-06, + "loss": 1.5611, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.687448024749756, + "learning_rate": 3.90536149471974e-06, + "loss": 1.5405, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.7373857498168945, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.6055, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.875298261642456, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6215, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.713200330734253, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6493, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.7608598470687866, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.541935682296753, + "learning_rate": 3.897238017871649e-06, + "loss": 1.5257, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.582045316696167, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.5253, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.6884682178497314, + "learning_rate": 3.893176279447604e-06, + "loss": 1.5329, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.6456501483917236, + "learning_rate": 3.891145410235581e-06, + "loss": 1.5167, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.60788893699646, + "learning_rate": 3.889114541023558e-06, + "loss": 1.5945, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.708646774291992, + "learning_rate": 3.887083671811536e-06, + "loss": 1.5704, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.666922092437744, + "learning_rate": 3.885052802599513e-06, + "loss": 1.52, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.6623599529266357, + "learning_rate": 3.88302193338749e-06, + "loss": 1.4791, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.9031424522399902, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.5954, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.763303279876709, + "learning_rate": 3.878960194963445e-06, + "loss": 1.5618, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.7647819519042969, + "eval_runtime": 37.1392, + "eval_samples_per_second": 26.926, + "eval_steps_per_second": 1.131, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 3.0733320713043213, + "learning_rate": 3.876929325751422e-06, + "loss": 1.5191, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.7809090614318848, + "learning_rate": 3.874898456539399e-06, + "loss": 1.5804, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.6357381343841553, + "learning_rate": 3.872867587327377e-06, + "loss": 1.4843, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.680595874786377, + "learning_rate": 3.870836718115354e-06, + "loss": 1.5339, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.5908069610595703, + "learning_rate": 3.868805848903331e-06, + "loss": 1.5049, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.680586338043213, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.5188, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.763528823852539, + "learning_rate": 3.864744110479286e-06, + "loss": 1.5395, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.8522050380706787, + "learning_rate": 3.862713241267262e-06, + "loss": 1.562, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.806962490081787, + "learning_rate": 3.86068237205524e-06, + "loss": 1.5419, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.9716053009033203, + "learning_rate": 3.858651502843217e-06, + "loss": 1.5691, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.7608845233917236, + "eval_runtime": 37.1433, + "eval_samples_per_second": 26.923, + "eval_steps_per_second": 1.131, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.613098382949829, + "learning_rate": 3.856620633631195e-06, + "loss": 1.5297, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.5690557956695557, + "learning_rate": 3.854589764419171e-06, + "loss": 1.4964, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.8001089096069336, + "learning_rate": 3.852558895207149e-06, + "loss": 1.5253, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.6715309619903564, + "learning_rate": 3.850528025995126e-06, + "loss": 1.601, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.7044479846954346, + "learning_rate": 3.848497156783104e-06, + "loss": 1.5526, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.7368102073669434, + "learning_rate": 3.84646628757108e-06, + "loss": 1.5088, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.9358696937561035, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.5202, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.858172655105591, + "learning_rate": 3.842404549147035e-06, + "loss": 1.5378, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.6723833084106445, + "learning_rate": 3.840373679935013e-06, + "loss": 1.5134, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.748187780380249, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.4797, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.75319242477417, + "eval_runtime": 37.1302, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.4821436405181885, + "learning_rate": 3.836311941510967e-06, + "loss": 1.5382, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.5585227012634277, + "learning_rate": 3.834281072298944e-06, + "loss": 1.5373, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.927334785461426, + "learning_rate": 3.832250203086921e-06, + "loss": 1.5447, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.6096436977386475, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.5062, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.677304744720459, + "learning_rate": 3.828188464662876e-06, + "loss": 1.5671, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.9606077671051025, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6138, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.6971657276153564, + "learning_rate": 3.82412672623883e-06, + "loss": 1.5544, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.6380839347839355, + "learning_rate": 3.822095857026808e-06, + "loss": 1.4964, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.7537858486175537, + "learning_rate": 3.820064987814785e-06, + "loss": 1.5429, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.6189136505126953, + "learning_rate": 3.818034118602762e-06, + "loss": 1.4956, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.7546342611312866, + "eval_runtime": 37.0828, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5744292736053467, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.5256, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.5503833293914795, + "learning_rate": 3.813972380178717e-06, + "loss": 1.5211, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.979384422302246, + "learning_rate": 3.811941510966694e-06, + "loss": 1.55, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.867161512374878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5655, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.8006832599639893, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.4902, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.594712734222412, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.4451, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.6632349491119385, + "learning_rate": 3.803818034118603e-06, + "loss": 1.531, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.859894275665283, + "learning_rate": 3.80178716490658e-06, + "loss": 1.5018, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.6405653953552246, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.5765, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.7285192012786865, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.5537, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.7618242502212524, + "eval_runtime": 37.0898, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.6882379055023193, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.5143, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.706550121307373, + "learning_rate": 3.793663688058489e-06, + "loss": 1.4857, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.790907382965088, + "learning_rate": 3.791632818846467e-06, + "loss": 1.5982, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.631096363067627, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.5656, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.6981775760650635, + "learning_rate": 3.787571080422421e-06, + "loss": 1.567, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.612086296081543, + "learning_rate": 3.785540211210398e-06, + "loss": 1.5723, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.718630075454712, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.5565, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.6265041828155518, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.5326, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.578124523162842, + "learning_rate": 3.77944760357433e-06, + "loss": 1.533, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.832695484161377, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6133, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.7516050338745117, + "eval_runtime": 37.3173, + "eval_samples_per_second": 26.797, + "eval_steps_per_second": 1.125, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.851625442504883, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.5836, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.773354995938262e-06, + "loss": 1.5069, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.5738255977630615, + "learning_rate": 3.771324126726239e-06, + "loss": 1.5531, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.583575963973999, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.5589, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.7729389667510986, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.5901, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.6322245597839355, + "learning_rate": 3.765231519090171e-06, + "loss": 1.5423, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.6382699012756348, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.4574, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.7393112182617188, + "learning_rate": 3.761169780666125e-06, + "loss": 1.5351, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.684398651123047, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.4873, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.89401912689209, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.5358, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.7523635625839233, + "eval_runtime": 37.3763, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.629870891571045, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.5566, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.6936254501342773, + "learning_rate": 3.753046303818034e-06, + "loss": 1.5016, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.5889103412628174, + "learning_rate": 3.751015434606012e-06, + "loss": 1.5775, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.610703468322754, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.5508, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6318750381469727, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.5892, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.6650729179382324, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.5106, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.577174186706543, + "learning_rate": 3.742891957757921e-06, + "loss": 1.54, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.7274489402770996, + "learning_rate": 3.740861088545898e-06, + "loss": 1.5803, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6849985122680664, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.5393, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.811479091644287, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.5862, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.7490049600601196, + "eval_runtime": 36.9083, + "eval_samples_per_second": 27.094, + "eval_steps_per_second": 1.138, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.640002489089966, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.5149, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.6350319385528564, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5132, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.689814567565918, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.473, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.7181100845336914, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.4353, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.760629415512085, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.4859, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.9322562217712402, + "learning_rate": 3.724614134849716e-06, + "loss": 1.4733, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.280315399169922, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.5769, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 3.1343884468078613, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.4766, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.8537609577178955, + "learning_rate": 3.718521527213648e-06, + "loss": 1.5586, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.895299196243286, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5232, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.7473852634429932, + "eval_runtime": 37.006, + "eval_samples_per_second": 27.023, + "eval_steps_per_second": 1.135, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.82454776763916, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5057, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.711296558380127, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.4187, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.6281023025512695, + "learning_rate": 3.710398050365557e-06, + "loss": 1.4997, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.775632381439209, + "learning_rate": 3.708367181153534e-06, + "loss": 1.5087, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.8021788597106934, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.474, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 3.09647536277771, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5015, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.810450315475464, + "learning_rate": 3.702274573517466e-06, + "loss": 1.4681, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.65166974067688, + "learning_rate": 3.700243704305443e-06, + "loss": 1.4484, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.755690574645996, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5084, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.8223068714141846, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.5015, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.7516940832138062, + "eval_runtime": 37.0227, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.8401107788085938, + "learning_rate": 3.694151096669375e-06, + "loss": 1.4597, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 3.010011672973633, + "learning_rate": 3.692120227457352e-06, + "loss": 1.5625, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8575570583343506, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.4828, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.8527276515960693, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5074, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.81260085105896, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.4132, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.988429069519043, + "learning_rate": 3.683996750609261e-06, + "loss": 1.4803, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.964179515838623, + "learning_rate": 3.681965881397238e-06, + "loss": 1.4871, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.7151682376861572, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.5061, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.764187812805176, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.5188, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.810554265975952, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.4779, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.7422841787338257, + "eval_runtime": 37.0641, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.7640810012817383, + "learning_rate": 3.673842404549147e-06, + "loss": 1.4827, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.7683937549591064, + "learning_rate": 3.671811535337125e-06, + "loss": 1.4775, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.86434268951416, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.4752, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 3.0976617336273193, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.5099, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 3.107388496398926, + "learning_rate": 3.665718927701056e-06, + "loss": 1.5646, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 3.032921314239502, + "learning_rate": 3.663688058489034e-06, + "loss": 1.5048, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.9377830028533936, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5333, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.706542730331421, + "learning_rate": 3.659626320064988e-06, + "loss": 1.4903, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.819936513900757, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.4441, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.805360794067383, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.5122, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.747745156288147, + "eval_runtime": 37.0801, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.848580837249756, + "learning_rate": 3.65353371242892e-06, + "loss": 1.4984, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.772721290588379, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5026, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.760772943496704, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.4005, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.937507390975952, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.5131, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.923210859298706, + "learning_rate": 3.645410235580829e-06, + "loss": 1.4661, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8029167652130127, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.518, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.752716541290283, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.5109, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.7767865657806396, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.4831, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8483617305755615, + "learning_rate": 3.637286758732738e-06, + "loss": 1.4872, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.8714234828948975, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.5109, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.7455583810806274, + "eval_runtime": 37.0892, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7373011112213135, + "learning_rate": 3.633225020308692e-06, + "loss": 1.55, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.942373752593994, + "learning_rate": 3.63119415109667e-06, + "loss": 1.5079, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.8790931701660156, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.4669, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.9564976692199707, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5555, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.811699390411377, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.4457, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.926544427871704, + "learning_rate": 3.623070674248579e-06, + "loss": 1.4368, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.8694913387298584, + "learning_rate": 3.621039805036556e-06, + "loss": 1.4875, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 3.0294361114501953, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.4815, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8643198013305664, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.5031, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 3.0244719982147217, + "learning_rate": 3.614947197400488e-06, + "loss": 1.4914, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.7504756450653076, + "eval_runtime": 37.0238, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.8641672134399414, + "learning_rate": 3.612916328188465e-06, + "loss": 1.5122, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.8713624477386475, + "learning_rate": 3.610885458976442e-06, + "loss": 1.43, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.990912914276123, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.5124, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.963945150375366, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.533, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.802168846130371, + "learning_rate": 3.604792851340374e-06, + "loss": 1.4531, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 3.0556633472442627, + "learning_rate": 3.602761982128351e-06, + "loss": 1.4618, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.95744252204895, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5356, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.813145399093628, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.5134, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.8382728099823, + "learning_rate": 3.596669374492283e-06, + "loss": 1.4413, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.821100950241089, + "learning_rate": 3.59463850528026e-06, + "loss": 1.4748, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.7488586902618408, + "eval_runtime": 37.0361, + "eval_samples_per_second": 27.001, + "eval_steps_per_second": 1.134, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.930708408355713, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.5177, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.9312615394592285, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.4493, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 3.0051841735839844, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.5208, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.793515205383301, + "learning_rate": 3.586515028432169e-06, + "loss": 1.4813, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.7595956325531006, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5379, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.992871046066284, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.5153, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.7270801067352295, + "learning_rate": 3.580422420796101e-06, + "loss": 1.4392, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.797985076904297, + "learning_rate": 3.578391551584078e-06, + "loss": 1.4728, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9595260620117188, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.5158, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.8131442070007324, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.4214, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.7503563165664673, + "eval_runtime": 37.3831, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.124, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8242709636688232, + "learning_rate": 3.57229894394801e-06, + "loss": 1.5027, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.8973116874694824, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.5294, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 3.1207027435302734, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.4902, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.803910732269287, + "learning_rate": 3.566206336311942e-06, + "loss": 1.466, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.813541889190674, + "learning_rate": 3.564175467099919e-06, + "loss": 1.4957, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.9068171977996826, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.4435, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.8491995334625244, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.4432, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.8785598278045654, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.4791, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7701117992401123, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.4607, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.4821, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.7518969774246216, + "eval_runtime": 37.4808, + "eval_samples_per_second": 26.68, + "eval_steps_per_second": 1.121, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.5228, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.8513107299804688, + "learning_rate": 3.551990251827783e-06, + "loss": 1.4812, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.937448024749756, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.4538, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.8307411670684814, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.4807, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.7986767292022705, + "learning_rate": 3.545897644191714e-06, + "loss": 1.4706, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 3.1135106086730957, + "learning_rate": 3.543866774979692e-06, + "loss": 1.4574, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.7756283283233643, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5114, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.8114547729492188, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.4458, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 3.0544545650482178, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.4934, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8990671634674072, + "learning_rate": 3.535743298131601e-06, + "loss": 1.4808, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.7468736171722412, + "eval_runtime": 36.9135, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.976139545440674, + "learning_rate": 3.533712428919578e-06, + "loss": 1.5123, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.909146785736084, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.5204, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7225229740142822, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.4746, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.6145517826080322, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.5049, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.8660519123077393, + "learning_rate": 3.525588952071487e-06, + "loss": 1.5173, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.979801654815674, + "learning_rate": 3.523558082859464e-06, + "loss": 1.5146, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.7500879764556885, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.514, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.6923978328704834, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.455, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.6728951930999756, + "learning_rate": 3.517465475223396e-06, + "loss": 1.4712, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.769507884979248, + "learning_rate": 3.515434606011373e-06, + "loss": 1.444, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.7491599321365356, + "eval_runtime": 37.066, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.9768483638763428, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.5005, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.993196725845337, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.5347, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.886883497238159, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.4959, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.9152941703796387, + "learning_rate": 3.507311129163282e-06, + "loss": 1.4641, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.966872215270996, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.4232, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.9853367805480957, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.4883, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.9353625774383545, + "learning_rate": 3.501218521527214e-06, + "loss": 1.4826, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.7860453128814697, + "learning_rate": 3.499187652315191e-06, + "loss": 1.42, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.9269113540649414, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.5087, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 3.052962303161621, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5042, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.7421623468399048, + "eval_runtime": 37.103, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.7328579425811768, + "learning_rate": 3.493095044679123e-06, + "loss": 1.4375, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.876138925552368, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5559, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.7500345706939697, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.4043, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.836596965789795, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.4808, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.797060966491699, + "learning_rate": 3.484971567831032e-06, + "loss": 1.405, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.928368091583252, + "learning_rate": 3.482940698619009e-06, + "loss": 1.4765, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.923375368118286, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.5134, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.8624660968780518, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.4881, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.9593892097473145, + "learning_rate": 3.476848090982941e-06, + "loss": 1.4869, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.8019728660583496, + "learning_rate": 3.474817221770918e-06, + "loss": 1.468, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.7386831045150757, + "eval_runtime": 37.1034, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.8965771198272705, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.5604, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.8114798069000244, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.5092, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.9554526805877686, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.4869, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.981499195098877, + "learning_rate": 3.466693744922827e-06, + "loss": 1.4454, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.8626201152801514, + "learning_rate": 3.464662875710805e-06, + "loss": 1.4701, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.9752259254455566, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.5004, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 3.1553750038146973, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.4652, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.9458937644958496, + "learning_rate": 3.458570268074736e-06, + "loss": 1.51, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.9626898765563965, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.484, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.8093032836914062, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.468, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.740411639213562, + "eval_runtime": 37.1036, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.928344488143921, + "learning_rate": 3.452477660438668e-06, + "loss": 1.4474, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 3.0413129329681396, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.4759, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.7930357456207275, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.4406, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.8075618743896484, + "learning_rate": 3.4463850528026e-06, + "loss": 1.48, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 3.019402503967285, + "learning_rate": 3.444354183590577e-06, + "loss": 1.4713, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 3.009504795074463, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5124, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.9452531337738037, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.464, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.6980783939361572, + "learning_rate": 3.438261575954509e-06, + "loss": 1.4143, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 3.039795160293579, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.4712, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.285968542098999, + "learning_rate": 3.434199837530463e-06, + "loss": 1.5227, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.7402914762496948, + "eval_runtime": 37.1273, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 3.035813093185425, + "learning_rate": 3.432168968318441e-06, + "loss": 1.4772, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 3.1172401905059814, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.4389, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 3.0828707218170166, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.5098, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.922488212585449, + "learning_rate": 3.426076360682372e-06, + "loss": 1.5239, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 3.03352689743042, + "learning_rate": 3.42404549147035e-06, + "loss": 1.5296, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.933380603790283, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.4851, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7250027656555176, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.3966, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.991846799850464, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.4863, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 3.0704684257507324, + "learning_rate": 3.415922014622259e-06, + "loss": 1.5366, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 3.0659193992614746, + "learning_rate": 3.413891145410236e-06, + "loss": 1.516, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.7458266019821167, + "eval_runtime": 37.06, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.9695920944213867, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.508, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 3.004558801651001, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.5092, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.8377230167388916, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.4909, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.860020875930786, + "learning_rate": 3.405767668562145e-06, + "loss": 1.4415, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8683009147644043, + "learning_rate": 3.403736799350122e-06, + "loss": 1.4997, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.91894268989563, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.4886, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.8158814907073975, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.4295, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.9528515338897705, + "learning_rate": 3.397644191714054e-06, + "loss": 1.5422, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.994856119155884, + "learning_rate": 3.395613322502031e-06, + "loss": 1.4909, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.8976950645446777, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.4959, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.7350144386291504, + "eval_runtime": 37.0807, + "eval_samples_per_second": 26.968, + "eval_steps_per_second": 1.133, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.73504638671875, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.4454, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.816823720932007, + "learning_rate": 3.389520714865963e-06, + "loss": 1.4864, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.706554889678955, + "learning_rate": 3.38748984565394e-06, + "loss": 1.483, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.9004647731781006, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.492, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8677759170532227, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.4786, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.83447527885437, + "learning_rate": 3.381397238017872e-06, + "loss": 1.4575, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.9942550659179688, + "learning_rate": 3.379366368805849e-06, + "loss": 1.4746, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.124676465988159, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.5406, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.9151222705841064, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.4748, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.734571933746338, + "learning_rate": 3.373273761169781e-06, + "loss": 1.4773, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.7389248609542847, + "eval_runtime": 37.1519, + "eval_samples_per_second": 26.917, + "eval_steps_per_second": 1.13, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.8919990062713623, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.4698, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.873870611190796, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.4509, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 3.0072309970855713, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5024, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.784576177597046, + "learning_rate": 3.36515028432169e-06, + "loss": 1.4405, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.8316547870635986, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.4697, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.8758575916290283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.495, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.879115343093872, + "learning_rate": 3.359057676685622e-06, + "loss": 1.5248, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.9559383392333984, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5229, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.6783695220947266, + "learning_rate": 3.354995938261576e-06, + "loss": 1.4251, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.8372304439544678, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.4649, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.7335349321365356, + "eval_runtime": 37.3498, + "eval_samples_per_second": 26.774, + "eval_steps_per_second": 1.125, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.729576826095581, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.4263, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.7717669010162354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.4467, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.8166871070861816, + "learning_rate": 3.346872461413485e-06, + "loss": 1.4937, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.845677137374878, + "learning_rate": 3.344841592201463e-06, + "loss": 1.4303, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.9527740478515625, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.4651, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.810873031616211, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.4521, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.77546763420105, + "learning_rate": 3.338748984565394e-06, + "loss": 1.4857, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.9098589420318604, + "learning_rate": 3.336718115353372e-06, + "loss": 1.4988, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.797712802886963, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.5307, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.916668176651001, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.4828, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.7372808456420898, + "eval_runtime": 36.888, + "eval_samples_per_second": 27.109, + "eval_steps_per_second": 1.139, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.8220417499542236, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.4636, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.851010322570801, + "learning_rate": 3.32859463850528e-06, + "loss": 1.4579, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.983619213104248, + "learning_rate": 3.326563769293258e-06, + "loss": 1.4801, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.750732898712158, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.4438, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.631105422973633, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.4439, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.944300889968872, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.437, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8841664791107178, + "learning_rate": 3.318440292445167e-06, + "loss": 1.4201, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.917189121246338, + "learning_rate": 3.316409423233144e-06, + "loss": 1.4489, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.112077236175537, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.4449, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.9847209453582764, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.5153, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.7323338985443115, + "eval_runtime": 36.9935, + "eval_samples_per_second": 27.032, + "eval_steps_per_second": 1.135, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.948901414871216, + "learning_rate": 3.310316815597076e-06, + "loss": 1.473, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.84930419921875, + "learning_rate": 3.308285946385053e-06, + "loss": 1.4796, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.9120237827301025, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5392, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.7088463306427, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.4206, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.7495498657226562, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.4759, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.9146509170532227, + "learning_rate": 3.300162469536962e-06, + "loss": 1.51, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.781975030899048, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.4667, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.7524054050445557, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.5036, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.7672629356384277, + "learning_rate": 3.294069861900894e-06, + "loss": 1.4827, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.875802516937256, + "learning_rate": 3.292038992688871e-06, + "loss": 1.519, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.7273142337799072, + "eval_runtime": 37.0032, + "eval_samples_per_second": 27.025, + "eval_steps_per_second": 1.135, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.8699231147766113, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.474, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.8670589923858643, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.5039, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.9946320056915283, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5226, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.767071008682251, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.4221, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8807036876678467, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.486, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.687720537185669, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.3944, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.8170154094696045, + "learning_rate": 3.277822908204712e-06, + "loss": 1.5213, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.905557632446289, + "learning_rate": 3.275792038992689e-06, + "loss": 1.4598, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.643657922744751, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.4282, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.798712968826294, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.4752, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.7295492887496948, + "eval_runtime": 37.0432, + "eval_samples_per_second": 26.996, + "eval_steps_per_second": 1.134, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.8480567932128906, + "learning_rate": 3.269699431356621e-06, + "loss": 1.4322, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.8311381340026855, + "learning_rate": 3.267668562144598e-06, + "loss": 1.4744, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.83522891998291, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.4578, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.757307767868042, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.4469, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.9117610454559326, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.5034, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.8141143321990967, + "learning_rate": 3.259545085296507e-06, + "loss": 1.473, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 3.050138473510742, + "learning_rate": 3.257514216084484e-06, + "loss": 1.528, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.660510540008545, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.4335, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.898832082748413, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.5335, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.955589532852173, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.4832, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.7293461561203003, + "eval_runtime": 37.0096, + "eval_samples_per_second": 27.02, + "eval_steps_per_second": 1.135, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.8524296283721924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.4728, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.9741404056549072, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.4805, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.738152265548706, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.4532, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.717066764831543, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.4741, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.8367488384246826, + "learning_rate": 3.241267262388302e-06, + "loss": 1.4773, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.987780809402466, + "learning_rate": 3.23923639317628e-06, + "loss": 1.4214, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.705444097518921, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.4537, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.8966188430786133, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.4552, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.9027600288391113, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.4883, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.965813159942627, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5318, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.7236050367355347, + "eval_runtime": 37.0306, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.6390798091888428, + "learning_rate": 3.229082047116166e-06, + "loss": 1.4431, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.822848081588745, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5163, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7289068698883057, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.4551, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.766545057296753, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.4648, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.8657522201538086, + "learning_rate": 3.220958570268075e-06, + "loss": 1.4825, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.7456932067871094, + "learning_rate": 3.218927701056052e-06, + "loss": 1.3909, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 3.0274853706359863, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.4461, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.990029811859131, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.4057, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 3.086756706237793, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5251, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 3.0227134227752686, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.427, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.7248852252960205, + "eval_runtime": 37.0619, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.815480947494507, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.469, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.9171998500823975, + "learning_rate": 3.206742485783916e-06, + "loss": 1.4977, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 3.087686777114868, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.4233, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.7349331378936768, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.4187, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.9189705848693848, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.4949, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 3.0483345985412598, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5041, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.7958145141601562, + "learning_rate": 3.196588139723802e-06, + "loss": 1.4009, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 3.081714153289795, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.4718, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.7897303104400635, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.4926, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.899045944213867, + "learning_rate": 3.190495532087734e-06, + "loss": 1.4524, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.7273337841033936, + "eval_runtime": 37.0682, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.979063034057617, + "learning_rate": 3.188464662875711e-06, + "loss": 1.4706, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.8633804321289062, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5273, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.8997795581817627, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.4638, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.906151294708252, + "learning_rate": 3.182372055239643e-06, + "loss": 1.4187, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 3.0212671756744385, + "learning_rate": 3.18034118602762e-06, + "loss": 1.4213, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 3.0891034603118896, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.4921, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.9302260875701904, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.4787, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.8741140365600586, + "learning_rate": 3.174248578391552e-06, + "loss": 1.4591, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 3.0567023754119873, + "learning_rate": 3.172217709179529e-06, + "loss": 1.4636, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.80245304107666, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.4523, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.7270530462265015, + "eval_runtime": 37.325, + "eval_samples_per_second": 26.792, + "eval_steps_per_second": 1.125, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.816023349761963, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.4579, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.7405545711517334, + "learning_rate": 3.166125101543461e-06, + "loss": 1.4225, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.9569272994995117, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.483, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 3.0433337688446045, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.4846, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.856781244277954, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.426, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.8880672454833984, + "learning_rate": 3.15800162469537e-06, + "loss": 1.4764, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 3.0363011360168457, + "learning_rate": 3.155970755483347e-06, + "loss": 1.4916, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 3.0171539783477783, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.4596, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 3.183234930038452, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.4936, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.802027463912964, + "learning_rate": 3.149878147847279e-06, + "loss": 1.4733, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.7253632545471191, + "eval_runtime": 37.4737, + "eval_samples_per_second": 26.685, + "eval_steps_per_second": 1.121, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.9667463302612305, + "learning_rate": 3.147847278635256e-06, + "loss": 1.508, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 3.1557469367980957, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.5459, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.7861125469207764, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.4614, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.7431211471557617, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.4297, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 3.140911340713501, + "learning_rate": 3.139723801787165e-06, + "loss": 1.4497, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.970118522644043, + "learning_rate": 3.137692932575143e-06, + "loss": 1.4409, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.94400954246521, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.4889, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.925431966781616, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.4205, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.997279167175293, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.4714, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.855996608734131, + "learning_rate": 3.129569455727051e-06, + "loss": 1.4925, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.7247990369796753, + "eval_runtime": 36.9312, + "eval_samples_per_second": 27.077, + "eval_steps_per_second": 1.137, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 3.011199951171875, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.5132, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.8528099060058594, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.4791, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.8827919960021973, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.495, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.8150227069854736, + "learning_rate": 3.12144597887896e-06, + "loss": 1.4563, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.9341959953308105, + "learning_rate": 3.119415109666938e-06, + "loss": 1.4829, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.9161651134490967, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.492, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.906963348388672, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.445, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.992582321166992, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.4988, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 3.005469799041748, + "learning_rate": 3.111291632818847e-06, + "loss": 1.4999, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.7577736377716064, + "learning_rate": 3.109260763606824e-06, + "loss": 1.5152, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.724539041519165, + "eval_runtime": 37.0601, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.870109796524048, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.4746, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.7745094299316406, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.4891, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.6818361282348633, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.4575, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.823237180709839, + "learning_rate": 3.101137286758733e-06, + "loss": 1.492, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.9089932441711426, + "learning_rate": 3.09910641754671e-06, + "loss": 1.4943, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.85476016998291, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.5049, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 3.07242488861084, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.4709, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 3.117357015609741, + "learning_rate": 3.093013809910642e-06, + "loss": 1.52, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.959383964538574, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.4775, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8594462871551514, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.5113, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.721728801727295, + "eval_runtime": 37.1106, + "eval_samples_per_second": 26.946, + "eval_steps_per_second": 1.132, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.8418593406677246, + "learning_rate": 3.086921202274574e-06, + "loss": 1.482, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.762467384338379, + "learning_rate": 3.084890333062551e-06, + "loss": 1.4672, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.8023135662078857, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.431, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.7467143535614014, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.4837, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.6832900047302246, + "learning_rate": 3.078797725426483e-06, + "loss": 1.4765, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.863823175430298, + "learning_rate": 3.07676685621446e-06, + "loss": 1.4039, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.9055254459381104, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.4655, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.8665876388549805, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.449, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 3.029784917831421, + "learning_rate": 3.070674248578392e-06, + "loss": 1.4665, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.8582639694213867, + "learning_rate": 3.068643379366369e-06, + "loss": 1.4832, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.722991704940796, + "eval_runtime": 37.0827, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.8655312061309814, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.4759, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.892972469329834, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.4363, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.7701165676116943, + "learning_rate": 3.062550771730301e-06, + "loss": 1.4798, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7921078205108643, + "learning_rate": 3.060519902518278e-06, + "loss": 1.4557, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.8233699798583984, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.4598, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.8705663681030273, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.3898, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.9667348861694336, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.4865, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 3.0962018966674805, + "learning_rate": 3.052396425670187e-06, + "loss": 1.4512, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 3.06290340423584, + "learning_rate": 3.050365556458164e-06, + "loss": 1.5129, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.8010520935058594, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.4525, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.7228375673294067, + "eval_runtime": 37.1149, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.73380446434021, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.4537, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.9469099044799805, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.4624, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.8733043670654297, + "learning_rate": 3.042242079610073e-06, + "loss": 1.4981, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.867926597595215, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.4966, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.8584275245666504, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.4161, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.8059520721435547, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.4511, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.916750192642212, + "learning_rate": 3.034118602761982e-06, + "loss": 1.4694, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.979952335357666, + "learning_rate": 3.03208773354996e-06, + "loss": 1.4624, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.807194709777832, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.4227, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.880439519882202, + "learning_rate": 3.028025995125914e-06, + "loss": 1.5094, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.7204533815383911, + "eval_runtime": 37.1346, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.7696263790130615, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.492, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.840989351272583, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.4229, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.948915481567383, + "learning_rate": 3.021933387489846e-06, + "loss": 1.4601, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.9163503646850586, + "learning_rate": 3.019902518277823e-06, + "loss": 1.4218, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 3.0191619396209717, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.4648, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.9332194328308105, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.4791, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.8832461833953857, + "learning_rate": 3.013809910641755e-06, + "loss": 1.43, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.886837959289551, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.4455, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.8201355934143066, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.4417, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.9708259105682373, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.4657, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.7213743925094604, + "eval_runtime": 37.0993, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.927104949951172, + "learning_rate": 3.005686433793664e-06, + "loss": 1.4406, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.9701600074768066, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.4692, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 3.0241730213165283, + "learning_rate": 3.001624695369618e-06, + "loss": 1.4336, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.8802919387817383, + "learning_rate": 2.999593826157596e-06, + "loss": 1.4948, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.7666738033294678, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.4518, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 3.1206812858581543, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.4707, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 3.069323778152466, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.4534, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 3.155106544494629, + "learning_rate": 2.991470349309505e-06, + "loss": 1.5153, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.923816680908203, + "learning_rate": 2.989439480097482e-06, + "loss": 1.4606, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 3.2247867584228516, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.4564, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.7207131385803223, + "eval_runtime": 37.0684, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.8193411827087402, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.4721, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.865691661834717, + "learning_rate": 2.983346872461414e-06, + "loss": 1.5072, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 3.125004529953003, + "learning_rate": 2.981316003249391e-06, + "loss": 1.4913, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.873455286026001, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.4699, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.986186981201172, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.4808, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.8694097995758057, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.495, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.8555796146392822, + "learning_rate": 2.9731925264013e-06, + "loss": 1.4481, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.8808774948120117, + "learning_rate": 2.971161657189277e-06, + "loss": 1.4361, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.8831286430358887, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.4489, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.7533552646636963, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.4386, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.7256641387939453, + "eval_runtime": 37.1621, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 2.783586025238037, + "learning_rate": 2.965069049553209e-06, + "loss": 1.4476, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.9043776988983154, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5077, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.838284492492676, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.4287, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.94248366355896, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.4987, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.9397339820861816, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.4809, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.8922536373138428, + "learning_rate": 2.954914703493095e-06, + "loss": 1.426, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.787680149078369, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.4549, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.8116507530212402, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.4324, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.7330825328826904, + "learning_rate": 2.948822095857027e-06, + "loss": 1.4244, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.806539535522461, + "learning_rate": 2.946791226645004e-06, + "loss": 1.4604, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.7229852676391602, + "eval_runtime": 37.378, + "eval_samples_per_second": 26.754, + "eval_steps_per_second": 1.124, + "step": 1300 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 4.052502232170496e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1300/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/latest new file mode 100644 index 0000000000000000000000000000000000000000..c2a0b5b17b996514a9bea0debe9403a641844bce --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/latest @@ -0,0 +1 @@ +global_step1400 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..9a8517f15a425295019af3621cb80a36648f823e --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/trainer_state.json @@ -0,0 +1,10953 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.5342465753424657, + "eval_steps": 10, + "global_step": 1400, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6624906063079834, + "learning_rate": 4.77051177904143e-06, + "loss": 1.704, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 3.0075268745422363, + "learning_rate": 4.768480909829407e-06, + "loss": 1.6322, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.6070828437805176, + "learning_rate": 4.766450040617385e-06, + "loss": 1.6871, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 3.044222831726074, + "learning_rate": 4.764419171405362e-06, + "loss": 1.6652, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.773167610168457, + "learning_rate": 4.762388302193339e-06, + "loss": 1.7063, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.880176305770874, + "learning_rate": 4.760357432981316e-06, + "loss": 1.6775, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.6402764320373535, + "learning_rate": 4.758326563769294e-06, + "loss": 1.6694, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.9245846271514893, + "learning_rate": 4.756295694557271e-06, + "loss": 1.6688, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.623044967651367, + "learning_rate": 4.754264825345248e-06, + "loss": 1.6819, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.739896059036255, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.6401, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.8707164525985718, + "eval_runtime": 36.9513, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.603599786758423, + "learning_rate": 4.750203086921203e-06, + "loss": 1.6412, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.5571675300598145, + "learning_rate": 4.74817221770918e-06, + "loss": 1.5779, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.87091326713562, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.6878, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.8140673637390137, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7023, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.6550583839416504, + "learning_rate": 4.742079610073111e-06, + "loss": 1.5995, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.6259925365448, + "learning_rate": 4.740048740861089e-06, + "loss": 1.6194, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.971848726272583, + "learning_rate": 4.738017871649066e-06, + "loss": 1.6634, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6344778537750244, + "learning_rate": 4.735987002437044e-06, + "loss": 1.6502, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 2.9362456798553467, + "learning_rate": 4.73395613322502e-06, + "loss": 1.6948, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.8098835945129395, + "learning_rate": 4.731925264012998e-06, + "loss": 1.7161, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.8724393844604492, + "eval_runtime": 37.0876, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.933009386062622, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7014, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.8216991424560547, + "learning_rate": 4.727863525588953e-06, + "loss": 1.6192, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.7028493881225586, + "learning_rate": 4.725832656376929e-06, + "loss": 1.607, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.816818952560425, + "learning_rate": 4.723801787164907e-06, + "loss": 1.6183, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.7707865238189697, + "learning_rate": 4.721770917952885e-06, + "loss": 1.6795, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.779562473297119, + "learning_rate": 4.719740048740861e-06, + "loss": 1.63, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.907338857650757, + "learning_rate": 4.717709179528839e-06, + "loss": 1.644, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 3.073558807373047, + "learning_rate": 4.715678310316816e-06, + "loss": 1.6821, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.7556493282318115, + "learning_rate": 4.713647441104794e-06, + "loss": 1.6501, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6718344688415527, + "learning_rate": 4.71161657189277e-06, + "loss": 1.6225, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.854687213897705, + "eval_runtime": 36.9773, + "eval_samples_per_second": 27.044, + "eval_steps_per_second": 1.136, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.763335943222046, + "learning_rate": 4.709585702680748e-06, + "loss": 1.6311, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.7669479846954346, + "learning_rate": 4.707554833468725e-06, + "loss": 1.6809, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.711562395095825, + "learning_rate": 4.705523964256703e-06, + "loss": 1.741, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.799954652786255, + "learning_rate": 4.703493095044679e-06, + "loss": 1.6409, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.7278690338134766, + "learning_rate": 4.701462225832657e-06, + "loss": 1.6325, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.6802427768707275, + "learning_rate": 4.699431356620634e-06, + "loss": 1.5988, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.8826160430908203, + "learning_rate": 4.697400487408612e-06, + "loss": 1.7119, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.784343719482422, + "learning_rate": 4.695369618196588e-06, + "loss": 1.6855, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.683337688446045, + "learning_rate": 4.693338748984566e-06, + "loss": 1.6588, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.6646857261657715, + "learning_rate": 4.691307879772543e-06, + "loss": 1.623, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.8511513471603394, + "eval_runtime": 36.9852, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.8181183338165283, + "learning_rate": 4.68927701056052e-06, + "loss": 1.6171, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.8270857334136963, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.678, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.7685446739196777, + "learning_rate": 4.685215272136475e-06, + "loss": 1.6575, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.7003183364868164, + "learning_rate": 4.683184402924452e-06, + "loss": 1.6405, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.668750286102295, + "learning_rate": 4.681153533712429e-06, + "loss": 1.6541, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 3.065800905227661, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7288, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.615601062774658, + "learning_rate": 4.677091795288384e-06, + "loss": 1.6306, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.5948891639709473, + "learning_rate": 4.675060926076361e-06, + "loss": 1.6018, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.7550079822540283, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7009, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.868687868118286, + "learning_rate": 4.670999187652316e-06, + "loss": 1.6034, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.8473671674728394, + "eval_runtime": 37.0015, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.705962657928467, + "learning_rate": 4.668968318440293e-06, + "loss": 1.6084, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.7901813983917236, + "learning_rate": 4.66693744922827e-06, + "loss": 1.6171, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.688774824142456, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.6774, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.734617233276367, + "learning_rate": 4.662875710804225e-06, + "loss": 1.5796, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.742017984390259, + "learning_rate": 4.660844841592202e-06, + "loss": 1.6671, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.6438794136047363, + "learning_rate": 4.658813972380179e-06, + "loss": 1.5835, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6441051959991455, + "learning_rate": 4.656783103168157e-06, + "loss": 1.6909, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.824204206466675, + "learning_rate": 4.654752233956134e-06, + "loss": 1.5758, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.556638479232788, + "learning_rate": 4.652721364744111e-06, + "loss": 1.6504, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.897812843322754, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.6406, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.8537043333053589, + "eval_runtime": 37.257, + "eval_samples_per_second": 26.841, + "eval_steps_per_second": 1.127, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.6594088077545166, + "learning_rate": 4.648659626320066e-06, + "loss": 1.6966, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.79524302482605, + "learning_rate": 4.646628757108043e-06, + "loss": 1.6645, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.80291748046875, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.6281, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.7723848819732666, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.5959, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 3.0625650882720947, + "learning_rate": 4.640536149471974e-06, + "loss": 1.6313, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.775726079940796, + "learning_rate": 4.638505280259952e-06, + "loss": 1.6563, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.837758779525757, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.6727, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 3.041355609893799, + "learning_rate": 4.634443541835907e-06, + "loss": 1.6282, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.5775389671325684, + "learning_rate": 4.632412672623883e-06, + "loss": 1.651, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 3.026287794113159, + "learning_rate": 4.630381803411861e-06, + "loss": 1.6458, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.8461565971374512, + "eval_runtime": 37.3881, + "eval_samples_per_second": 26.747, + "eval_steps_per_second": 1.123, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.6781764030456543, + "learning_rate": 4.628350934199838e-06, + "loss": 1.5232, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.763350486755371, + "learning_rate": 4.626320064987816e-06, + "loss": 1.6186, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.7063372135162354, + "learning_rate": 4.624289195775792e-06, + "loss": 1.6784, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.824563980102539, + "learning_rate": 4.62225832656377e-06, + "loss": 1.6219, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.5335569381713867, + "learning_rate": 4.620227457351747e-06, + "loss": 1.6152, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.773644208908081, + "learning_rate": 4.618196588139724e-06, + "loss": 1.6179, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.9841721057891846, + "learning_rate": 4.616165718927701e-06, + "loss": 1.6258, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.684171676635742, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.6066, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.8817965984344482, + "learning_rate": 4.612103980503656e-06, + "loss": 1.6697, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.9846866130828857, + "learning_rate": 4.610073111291633e-06, + "loss": 1.6541, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.8485002517700195, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.904170513153076, + "learning_rate": 4.60804224207961e-06, + "loss": 1.6362, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.684476375579834, + "learning_rate": 4.606011372867588e-06, + "loss": 1.6303, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6692216396331787, + "learning_rate": 4.603980503655565e-06, + "loss": 1.638, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.73555588722229, + "learning_rate": 4.601949634443542e-06, + "loss": 1.6125, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.729525089263916, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.565, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.5972018241882324, + "learning_rate": 4.597887896019497e-06, + "loss": 1.5856, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7859137058258057, + "learning_rate": 4.595857026807474e-06, + "loss": 1.6325, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.662587881088257, + "learning_rate": 4.593826157595451e-06, + "loss": 1.6172, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.6380577087402344, + "learning_rate": 4.591795288383429e-06, + "loss": 1.6005, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8375349044799805, + "learning_rate": 4.589764419171406e-06, + "loss": 1.6507, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.833074688911438, + "eval_runtime": 37.0695, + "eval_samples_per_second": 26.976, + "eval_steps_per_second": 1.133, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.786713123321533, + "learning_rate": 4.587733549959383e-06, + "loss": 1.6283, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.6959030628204346, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.5994, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.6100692749023438, + "learning_rate": 4.583671811535338e-06, + "loss": 1.5884, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.7281863689422607, + "learning_rate": 4.581640942323315e-06, + "loss": 1.5737, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.7519376277923584, + "learning_rate": 4.579610073111292e-06, + "loss": 1.6531, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.9314866065979004, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.6755, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.9250576496124268, + "learning_rate": 4.575548334687247e-06, + "loss": 1.6604, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.6382815837860107, + "learning_rate": 4.573517465475224e-06, + "loss": 1.6773, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.6891751289367676, + "learning_rate": 4.571486596263201e-06, + "loss": 1.5849, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.7018136978149414, + "learning_rate": 4.569455727051178e-06, + "loss": 1.6358, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.838019847869873, + "eval_runtime": 37.0496, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.671952247619629, + "learning_rate": 4.567424857839156e-06, + "loss": 1.6207, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.882539749145508, + "learning_rate": 4.565393988627132e-06, + "loss": 1.6618, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.684070587158203, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.6502, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.5295751094818115, + "learning_rate": 4.561332250203087e-06, + "loss": 1.6078, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.0268194675445557, + "learning_rate": 4.559301380991065e-06, + "loss": 1.6312, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6895201206207275, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.5848, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.869950771331787, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.5951, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.909505605697632, + "learning_rate": 4.553208773354996e-06, + "loss": 1.6589, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.7604334354400635, + "learning_rate": 4.551177904142974e-06, + "loss": 1.5767, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.8792080879211426, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.6864, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.836847186088562, + "eval_runtime": 36.9142, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.607037305831909, + "learning_rate": 4.547116165718929e-06, + "loss": 1.5925, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.9338927268981934, + "learning_rate": 4.545085296506905e-06, + "loss": 1.6433, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.7723233699798584, + "learning_rate": 4.543054427294882e-06, + "loss": 1.5963, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.4122188091278076, + "learning_rate": 4.54102355808286e-06, + "loss": 1.5978, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.7582709789276123, + "learning_rate": 4.538992688870837e-06, + "loss": 1.581, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.6983039379119873, + "learning_rate": 4.536961819658814e-06, + "loss": 1.6253, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.6036365032196045, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.5957, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.759645700454712, + "learning_rate": 4.532900081234769e-06, + "loss": 1.6199, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.797744035720825, + "learning_rate": 4.530869212022746e-06, + "loss": 1.5854, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.752868890762329, + "learning_rate": 4.528838342810723e-06, + "loss": 1.619, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.8305871486663818, + "eval_runtime": 37.0307, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.7421889305114746, + "learning_rate": 4.526807473598701e-06, + "loss": 1.6182, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.6852707862854004, + "learning_rate": 4.524776604386678e-06, + "loss": 1.6135, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.861720323562622, + "learning_rate": 4.522745735174655e-06, + "loss": 1.6099, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.706851005554199, + "learning_rate": 4.520714865962632e-06, + "loss": 1.6597, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7992191314697266, + "learning_rate": 4.51868399675061e-06, + "loss": 1.6077, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.6248183250427246, + "learning_rate": 4.516653127538587e-06, + "loss": 1.6485, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.696040391921997, + "learning_rate": 4.514622258326564e-06, + "loss": 1.5853, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.5958261489868164, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.571, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.9715874195098877, + "learning_rate": 4.510560519902519e-06, + "loss": 1.6044, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.664834499359131, + "learning_rate": 4.508529650690496e-06, + "loss": 1.6568, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.8277279138565063, + "eval_runtime": 37.0489, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.869558811187744, + "learning_rate": 4.506498781478473e-06, + "loss": 1.5654, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.5814828872680664, + "learning_rate": 4.504467912266451e-06, + "loss": 1.5484, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.6158947944641113, + "learning_rate": 4.502437043054428e-06, + "loss": 1.6202, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.7424654960632324, + "learning_rate": 4.500406173842405e-06, + "loss": 1.6755, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.8670949935913086, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.6144, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.7914652824401855, + "learning_rate": 4.49634443541836e-06, + "loss": 1.6164, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.866380214691162, + "learning_rate": 4.494313566206336e-06, + "loss": 1.5961, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.966474771499634, + "learning_rate": 4.492282696994314e-06, + "loss": 1.6681, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.751291513442993, + "learning_rate": 4.490251827782291e-06, + "loss": 1.6338, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.58840012550354, + "learning_rate": 4.488220958570269e-06, + "loss": 1.5619, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.824889898300171, + "eval_runtime": 37.0821, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.786945104598999, + "learning_rate": 4.486190089358245e-06, + "loss": 1.5712, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4536845684051514, + "learning_rate": 4.484159220146223e-06, + "loss": 1.5487, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.780278205871582, + "learning_rate": 4.4821283509342e-06, + "loss": 1.6269, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.5427017211914062, + "learning_rate": 4.480097481722178e-06, + "loss": 1.6254, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.7540087699890137, + "learning_rate": 4.478066612510154e-06, + "loss": 1.6112, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.616352081298828, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.5637, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.686035633087158, + "learning_rate": 4.474004874086109e-06, + "loss": 1.5974, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.982365846633911, + "learning_rate": 4.471974004874087e-06, + "loss": 1.7034, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.6106085777282715, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.5482, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.7798519134521484, + "learning_rate": 4.467912266450041e-06, + "loss": 1.6475, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.820489764213562, + "eval_runtime": 37.0672, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.8634185791015625, + "learning_rate": 4.465881397238018e-06, + "loss": 1.6258, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.633626699447632, + "learning_rate": 4.463850528025995e-06, + "loss": 1.5806, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.719074249267578, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.6404, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.7545487880706787, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6404, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.6559669971466064, + "learning_rate": 4.457757920389927e-06, + "loss": 1.6104, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.9004898071289062, + "learning_rate": 4.455727051177904e-06, + "loss": 1.5513, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.6640021800994873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.5859, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.842026472091675, + "learning_rate": 4.451665312753859e-06, + "loss": 1.6241, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.784647226333618, + "learning_rate": 4.449634443541836e-06, + "loss": 1.627, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.7067618370056152, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.6276, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.8115602731704712, + "eval_runtime": 37.078, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.6582882404327393, + "learning_rate": 4.445572705117791e-06, + "loss": 1.5746, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7881102561950684, + "learning_rate": 4.443541835905768e-06, + "loss": 1.5749, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.9323439598083496, + "learning_rate": 4.441510966693745e-06, + "loss": 1.6421, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.63948655128479, + "learning_rate": 4.439480097481723e-06, + "loss": 1.5853, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.7746710777282715, + "learning_rate": 4.4374492282697e-06, + "loss": 1.6455, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.093895435333252, + "learning_rate": 4.435418359057677e-06, + "loss": 1.5311, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.526385545730591, + "learning_rate": 4.433387489845654e-06, + "loss": 1.6082, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.6604297161102295, + "learning_rate": 4.431356620633632e-06, + "loss": 1.61, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.6636247634887695, + "learning_rate": 4.429325751421609e-06, + "loss": 1.5815, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.5567514896392822, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6086, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.8170957565307617, + "eval_runtime": 37.4095, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.695563793182373, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.5836, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6009771823883057, + "learning_rate": 4.423233143785541e-06, + "loss": 1.5264, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.0490736961364746, + "learning_rate": 4.421202274573518e-06, + "loss": 1.5982, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.8609886169433594, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6265, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4415369033813477, + "learning_rate": 4.417140536149473e-06, + "loss": 1.4987, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.771657943725586, + "learning_rate": 4.415109666937449e-06, + "loss": 1.5494, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.739617347717285, + "learning_rate": 4.413078797725427e-06, + "loss": 1.6274, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.805098295211792, + "learning_rate": 4.411047928513404e-06, + "loss": 1.5686, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.7248735427856445, + "learning_rate": 4.409017059301382e-06, + "loss": 1.6027, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6653318405151367, + "learning_rate": 4.406986190089358e-06, + "loss": 1.5909, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.8076283931732178, + "eval_runtime": 37.5073, + "eval_samples_per_second": 26.661, + "eval_steps_per_second": 1.12, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.626783847808838, + "learning_rate": 4.404955320877336e-06, + "loss": 1.6205, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.6929268836975098, + "learning_rate": 4.402924451665313e-06, + "loss": 1.6146, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.6806800365448, + "learning_rate": 4.400893582453291e-06, + "loss": 1.5702, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.5407259464263916, + "learning_rate": 4.398862713241267e-06, + "loss": 1.5525, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.639122247695923, + "learning_rate": 4.396831844029245e-06, + "loss": 1.5826, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.669396162033081, + "learning_rate": 4.394800974817222e-06, + "loss": 1.5982, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.6377906799316406, + "learning_rate": 4.392770105605199e-06, + "loss": 1.5797, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6769638061523438, + "learning_rate": 4.390739236393176e-06, + "loss": 1.607, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.6347222328186035, + "learning_rate": 4.388708367181154e-06, + "loss": 1.6196, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.670680284500122, + "learning_rate": 4.386677497969131e-06, + "loss": 1.619, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.8124384880065918, + "eval_runtime": 37.4751, + "eval_samples_per_second": 26.684, + "eval_steps_per_second": 1.121, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.5886521339416504, + "learning_rate": 4.384646628757108e-06, + "loss": 1.6258, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.644566297531128, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6447, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.6432394981384277, + "learning_rate": 4.380584890333063e-06, + "loss": 1.5888, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.5602784156799316, + "learning_rate": 4.37855402112104e-06, + "loss": 1.5544, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.4874439239501953, + "learning_rate": 4.376523151909017e-06, + "loss": 1.5075, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.694748640060425, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.6498, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.582211494445801, + "learning_rate": 4.372461413484972e-06, + "loss": 1.5877, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.675713539123535, + "learning_rate": 4.370430544272949e-06, + "loss": 1.6045, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.7157161235809326, + "learning_rate": 4.368399675060926e-06, + "loss": 1.623, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.795076370239258, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6141, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.8062744140625, + "eval_runtime": 37.2435, + "eval_samples_per_second": 26.85, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6296374797821045, + "learning_rate": 4.364337936636881e-06, + "loss": 1.6087, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.362307067424858e-06, + "loss": 1.5182, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.7965915203094482, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.6025, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.807307481765747, + "learning_rate": 4.358245329000813e-06, + "loss": 1.574, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.584615468978882, + "learning_rate": 4.35621445978879e-06, + "loss": 1.5922, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.8472282886505127, + "learning_rate": 4.354183590576767e-06, + "loss": 1.5914, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.8471736907958984, + "learning_rate": 4.352152721364745e-06, + "loss": 1.5749, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.569225549697876, + "learning_rate": 4.350121852152722e-06, + "loss": 1.5591, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.688157796859741, + "learning_rate": 4.348090982940699e-06, + "loss": 1.6042, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.5865259170532227, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.6307, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.8040707111358643, + "eval_runtime": 36.9511, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.71850323677063, + "learning_rate": 4.344029244516654e-06, + "loss": 1.6043, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.7158093452453613, + "learning_rate": 4.341998375304631e-06, + "loss": 1.5907, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6879327297210693, + "learning_rate": 4.339967506092607e-06, + "loss": 1.6365, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.827277660369873, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.5928, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6709842681884766, + "learning_rate": 4.335905767668562e-06, + "loss": 1.5812, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.7071495056152344, + "learning_rate": 4.33387489845654e-06, + "loss": 1.5571, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6008732318878174, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.6048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.6238059997558594, + "learning_rate": 4.329813160032495e-06, + "loss": 1.5962, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.567065954208374, + "learning_rate": 4.327782290820471e-06, + "loss": 1.5713, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.573068857192993, + "learning_rate": 4.325751421608449e-06, + "loss": 1.536, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.8103471994400024, + "eval_runtime": 37.0956, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6245224475860596, + "learning_rate": 4.323720552396426e-06, + "loss": 1.5986, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.6243810653686523, + "learning_rate": 4.321689683184404e-06, + "loss": 1.589, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.5303313732147217, + "learning_rate": 4.31965881397238e-06, + "loss": 1.5793, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.5617806911468506, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.5704, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.7429656982421875, + "learning_rate": 4.315597075548335e-06, + "loss": 1.6174, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.6753251552581787, + "learning_rate": 4.313566206336312e-06, + "loss": 1.5879, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.5944488048553467, + "learning_rate": 4.311535337124289e-06, + "loss": 1.6082, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.9002904891967773, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.5521, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.5384247303009033, + "learning_rate": 4.307473598700244e-06, + "loss": 1.5751, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6015477180480957, + "learning_rate": 4.305442729488221e-06, + "loss": 1.5966, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8091154098510742, + "eval_runtime": 37.1075, + "eval_samples_per_second": 26.949, + "eval_steps_per_second": 1.132, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.6002886295318604, + "learning_rate": 4.303411860276198e-06, + "loss": 1.5403, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7755627632141113, + "learning_rate": 4.301380991064176e-06, + "loss": 1.6192, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.8351540565490723, + "learning_rate": 4.299350121852153e-06, + "loss": 1.6217, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.684007406234741, + "learning_rate": 4.29731925264013e-06, + "loss": 1.6242, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.8019614219665527, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6173, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.9122867584228516, + "learning_rate": 4.293257514216085e-06, + "loss": 1.5701, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.545732021331787, + "learning_rate": 4.291226645004062e-06, + "loss": 1.5494, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.9499566555023193, + "learning_rate": 4.289195775792039e-06, + "loss": 1.5848, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.8071892261505127, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6114, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.763864517211914, + "learning_rate": 4.285134037367994e-06, + "loss": 1.5768, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8057912588119507, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 3.001190423965454, + "learning_rate": 4.283103168155971e-06, + "loss": 1.5601, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.7640297412872314, + "learning_rate": 4.281072298943948e-06, + "loss": 1.5619, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.7372794151306152, + "learning_rate": 4.279041429731926e-06, + "loss": 1.562, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 3.0390751361846924, + "learning_rate": 4.277010560519903e-06, + "loss": 1.5519, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.758798599243164, + "learning_rate": 4.27497969130788e-06, + "loss": 1.527, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.96443772315979, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.5899, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.7941088676452637, + "learning_rate": 4.270917952883835e-06, + "loss": 1.617, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 3.0385100841522217, + "learning_rate": 4.268887083671812e-06, + "loss": 1.5996, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.6946637630462646, + "learning_rate": 4.266856214459789e-06, + "loss": 1.5641, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.776723861694336, + "learning_rate": 4.264825345247766e-06, + "loss": 1.6451, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.7911298274993896, + "eval_runtime": 37.0623, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.851224184036255, + "learning_rate": 4.262794476035744e-06, + "loss": 1.5302, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9045867919921875, + "learning_rate": 4.26076360682372e-06, + "loss": 1.5808, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.6568782329559326, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.5233, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.94191837310791, + "learning_rate": 4.256701868399675e-06, + "loss": 1.5912, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.726778984069824, + "learning_rate": 4.254670999187653e-06, + "loss": 1.5823, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.8704583644866943, + "learning_rate": 4.252640129975629e-06, + "loss": 1.5931, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.5721547603607178, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6128, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6883509159088135, + "learning_rate": 4.248578391551584e-06, + "loss": 1.5479, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.7186272144317627, + "learning_rate": 4.246547522339562e-06, + "loss": 1.5203, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.8326518535614014, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.5547, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.7948570251464844, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.7870523929595947, + "learning_rate": 4.242485783915517e-06, + "loss": 1.5891, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.866610288619995, + "learning_rate": 4.240454914703493e-06, + "loss": 1.5232, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.94161057472229, + "learning_rate": 4.23842404549147e-06, + "loss": 1.52, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.7331979274749756, + "learning_rate": 4.236393176279448e-06, + "loss": 1.5665, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.5375845432281494, + "learning_rate": 4.234362307067425e-06, + "loss": 1.5274, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.691927433013916, + "learning_rate": 4.232331437855402e-06, + "loss": 1.604, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7578394412994385, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.6, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.594538688659668, + "learning_rate": 4.228269699431357e-06, + "loss": 1.5928, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.5408968925476074, + "learning_rate": 4.226238830219334e-06, + "loss": 1.5908, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.689786672592163, + "learning_rate": 4.224207961007311e-06, + "loss": 1.5757, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.792494297027588, + "eval_runtime": 37.0822, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.735539197921753, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.5811, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.58634614944458, + "learning_rate": 4.220146222583266e-06, + "loss": 1.5808, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.7899138927459717, + "learning_rate": 4.218115353371243e-06, + "loss": 1.5801, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.6507744789123535, + "learning_rate": 4.21608448415922e-06, + "loss": 1.5585, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.6112217903137207, + "learning_rate": 4.214053614947198e-06, + "loss": 1.5329, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.729405164718628, + "learning_rate": 4.212022745735175e-06, + "loss": 1.6177, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.6431992053985596, + "learning_rate": 4.209991876523152e-06, + "loss": 1.5234, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.726243734359741, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.5926, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.6574180126190186, + "learning_rate": 4.205930138099107e-06, + "loss": 1.5424, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.6570956707000732, + "learning_rate": 4.203899268887084e-06, + "loss": 1.6119, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.781999945640564, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.528052806854248, + "learning_rate": 4.201868399675061e-06, + "loss": 1.598, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.6075727939605713, + "learning_rate": 4.199837530463039e-06, + "loss": 1.5886, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.588737964630127, + "learning_rate": 4.197806661251016e-06, + "loss": 1.507, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.773618221282959, + "learning_rate": 4.195775792038993e-06, + "loss": 1.5794, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.624725580215454, + "learning_rate": 4.19374492282697e-06, + "loss": 1.5495, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.662397861480713, + "learning_rate": 4.191714053614948e-06, + "loss": 1.5203, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.6617255210876465, + "learning_rate": 4.189683184402924e-06, + "loss": 1.5885, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.5784451961517334, + "learning_rate": 4.187652315190902e-06, + "loss": 1.5689, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.726011037826538, + "learning_rate": 4.185621445978879e-06, + "loss": 1.5913, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.6117324829101562, + "learning_rate": 4.183590576766857e-06, + "loss": 1.5453, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.7893352508544922, + "eval_runtime": 37.4096, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7025442123413086, + "learning_rate": 4.181559707554833e-06, + "loss": 1.5868, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.656102180480957, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6141, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.741962194442749, + "learning_rate": 4.177497969130788e-06, + "loss": 1.5516, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5887229442596436, + "learning_rate": 4.175467099918766e-06, + "loss": 1.5525, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.542985439300537, + "learning_rate": 4.173436230706742e-06, + "loss": 1.592, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.7797129154205322, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.5349, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.632887363433838, + "learning_rate": 4.169374492282697e-06, + "loss": 1.5548, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.691690444946289, + "learning_rate": 4.167343623070675e-06, + "loss": 1.5352, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.8922817707061768, + "learning_rate": 4.165312753858652e-06, + "loss": 1.5717, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.5517446994781494, + "learning_rate": 4.163281884646629e-06, + "loss": 1.5658, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.780922770500183, + "eval_runtime": 37.3755, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.718374490737915, + "learning_rate": 4.161251015434607e-06, + "loss": 1.5578, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.6249442100524902, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6018, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.7051141262054443, + "learning_rate": 4.157189277010561e-06, + "loss": 1.5182, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.677109956741333, + "learning_rate": 4.155158407798538e-06, + "loss": 1.5423, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.591059684753418, + "learning_rate": 4.153127538586516e-06, + "loss": 1.5694, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.699953556060791, + "learning_rate": 4.151096669374492e-06, + "loss": 1.5758, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.8837342262268066, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.5509, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.6566851139068604, + "learning_rate": 4.147034930950447e-06, + "loss": 1.5577, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.901057004928589, + "learning_rate": 4.145004061738425e-06, + "loss": 1.5797, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.796971321105957, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.5608, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.7797813415527344, + "eval_runtime": 36.8425, + "eval_samples_per_second": 27.143, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.605431318283081, + "learning_rate": 4.140942323314379e-06, + "loss": 1.491, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8084020614624023, + "learning_rate": 4.138911454102356e-06, + "loss": 1.5074, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.625511407852173, + "learning_rate": 4.136880584890333e-06, + "loss": 1.5404, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.6944286823272705, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.589, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.81792950630188, + "learning_rate": 4.132818846466288e-06, + "loss": 1.5646, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5813655853271484, + "learning_rate": 4.130787977254265e-06, + "loss": 1.5697, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.6733286380767822, + "learning_rate": 4.128757108042242e-06, + "loss": 1.5786, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.6972577571868896, + "learning_rate": 4.12672623883022e-06, + "loss": 1.5857, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7381093502044678, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6365, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.6162049770355225, + "learning_rate": 4.122664500406174e-06, + "loss": 1.5598, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.7869179248809814, + "eval_runtime": 36.974, + "eval_samples_per_second": 27.046, + "eval_steps_per_second": 1.136, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.7157270908355713, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.5533, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.560105085372925, + "learning_rate": 4.118602761982129e-06, + "loss": 1.5821, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.5261518955230713, + "learning_rate": 4.116571892770106e-06, + "loss": 1.5195, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7377054691314697, + "learning_rate": 4.114541023558083e-06, + "loss": 1.5066, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.632946252822876, + "learning_rate": 4.112510154346061e-06, + "loss": 1.5948, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.809671640396118, + "learning_rate": 4.110479285134038e-06, + "loss": 1.5435, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.843156099319458, + "learning_rate": 4.108448415922015e-06, + "loss": 1.5667, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.743614673614502, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5958, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.9596827030181885, + "learning_rate": 4.10438667749797e-06, + "loss": 1.5807, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.7321834564208984, + "learning_rate": 4.102355808285947e-06, + "loss": 1.4744, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.784859538078308, + "eval_runtime": 37.0146, + "eval_samples_per_second": 27.016, + "eval_steps_per_second": 1.135, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.010359764099121, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6146, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.744251251220703, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.5401, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.592690944671631, + "learning_rate": 4.096263200649879e-06, + "loss": 1.5508, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.786736011505127, + "learning_rate": 4.094232331437856e-06, + "loss": 1.5717, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.8773610591888428, + "learning_rate": 4.092201462225833e-06, + "loss": 1.5201, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.818610906600952, + "learning_rate": 4.090170593013811e-06, + "loss": 1.5282, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.9548542499542236, + "learning_rate": 4.088139723801787e-06, + "loss": 1.553, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.678842306137085, + "learning_rate": 4.086108854589765e-06, + "loss": 1.531, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.7757408618927, + "learning_rate": 4.084077985377742e-06, + "loss": 1.5371, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.6954638957977295, + "learning_rate": 4.08204711616572e-06, + "loss": 1.5462, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.781003475189209, + "eval_runtime": 37.0687, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6897518634796143, + "learning_rate": 4.080016246953696e-06, + "loss": 1.5979, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.581172466278076, + "learning_rate": 4.077985377741674e-06, + "loss": 1.5007, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.828234910964966, + "learning_rate": 4.075954508529651e-06, + "loss": 1.5857, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.6171951293945312, + "learning_rate": 4.073923639317629e-06, + "loss": 1.5937, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.6476316452026367, + "learning_rate": 4.071892770105605e-06, + "loss": 1.5864, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.7525634765625, + "learning_rate": 4.069861900893583e-06, + "loss": 1.5095, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.6881322860717773, + "learning_rate": 4.06783103168156e-06, + "loss": 1.4936, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.6104350090026855, + "learning_rate": 4.065800162469538e-06, + "loss": 1.5229, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.866706132888794, + "learning_rate": 4.063769293257514e-06, + "loss": 1.539, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.803999900817871, + "learning_rate": 4.061738424045492e-06, + "loss": 1.5792, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.7793465852737427, + "eval_runtime": 37.0749, + "eval_samples_per_second": 26.972, + "eval_steps_per_second": 1.133, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.670285940170288, + "learning_rate": 4.059707554833469e-06, + "loss": 1.5523, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.871267557144165, + "learning_rate": 4.057676685621446e-06, + "loss": 1.5623, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.6631205081939697, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6094, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.7580034732818604, + "learning_rate": 4.053614947197401e-06, + "loss": 1.5725, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.6412739753723145, + "learning_rate": 4.051584077985378e-06, + "loss": 1.4959, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.699176073074341, + "learning_rate": 4.049553208773355e-06, + "loss": 1.5586, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.622969150543213, + "learning_rate": 4.047522339561333e-06, + "loss": 1.5713, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.892338991165161, + "learning_rate": 4.04549147034931e-06, + "loss": 1.5471, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.77056622505188, + "learning_rate": 4.043460601137287e-06, + "loss": 1.5236, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.7122550010681152, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6562, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.7708193063735962, + "eval_runtime": 37.0743, + "eval_samples_per_second": 26.973, + "eval_steps_per_second": 1.133, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.702824592590332, + "learning_rate": 4.039398862713242e-06, + "loss": 1.534, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.6822307109832764, + "learning_rate": 4.037367993501219e-06, + "loss": 1.5368, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.7274515628814697, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6218, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.8114798069000244, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.532, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.8936192989349365, + "learning_rate": 4.031275385865151e-06, + "loss": 1.5124, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.7851672172546387, + "learning_rate": 4.029244516653128e-06, + "loss": 1.5596, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.815023183822632, + "learning_rate": 4.027213647441105e-06, + "loss": 1.5706, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5958921909332275, + "learning_rate": 4.025182778229083e-06, + "loss": 1.5161, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.779360771179199, + "learning_rate": 4.02315190901706e-06, + "loss": 1.5057, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.7411787509918213, + "learning_rate": 4.021121039805037e-06, + "loss": 1.5396, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.7778708934783936, + "eval_runtime": 37.0791, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.5709781646728516, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.5474, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8583388328552246, + "learning_rate": 4.017059301380992e-06, + "loss": 1.6149, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.6877572536468506, + "learning_rate": 4.015028432168969e-06, + "loss": 1.5292, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.680612087249756, + "learning_rate": 4.012997562956945e-06, + "loss": 1.596, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.653559446334839, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.4901, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.542787790298462, + "learning_rate": 4.0089358245329e-06, + "loss": 1.5295, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5476601123809814, + "learning_rate": 4.006904955320878e-06, + "loss": 1.5393, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.6367862224578857, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.5321, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.8112173080444336, + "learning_rate": 4.002843216896833e-06, + "loss": 1.5963, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.7136871814727783, + "learning_rate": 4.000812347684809e-06, + "loss": 1.5839, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.7763899564743042, + "eval_runtime": 37.2537, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.9095280170440674, + "learning_rate": 3.998781478472787e-06, + "loss": 1.5343, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.6196115016937256, + "learning_rate": 3.996750609260764e-06, + "loss": 1.5411, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6413371562957764, + "learning_rate": 3.994719740048742e-06, + "loss": 1.4862, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 3.0759036540985107, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6471, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.599147081375122, + "learning_rate": 3.990658001624696e-06, + "loss": 1.5556, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.726978302001953, + "learning_rate": 3.988627132412673e-06, + "loss": 1.5351, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.6760807037353516, + "learning_rate": 3.98659626320065e-06, + "loss": 1.5681, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.680483818054199, + "learning_rate": 3.984565393988627e-06, + "loss": 1.5225, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6649010181427, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.5325, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.638085126876831, + "learning_rate": 3.980503655564582e-06, + "loss": 1.564, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.7698904275894165, + "eval_runtime": 37.4492, + "eval_samples_per_second": 26.703, + "eval_steps_per_second": 1.122, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.647825241088867, + "learning_rate": 3.978472786352559e-06, + "loss": 1.5782, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.748222589492798, + "learning_rate": 3.976441917140536e-06, + "loss": 1.556, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.6319026947021484, + "learning_rate": 3.974411047928514e-06, + "loss": 1.5378, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.6759872436523438, + "learning_rate": 3.972380178716491e-06, + "loss": 1.6018, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.6760940551757812, + "learning_rate": 3.970349309504468e-06, + "loss": 1.5742, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.5565083026885986, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.565, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.5986931324005127, + "learning_rate": 3.966287571080423e-06, + "loss": 1.5124, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.713092803955078, + "learning_rate": 3.9642567018684e-06, + "loss": 1.523, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.6203417778015137, + "learning_rate": 3.962225832656377e-06, + "loss": 1.5601, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.604440927505493, + "learning_rate": 3.960194963444355e-06, + "loss": 1.5485, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.7633332014083862, + "eval_runtime": 37.49, + "eval_samples_per_second": 26.674, + "eval_steps_per_second": 1.12, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.5513081550598145, + "learning_rate": 3.958164094232332e-06, + "loss": 1.479, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.7990963459014893, + "learning_rate": 3.956133225020309e-06, + "loss": 1.5355, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.651113271713257, + "learning_rate": 3.954102355808286e-06, + "loss": 1.5406, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.7236671447753906, + "learning_rate": 3.952071486596264e-06, + "loss": 1.5314, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.697598695755005, + "learning_rate": 3.950040617384241e-06, + "loss": 1.5398, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.8271944522857666, + "learning_rate": 3.948009748172218e-06, + "loss": 1.54, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.733125925064087, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.5236, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.7555017471313477, + "learning_rate": 3.943948009748173e-06, + "loss": 1.5808, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.6592960357666016, + "learning_rate": 3.94191714053615e-06, + "loss": 1.5398, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.6870126724243164, + "learning_rate": 3.939886271324127e-06, + "loss": 1.5517, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.760675311088562, + "eval_runtime": 36.9372, + "eval_samples_per_second": 27.073, + "eval_steps_per_second": 1.137, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.7469022274017334, + "learning_rate": 3.937855402112104e-06, + "loss": 1.605, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.5255792140960693, + "learning_rate": 3.935824532900082e-06, + "loss": 1.5061, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.589054822921753, + "learning_rate": 3.933793663688058e-06, + "loss": 1.4971, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.7874889373779297, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.5643, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.819451093673706, + "learning_rate": 3.929731925264013e-06, + "loss": 1.5797, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.6516339778900146, + "learning_rate": 3.927701056051991e-06, + "loss": 1.5048, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.979091167449951, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.5174, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.743058204650879, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.5432, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.6864030361175537, + "learning_rate": 3.921608448415922e-06, + "loss": 1.5147, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.7751128673553467, + "learning_rate": 3.9195775792039e-06, + "loss": 1.4857, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.7609487771987915, + "eval_runtime": 37.09, + "eval_samples_per_second": 26.961, + "eval_steps_per_second": 1.132, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.626188278198242, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.5428, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.554067850112915, + "learning_rate": 3.915515840779855e-06, + "loss": 1.5153, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.8580875396728516, + "learning_rate": 3.913484971567831e-06, + "loss": 1.5562, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.7405097484588623, + "learning_rate": 3.911454102355808e-06, + "loss": 1.565, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.5650815963745117, + "learning_rate": 3.909423233143786e-06, + "loss": 1.4756, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.9405910968780518, + "learning_rate": 3.907392363931763e-06, + "loss": 1.5611, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.687448024749756, + "learning_rate": 3.90536149471974e-06, + "loss": 1.5405, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.7373857498168945, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.6055, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.875298261642456, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6215, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.713200330734253, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6493, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.7608598470687866, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.541935682296753, + "learning_rate": 3.897238017871649e-06, + "loss": 1.5257, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.582045316696167, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.5253, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.6884682178497314, + "learning_rate": 3.893176279447604e-06, + "loss": 1.5329, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.6456501483917236, + "learning_rate": 3.891145410235581e-06, + "loss": 1.5167, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.60788893699646, + "learning_rate": 3.889114541023558e-06, + "loss": 1.5945, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.708646774291992, + "learning_rate": 3.887083671811536e-06, + "loss": 1.5704, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.666922092437744, + "learning_rate": 3.885052802599513e-06, + "loss": 1.52, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.6623599529266357, + "learning_rate": 3.88302193338749e-06, + "loss": 1.4791, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.9031424522399902, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.5954, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.763303279876709, + "learning_rate": 3.878960194963445e-06, + "loss": 1.5618, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.7647819519042969, + "eval_runtime": 37.1392, + "eval_samples_per_second": 26.926, + "eval_steps_per_second": 1.131, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 3.0733320713043213, + "learning_rate": 3.876929325751422e-06, + "loss": 1.5191, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.7809090614318848, + "learning_rate": 3.874898456539399e-06, + "loss": 1.5804, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.6357381343841553, + "learning_rate": 3.872867587327377e-06, + "loss": 1.4843, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.680595874786377, + "learning_rate": 3.870836718115354e-06, + "loss": 1.5339, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.5908069610595703, + "learning_rate": 3.868805848903331e-06, + "loss": 1.5049, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.680586338043213, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.5188, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.763528823852539, + "learning_rate": 3.864744110479286e-06, + "loss": 1.5395, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.8522050380706787, + "learning_rate": 3.862713241267262e-06, + "loss": 1.562, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.806962490081787, + "learning_rate": 3.86068237205524e-06, + "loss": 1.5419, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.9716053009033203, + "learning_rate": 3.858651502843217e-06, + "loss": 1.5691, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.7608845233917236, + "eval_runtime": 37.1433, + "eval_samples_per_second": 26.923, + "eval_steps_per_second": 1.131, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.613098382949829, + "learning_rate": 3.856620633631195e-06, + "loss": 1.5297, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.5690557956695557, + "learning_rate": 3.854589764419171e-06, + "loss": 1.4964, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.8001089096069336, + "learning_rate": 3.852558895207149e-06, + "loss": 1.5253, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.6715309619903564, + "learning_rate": 3.850528025995126e-06, + "loss": 1.601, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.7044479846954346, + "learning_rate": 3.848497156783104e-06, + "loss": 1.5526, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.7368102073669434, + "learning_rate": 3.84646628757108e-06, + "loss": 1.5088, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.9358696937561035, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.5202, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.858172655105591, + "learning_rate": 3.842404549147035e-06, + "loss": 1.5378, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.6723833084106445, + "learning_rate": 3.840373679935013e-06, + "loss": 1.5134, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.748187780380249, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.4797, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.75319242477417, + "eval_runtime": 37.1302, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.4821436405181885, + "learning_rate": 3.836311941510967e-06, + "loss": 1.5382, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.5585227012634277, + "learning_rate": 3.834281072298944e-06, + "loss": 1.5373, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.927334785461426, + "learning_rate": 3.832250203086921e-06, + "loss": 1.5447, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.6096436977386475, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.5062, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.677304744720459, + "learning_rate": 3.828188464662876e-06, + "loss": 1.5671, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.9606077671051025, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6138, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.6971657276153564, + "learning_rate": 3.82412672623883e-06, + "loss": 1.5544, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.6380839347839355, + "learning_rate": 3.822095857026808e-06, + "loss": 1.4964, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.7537858486175537, + "learning_rate": 3.820064987814785e-06, + "loss": 1.5429, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.6189136505126953, + "learning_rate": 3.818034118602762e-06, + "loss": 1.4956, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.7546342611312866, + "eval_runtime": 37.0828, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5744292736053467, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.5256, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.5503833293914795, + "learning_rate": 3.813972380178717e-06, + "loss": 1.5211, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.979384422302246, + "learning_rate": 3.811941510966694e-06, + "loss": 1.55, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.867161512374878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5655, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.8006832599639893, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.4902, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.594712734222412, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.4451, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.6632349491119385, + "learning_rate": 3.803818034118603e-06, + "loss": 1.531, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.859894275665283, + "learning_rate": 3.80178716490658e-06, + "loss": 1.5018, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.6405653953552246, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.5765, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.7285192012786865, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.5537, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.7618242502212524, + "eval_runtime": 37.0898, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.6882379055023193, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.5143, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.706550121307373, + "learning_rate": 3.793663688058489e-06, + "loss": 1.4857, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.790907382965088, + "learning_rate": 3.791632818846467e-06, + "loss": 1.5982, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.631096363067627, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.5656, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.6981775760650635, + "learning_rate": 3.787571080422421e-06, + "loss": 1.567, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.612086296081543, + "learning_rate": 3.785540211210398e-06, + "loss": 1.5723, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.718630075454712, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.5565, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.6265041828155518, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.5326, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.578124523162842, + "learning_rate": 3.77944760357433e-06, + "loss": 1.533, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.832695484161377, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6133, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.7516050338745117, + "eval_runtime": 37.3173, + "eval_samples_per_second": 26.797, + "eval_steps_per_second": 1.125, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.851625442504883, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.5836, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.773354995938262e-06, + "loss": 1.5069, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.5738255977630615, + "learning_rate": 3.771324126726239e-06, + "loss": 1.5531, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.583575963973999, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.5589, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.7729389667510986, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.5901, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.6322245597839355, + "learning_rate": 3.765231519090171e-06, + "loss": 1.5423, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.6382699012756348, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.4574, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.7393112182617188, + "learning_rate": 3.761169780666125e-06, + "loss": 1.5351, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.684398651123047, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.4873, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.89401912689209, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.5358, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.7523635625839233, + "eval_runtime": 37.3763, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.629870891571045, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.5566, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.6936254501342773, + "learning_rate": 3.753046303818034e-06, + "loss": 1.5016, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.5889103412628174, + "learning_rate": 3.751015434606012e-06, + "loss": 1.5775, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.610703468322754, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.5508, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6318750381469727, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.5892, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.6650729179382324, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.5106, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.577174186706543, + "learning_rate": 3.742891957757921e-06, + "loss": 1.54, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.7274489402770996, + "learning_rate": 3.740861088545898e-06, + "loss": 1.5803, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6849985122680664, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.5393, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.811479091644287, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.5862, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.7490049600601196, + "eval_runtime": 36.9083, + "eval_samples_per_second": 27.094, + "eval_steps_per_second": 1.138, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.640002489089966, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.5149, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.6350319385528564, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5132, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.689814567565918, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.473, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.7181100845336914, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.4353, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.760629415512085, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.4859, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.9322562217712402, + "learning_rate": 3.724614134849716e-06, + "loss": 1.4733, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.280315399169922, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.5769, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 3.1343884468078613, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.4766, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.8537609577178955, + "learning_rate": 3.718521527213648e-06, + "loss": 1.5586, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.895299196243286, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5232, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.7473852634429932, + "eval_runtime": 37.006, + "eval_samples_per_second": 27.023, + "eval_steps_per_second": 1.135, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.82454776763916, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5057, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.711296558380127, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.4187, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.6281023025512695, + "learning_rate": 3.710398050365557e-06, + "loss": 1.4997, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.775632381439209, + "learning_rate": 3.708367181153534e-06, + "loss": 1.5087, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.8021788597106934, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.474, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 3.09647536277771, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5015, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.810450315475464, + "learning_rate": 3.702274573517466e-06, + "loss": 1.4681, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.65166974067688, + "learning_rate": 3.700243704305443e-06, + "loss": 1.4484, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.755690574645996, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5084, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.8223068714141846, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.5015, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.7516940832138062, + "eval_runtime": 37.0227, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.8401107788085938, + "learning_rate": 3.694151096669375e-06, + "loss": 1.4597, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 3.010011672973633, + "learning_rate": 3.692120227457352e-06, + "loss": 1.5625, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8575570583343506, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.4828, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.8527276515960693, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5074, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.81260085105896, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.4132, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.988429069519043, + "learning_rate": 3.683996750609261e-06, + "loss": 1.4803, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.964179515838623, + "learning_rate": 3.681965881397238e-06, + "loss": 1.4871, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.7151682376861572, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.5061, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.764187812805176, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.5188, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.810554265975952, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.4779, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.7422841787338257, + "eval_runtime": 37.0641, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.7640810012817383, + "learning_rate": 3.673842404549147e-06, + "loss": 1.4827, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.7683937549591064, + "learning_rate": 3.671811535337125e-06, + "loss": 1.4775, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.86434268951416, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.4752, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 3.0976617336273193, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.5099, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 3.107388496398926, + "learning_rate": 3.665718927701056e-06, + "loss": 1.5646, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 3.032921314239502, + "learning_rate": 3.663688058489034e-06, + "loss": 1.5048, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.9377830028533936, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5333, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.706542730331421, + "learning_rate": 3.659626320064988e-06, + "loss": 1.4903, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.819936513900757, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.4441, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.805360794067383, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.5122, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.747745156288147, + "eval_runtime": 37.0801, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.848580837249756, + "learning_rate": 3.65353371242892e-06, + "loss": 1.4984, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.772721290588379, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5026, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.760772943496704, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.4005, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.937507390975952, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.5131, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.923210859298706, + "learning_rate": 3.645410235580829e-06, + "loss": 1.4661, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8029167652130127, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.518, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.752716541290283, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.5109, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.7767865657806396, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.4831, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8483617305755615, + "learning_rate": 3.637286758732738e-06, + "loss": 1.4872, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.8714234828948975, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.5109, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.7455583810806274, + "eval_runtime": 37.0892, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7373011112213135, + "learning_rate": 3.633225020308692e-06, + "loss": 1.55, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.942373752593994, + "learning_rate": 3.63119415109667e-06, + "loss": 1.5079, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.8790931701660156, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.4669, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.9564976692199707, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5555, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.811699390411377, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.4457, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.926544427871704, + "learning_rate": 3.623070674248579e-06, + "loss": 1.4368, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.8694913387298584, + "learning_rate": 3.621039805036556e-06, + "loss": 1.4875, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 3.0294361114501953, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.4815, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8643198013305664, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.5031, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 3.0244719982147217, + "learning_rate": 3.614947197400488e-06, + "loss": 1.4914, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.7504756450653076, + "eval_runtime": 37.0238, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.8641672134399414, + "learning_rate": 3.612916328188465e-06, + "loss": 1.5122, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.8713624477386475, + "learning_rate": 3.610885458976442e-06, + "loss": 1.43, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.990912914276123, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.5124, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.963945150375366, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.533, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.802168846130371, + "learning_rate": 3.604792851340374e-06, + "loss": 1.4531, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 3.0556633472442627, + "learning_rate": 3.602761982128351e-06, + "loss": 1.4618, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.95744252204895, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5356, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.813145399093628, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.5134, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.8382728099823, + "learning_rate": 3.596669374492283e-06, + "loss": 1.4413, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.821100950241089, + "learning_rate": 3.59463850528026e-06, + "loss": 1.4748, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.7488586902618408, + "eval_runtime": 37.0361, + "eval_samples_per_second": 27.001, + "eval_steps_per_second": 1.134, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.930708408355713, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.5177, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.9312615394592285, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.4493, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 3.0051841735839844, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.5208, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.793515205383301, + "learning_rate": 3.586515028432169e-06, + "loss": 1.4813, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.7595956325531006, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5379, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.992871046066284, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.5153, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.7270801067352295, + "learning_rate": 3.580422420796101e-06, + "loss": 1.4392, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.797985076904297, + "learning_rate": 3.578391551584078e-06, + "loss": 1.4728, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9595260620117188, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.5158, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.8131442070007324, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.4214, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.7503563165664673, + "eval_runtime": 37.3831, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.124, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8242709636688232, + "learning_rate": 3.57229894394801e-06, + "loss": 1.5027, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.8973116874694824, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.5294, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 3.1207027435302734, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.4902, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.803910732269287, + "learning_rate": 3.566206336311942e-06, + "loss": 1.466, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.813541889190674, + "learning_rate": 3.564175467099919e-06, + "loss": 1.4957, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.9068171977996826, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.4435, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.8491995334625244, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.4432, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.8785598278045654, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.4791, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7701117992401123, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.4607, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.4821, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.7518969774246216, + "eval_runtime": 37.4808, + "eval_samples_per_second": 26.68, + "eval_steps_per_second": 1.121, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.5228, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.8513107299804688, + "learning_rate": 3.551990251827783e-06, + "loss": 1.4812, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.937448024749756, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.4538, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.8307411670684814, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.4807, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.7986767292022705, + "learning_rate": 3.545897644191714e-06, + "loss": 1.4706, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 3.1135106086730957, + "learning_rate": 3.543866774979692e-06, + "loss": 1.4574, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.7756283283233643, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5114, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.8114547729492188, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.4458, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 3.0544545650482178, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.4934, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8990671634674072, + "learning_rate": 3.535743298131601e-06, + "loss": 1.4808, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.7468736171722412, + "eval_runtime": 36.9135, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.976139545440674, + "learning_rate": 3.533712428919578e-06, + "loss": 1.5123, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.909146785736084, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.5204, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7225229740142822, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.4746, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.6145517826080322, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.5049, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.8660519123077393, + "learning_rate": 3.525588952071487e-06, + "loss": 1.5173, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.979801654815674, + "learning_rate": 3.523558082859464e-06, + "loss": 1.5146, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.7500879764556885, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.514, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.6923978328704834, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.455, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.6728951930999756, + "learning_rate": 3.517465475223396e-06, + "loss": 1.4712, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.769507884979248, + "learning_rate": 3.515434606011373e-06, + "loss": 1.444, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.7491599321365356, + "eval_runtime": 37.066, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.9768483638763428, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.5005, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.993196725845337, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.5347, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.886883497238159, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.4959, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.9152941703796387, + "learning_rate": 3.507311129163282e-06, + "loss": 1.4641, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.966872215270996, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.4232, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.9853367805480957, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.4883, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.9353625774383545, + "learning_rate": 3.501218521527214e-06, + "loss": 1.4826, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.7860453128814697, + "learning_rate": 3.499187652315191e-06, + "loss": 1.42, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.9269113540649414, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.5087, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 3.052962303161621, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5042, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.7421623468399048, + "eval_runtime": 37.103, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.7328579425811768, + "learning_rate": 3.493095044679123e-06, + "loss": 1.4375, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.876138925552368, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5559, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.7500345706939697, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.4043, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.836596965789795, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.4808, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.797060966491699, + "learning_rate": 3.484971567831032e-06, + "loss": 1.405, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.928368091583252, + "learning_rate": 3.482940698619009e-06, + "loss": 1.4765, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.923375368118286, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.5134, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.8624660968780518, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.4881, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.9593892097473145, + "learning_rate": 3.476848090982941e-06, + "loss": 1.4869, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.8019728660583496, + "learning_rate": 3.474817221770918e-06, + "loss": 1.468, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.7386831045150757, + "eval_runtime": 37.1034, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.8965771198272705, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.5604, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.8114798069000244, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.5092, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.9554526805877686, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.4869, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.981499195098877, + "learning_rate": 3.466693744922827e-06, + "loss": 1.4454, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.8626201152801514, + "learning_rate": 3.464662875710805e-06, + "loss": 1.4701, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.9752259254455566, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.5004, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 3.1553750038146973, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.4652, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.9458937644958496, + "learning_rate": 3.458570268074736e-06, + "loss": 1.51, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.9626898765563965, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.484, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.8093032836914062, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.468, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.740411639213562, + "eval_runtime": 37.1036, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.928344488143921, + "learning_rate": 3.452477660438668e-06, + "loss": 1.4474, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 3.0413129329681396, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.4759, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.7930357456207275, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.4406, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.8075618743896484, + "learning_rate": 3.4463850528026e-06, + "loss": 1.48, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 3.019402503967285, + "learning_rate": 3.444354183590577e-06, + "loss": 1.4713, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 3.009504795074463, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5124, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.9452531337738037, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.464, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.6980783939361572, + "learning_rate": 3.438261575954509e-06, + "loss": 1.4143, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 3.039795160293579, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.4712, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.285968542098999, + "learning_rate": 3.434199837530463e-06, + "loss": 1.5227, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.7402914762496948, + "eval_runtime": 37.1273, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 3.035813093185425, + "learning_rate": 3.432168968318441e-06, + "loss": 1.4772, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 3.1172401905059814, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.4389, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 3.0828707218170166, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.5098, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.922488212585449, + "learning_rate": 3.426076360682372e-06, + "loss": 1.5239, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 3.03352689743042, + "learning_rate": 3.42404549147035e-06, + "loss": 1.5296, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.933380603790283, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.4851, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7250027656555176, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.3966, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.991846799850464, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.4863, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 3.0704684257507324, + "learning_rate": 3.415922014622259e-06, + "loss": 1.5366, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 3.0659193992614746, + "learning_rate": 3.413891145410236e-06, + "loss": 1.516, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.7458266019821167, + "eval_runtime": 37.06, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.9695920944213867, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.508, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 3.004558801651001, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.5092, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.8377230167388916, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.4909, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.860020875930786, + "learning_rate": 3.405767668562145e-06, + "loss": 1.4415, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8683009147644043, + "learning_rate": 3.403736799350122e-06, + "loss": 1.4997, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.91894268989563, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.4886, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.8158814907073975, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.4295, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.9528515338897705, + "learning_rate": 3.397644191714054e-06, + "loss": 1.5422, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.994856119155884, + "learning_rate": 3.395613322502031e-06, + "loss": 1.4909, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.8976950645446777, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.4959, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.7350144386291504, + "eval_runtime": 37.0807, + "eval_samples_per_second": 26.968, + "eval_steps_per_second": 1.133, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.73504638671875, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.4454, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.816823720932007, + "learning_rate": 3.389520714865963e-06, + "loss": 1.4864, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.706554889678955, + "learning_rate": 3.38748984565394e-06, + "loss": 1.483, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.9004647731781006, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.492, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8677759170532227, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.4786, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.83447527885437, + "learning_rate": 3.381397238017872e-06, + "loss": 1.4575, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.9942550659179688, + "learning_rate": 3.379366368805849e-06, + "loss": 1.4746, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.124676465988159, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.5406, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.9151222705841064, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.4748, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.734571933746338, + "learning_rate": 3.373273761169781e-06, + "loss": 1.4773, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.7389248609542847, + "eval_runtime": 37.1519, + "eval_samples_per_second": 26.917, + "eval_steps_per_second": 1.13, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.8919990062713623, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.4698, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.873870611190796, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.4509, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 3.0072309970855713, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5024, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.784576177597046, + "learning_rate": 3.36515028432169e-06, + "loss": 1.4405, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.8316547870635986, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.4697, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.8758575916290283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.495, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.879115343093872, + "learning_rate": 3.359057676685622e-06, + "loss": 1.5248, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.9559383392333984, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5229, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.6783695220947266, + "learning_rate": 3.354995938261576e-06, + "loss": 1.4251, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.8372304439544678, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.4649, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.7335349321365356, + "eval_runtime": 37.3498, + "eval_samples_per_second": 26.774, + "eval_steps_per_second": 1.125, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.729576826095581, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.4263, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.7717669010162354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.4467, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.8166871070861816, + "learning_rate": 3.346872461413485e-06, + "loss": 1.4937, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.845677137374878, + "learning_rate": 3.344841592201463e-06, + "loss": 1.4303, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.9527740478515625, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.4651, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.810873031616211, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.4521, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.77546763420105, + "learning_rate": 3.338748984565394e-06, + "loss": 1.4857, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.9098589420318604, + "learning_rate": 3.336718115353372e-06, + "loss": 1.4988, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.797712802886963, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.5307, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.916668176651001, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.4828, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.7372808456420898, + "eval_runtime": 36.888, + "eval_samples_per_second": 27.109, + "eval_steps_per_second": 1.139, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.8220417499542236, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.4636, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.851010322570801, + "learning_rate": 3.32859463850528e-06, + "loss": 1.4579, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.983619213104248, + "learning_rate": 3.326563769293258e-06, + "loss": 1.4801, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.750732898712158, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.4438, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.631105422973633, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.4439, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.944300889968872, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.437, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8841664791107178, + "learning_rate": 3.318440292445167e-06, + "loss": 1.4201, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.917189121246338, + "learning_rate": 3.316409423233144e-06, + "loss": 1.4489, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.112077236175537, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.4449, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.9847209453582764, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.5153, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.7323338985443115, + "eval_runtime": 36.9935, + "eval_samples_per_second": 27.032, + "eval_steps_per_second": 1.135, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.948901414871216, + "learning_rate": 3.310316815597076e-06, + "loss": 1.473, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.84930419921875, + "learning_rate": 3.308285946385053e-06, + "loss": 1.4796, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.9120237827301025, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5392, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.7088463306427, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.4206, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.7495498657226562, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.4759, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.9146509170532227, + "learning_rate": 3.300162469536962e-06, + "loss": 1.51, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.781975030899048, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.4667, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.7524054050445557, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.5036, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.7672629356384277, + "learning_rate": 3.294069861900894e-06, + "loss": 1.4827, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.875802516937256, + "learning_rate": 3.292038992688871e-06, + "loss": 1.519, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.7273142337799072, + "eval_runtime": 37.0032, + "eval_samples_per_second": 27.025, + "eval_steps_per_second": 1.135, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.8699231147766113, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.474, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.8670589923858643, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.5039, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.9946320056915283, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5226, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.767071008682251, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.4221, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8807036876678467, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.486, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.687720537185669, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.3944, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.8170154094696045, + "learning_rate": 3.277822908204712e-06, + "loss": 1.5213, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.905557632446289, + "learning_rate": 3.275792038992689e-06, + "loss": 1.4598, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.643657922744751, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.4282, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.798712968826294, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.4752, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.7295492887496948, + "eval_runtime": 37.0432, + "eval_samples_per_second": 26.996, + "eval_steps_per_second": 1.134, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.8480567932128906, + "learning_rate": 3.269699431356621e-06, + "loss": 1.4322, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.8311381340026855, + "learning_rate": 3.267668562144598e-06, + "loss": 1.4744, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.83522891998291, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.4578, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.757307767868042, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.4469, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.9117610454559326, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.5034, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.8141143321990967, + "learning_rate": 3.259545085296507e-06, + "loss": 1.473, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 3.050138473510742, + "learning_rate": 3.257514216084484e-06, + "loss": 1.528, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.660510540008545, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.4335, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.898832082748413, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.5335, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.955589532852173, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.4832, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.7293461561203003, + "eval_runtime": 37.0096, + "eval_samples_per_second": 27.02, + "eval_steps_per_second": 1.135, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.8524296283721924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.4728, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.9741404056549072, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.4805, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.738152265548706, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.4532, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.717066764831543, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.4741, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.8367488384246826, + "learning_rate": 3.241267262388302e-06, + "loss": 1.4773, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.987780809402466, + "learning_rate": 3.23923639317628e-06, + "loss": 1.4214, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.705444097518921, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.4537, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.8966188430786133, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.4552, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.9027600288391113, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.4883, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.965813159942627, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5318, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.7236050367355347, + "eval_runtime": 37.0306, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.6390798091888428, + "learning_rate": 3.229082047116166e-06, + "loss": 1.4431, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.822848081588745, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5163, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7289068698883057, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.4551, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.766545057296753, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.4648, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.8657522201538086, + "learning_rate": 3.220958570268075e-06, + "loss": 1.4825, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.7456932067871094, + "learning_rate": 3.218927701056052e-06, + "loss": 1.3909, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 3.0274853706359863, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.4461, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.990029811859131, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.4057, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 3.086756706237793, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5251, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 3.0227134227752686, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.427, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.7248852252960205, + "eval_runtime": 37.0619, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.815480947494507, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.469, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.9171998500823975, + "learning_rate": 3.206742485783916e-06, + "loss": 1.4977, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 3.087686777114868, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.4233, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.7349331378936768, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.4187, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.9189705848693848, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.4949, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 3.0483345985412598, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5041, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.7958145141601562, + "learning_rate": 3.196588139723802e-06, + "loss": 1.4009, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 3.081714153289795, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.4718, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.7897303104400635, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.4926, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.899045944213867, + "learning_rate": 3.190495532087734e-06, + "loss": 1.4524, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.7273337841033936, + "eval_runtime": 37.0682, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.979063034057617, + "learning_rate": 3.188464662875711e-06, + "loss": 1.4706, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.8633804321289062, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5273, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.8997795581817627, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.4638, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.906151294708252, + "learning_rate": 3.182372055239643e-06, + "loss": 1.4187, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 3.0212671756744385, + "learning_rate": 3.18034118602762e-06, + "loss": 1.4213, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 3.0891034603118896, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.4921, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.9302260875701904, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.4787, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.8741140365600586, + "learning_rate": 3.174248578391552e-06, + "loss": 1.4591, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 3.0567023754119873, + "learning_rate": 3.172217709179529e-06, + "loss": 1.4636, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.80245304107666, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.4523, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.7270530462265015, + "eval_runtime": 37.325, + "eval_samples_per_second": 26.792, + "eval_steps_per_second": 1.125, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.816023349761963, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.4579, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.7405545711517334, + "learning_rate": 3.166125101543461e-06, + "loss": 1.4225, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.9569272994995117, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.483, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 3.0433337688446045, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.4846, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.856781244277954, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.426, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.8880672454833984, + "learning_rate": 3.15800162469537e-06, + "loss": 1.4764, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 3.0363011360168457, + "learning_rate": 3.155970755483347e-06, + "loss": 1.4916, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 3.0171539783477783, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.4596, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 3.183234930038452, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.4936, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.802027463912964, + "learning_rate": 3.149878147847279e-06, + "loss": 1.4733, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.7253632545471191, + "eval_runtime": 37.4737, + "eval_samples_per_second": 26.685, + "eval_steps_per_second": 1.121, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.9667463302612305, + "learning_rate": 3.147847278635256e-06, + "loss": 1.508, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 3.1557469367980957, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.5459, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.7861125469207764, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.4614, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.7431211471557617, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.4297, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 3.140911340713501, + "learning_rate": 3.139723801787165e-06, + "loss": 1.4497, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.970118522644043, + "learning_rate": 3.137692932575143e-06, + "loss": 1.4409, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.94400954246521, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.4889, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.925431966781616, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.4205, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.997279167175293, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.4714, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.855996608734131, + "learning_rate": 3.129569455727051e-06, + "loss": 1.4925, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.7247990369796753, + "eval_runtime": 36.9312, + "eval_samples_per_second": 27.077, + "eval_steps_per_second": 1.137, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 3.011199951171875, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.5132, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.8528099060058594, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.4791, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.8827919960021973, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.495, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.8150227069854736, + "learning_rate": 3.12144597887896e-06, + "loss": 1.4563, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.9341959953308105, + "learning_rate": 3.119415109666938e-06, + "loss": 1.4829, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.9161651134490967, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.492, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.906963348388672, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.445, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.992582321166992, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.4988, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 3.005469799041748, + "learning_rate": 3.111291632818847e-06, + "loss": 1.4999, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.7577736377716064, + "learning_rate": 3.109260763606824e-06, + "loss": 1.5152, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.724539041519165, + "eval_runtime": 37.0601, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.870109796524048, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.4746, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.7745094299316406, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.4891, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.6818361282348633, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.4575, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.823237180709839, + "learning_rate": 3.101137286758733e-06, + "loss": 1.492, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.9089932441711426, + "learning_rate": 3.09910641754671e-06, + "loss": 1.4943, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.85476016998291, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.5049, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 3.07242488861084, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.4709, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 3.117357015609741, + "learning_rate": 3.093013809910642e-06, + "loss": 1.52, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.959383964538574, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.4775, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8594462871551514, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.5113, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.721728801727295, + "eval_runtime": 37.1106, + "eval_samples_per_second": 26.946, + "eval_steps_per_second": 1.132, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.8418593406677246, + "learning_rate": 3.086921202274574e-06, + "loss": 1.482, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.762467384338379, + "learning_rate": 3.084890333062551e-06, + "loss": 1.4672, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.8023135662078857, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.431, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.7467143535614014, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.4837, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.6832900047302246, + "learning_rate": 3.078797725426483e-06, + "loss": 1.4765, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.863823175430298, + "learning_rate": 3.07676685621446e-06, + "loss": 1.4039, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.9055254459381104, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.4655, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.8665876388549805, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.449, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 3.029784917831421, + "learning_rate": 3.070674248578392e-06, + "loss": 1.4665, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.8582639694213867, + "learning_rate": 3.068643379366369e-06, + "loss": 1.4832, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.722991704940796, + "eval_runtime": 37.0827, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.8655312061309814, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.4759, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.892972469329834, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.4363, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.7701165676116943, + "learning_rate": 3.062550771730301e-06, + "loss": 1.4798, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7921078205108643, + "learning_rate": 3.060519902518278e-06, + "loss": 1.4557, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.8233699798583984, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.4598, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.8705663681030273, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.3898, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.9667348861694336, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.4865, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 3.0962018966674805, + "learning_rate": 3.052396425670187e-06, + "loss": 1.4512, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 3.06290340423584, + "learning_rate": 3.050365556458164e-06, + "loss": 1.5129, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.8010520935058594, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.4525, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.7228375673294067, + "eval_runtime": 37.1149, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.73380446434021, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.4537, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.9469099044799805, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.4624, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.8733043670654297, + "learning_rate": 3.042242079610073e-06, + "loss": 1.4981, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.867926597595215, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.4966, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.8584275245666504, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.4161, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.8059520721435547, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.4511, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.916750192642212, + "learning_rate": 3.034118602761982e-06, + "loss": 1.4694, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.979952335357666, + "learning_rate": 3.03208773354996e-06, + "loss": 1.4624, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.807194709777832, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.4227, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.880439519882202, + "learning_rate": 3.028025995125914e-06, + "loss": 1.5094, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.7204533815383911, + "eval_runtime": 37.1346, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.7696263790130615, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.492, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.840989351272583, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.4229, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.948915481567383, + "learning_rate": 3.021933387489846e-06, + "loss": 1.4601, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.9163503646850586, + "learning_rate": 3.019902518277823e-06, + "loss": 1.4218, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 3.0191619396209717, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.4648, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.9332194328308105, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.4791, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.8832461833953857, + "learning_rate": 3.013809910641755e-06, + "loss": 1.43, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.886837959289551, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.4455, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.8201355934143066, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.4417, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.9708259105682373, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.4657, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.7213743925094604, + "eval_runtime": 37.0993, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.927104949951172, + "learning_rate": 3.005686433793664e-06, + "loss": 1.4406, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.9701600074768066, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.4692, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 3.0241730213165283, + "learning_rate": 3.001624695369618e-06, + "loss": 1.4336, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.8802919387817383, + "learning_rate": 2.999593826157596e-06, + "loss": 1.4948, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.7666738033294678, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.4518, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 3.1206812858581543, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.4707, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 3.069323778152466, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.4534, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 3.155106544494629, + "learning_rate": 2.991470349309505e-06, + "loss": 1.5153, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.923816680908203, + "learning_rate": 2.989439480097482e-06, + "loss": 1.4606, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 3.2247867584228516, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.4564, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.7207131385803223, + "eval_runtime": 37.0684, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.8193411827087402, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.4721, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.865691661834717, + "learning_rate": 2.983346872461414e-06, + "loss": 1.5072, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 3.125004529953003, + "learning_rate": 2.981316003249391e-06, + "loss": 1.4913, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.873455286026001, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.4699, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.986186981201172, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.4808, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.8694097995758057, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.495, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.8555796146392822, + "learning_rate": 2.9731925264013e-06, + "loss": 1.4481, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.8808774948120117, + "learning_rate": 2.971161657189277e-06, + "loss": 1.4361, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.8831286430358887, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.4489, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.7533552646636963, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.4386, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.7256641387939453, + "eval_runtime": 37.1621, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 2.783586025238037, + "learning_rate": 2.965069049553209e-06, + "loss": 1.4476, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.9043776988983154, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5077, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.838284492492676, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.4287, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.94248366355896, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.4987, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.9397339820861816, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.4809, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.8922536373138428, + "learning_rate": 2.954914703493095e-06, + "loss": 1.426, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.787680149078369, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.4549, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.8116507530212402, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.4324, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.7330825328826904, + "learning_rate": 2.948822095857027e-06, + "loss": 1.4244, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.806539535522461, + "learning_rate": 2.946791226645004e-06, + "loss": 1.4604, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.7229852676391602, + "eval_runtime": 37.378, + "eval_samples_per_second": 26.754, + "eval_steps_per_second": 1.124, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.9335062503814697, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.4433, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.8057737350463867, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.4242, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 2.7581822872161865, + "learning_rate": 2.940698619008936e-06, + "loss": 1.4374, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.931168794631958, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.4588, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.865867853164673, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.4731, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.8826231956481934, + "learning_rate": 2.934606011372868e-06, + "loss": 1.4568, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.922809600830078, + "learning_rate": 2.932575142160845e-06, + "loss": 1.4747, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.9709854125976562, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.4631, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8259403705596924, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.449, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.850677490234375, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.4725, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.7200536727905273, + "eval_runtime": 36.8614, + "eval_samples_per_second": 27.129, + "eval_steps_per_second": 1.139, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.8449628353118896, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.5022, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 3.004880666732788, + "learning_rate": 2.922420796100731e-06, + "loss": 1.4606, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.737196445465088, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.3911, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.812662124633789, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.4873, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.7756738662719727, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.3975, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.8784940242767334, + "learning_rate": 2.91429731925264e-06, + "loss": 1.4726, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 2.946704626083374, + "learning_rate": 2.912266450040618e-06, + "loss": 1.4841, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.8813061714172363, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.4656, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.9221038818359375, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.4505, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 2.9353513717651367, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.4585, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.7200087308883667, + "eval_runtime": 36.9952, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.7565391063690186, + "learning_rate": 2.904142973192527e-06, + "loss": 1.4199, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 2.824780225753784, + "learning_rate": 2.902112103980504e-06, + "loss": 1.5003, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.963022232055664, + "learning_rate": 2.900081234768481e-06, + "loss": 1.4649, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.744490385055542, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.4677, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.8483076095581055, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.4291, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.878147602081299, + "learning_rate": 2.893988627132413e-06, + "loss": 1.4278, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.866380214691162, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.4491, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.802532911300659, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.397, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.863152265548706, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.4573, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.800173044204712, + "learning_rate": 2.885865150284322e-06, + "loss": 1.4466, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.7167619466781616, + "eval_runtime": 37.0136, + "eval_samples_per_second": 27.017, + "eval_steps_per_second": 1.135, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.975752830505371, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.4918, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.7768948078155518, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.4549, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.761004686355591, + "learning_rate": 2.879772542648254e-06, + "loss": 1.4039, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.9454665184020996, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.4542, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.85044264793396, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.3875, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.9699220657348633, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.5084, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.8547441959381104, + "learning_rate": 2.871649065800163e-06, + "loss": 1.4839, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 3.0871565341949463, + "learning_rate": 2.86961819658814e-06, + "loss": 1.5105, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 3.662945032119751, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.4483, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.916095495223999, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.455, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.7137683629989624, + "eval_runtime": 37.0176, + "eval_samples_per_second": 27.014, + "eval_steps_per_second": 1.135, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.857478141784668, + "learning_rate": 2.863525588952072e-06, + "loss": 1.4531, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.9542508125305176, + "learning_rate": 2.861494719740049e-06, + "loss": 1.488, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.677672863006592, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.3725, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.8790128231048584, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.4252, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.95095157623291, + "learning_rate": 2.855402112103981e-06, + "loss": 1.4532, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.997403383255005, + "learning_rate": 2.853371242891958e-06, + "loss": 1.4581, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.926973581314087, + "learning_rate": 2.851340373679935e-06, + "loss": 1.4368, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.8602488040924072, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.4427, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.875828504562378, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.4757, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.825498580932617, + "learning_rate": 2.845247766043867e-06, + "loss": 1.4535, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.7157984972000122, + "eval_runtime": 37.0329, + "eval_samples_per_second": 27.003, + "eval_steps_per_second": 1.134, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.8068923950195312, + "learning_rate": 2.843216896831844e-06, + "loss": 1.4391, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.8095016479492188, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.397, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.7768023014068604, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.4415, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.7868361473083496, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.4726, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 3.085806131362915, + "learning_rate": 2.835093419983753e-06, + "loss": 1.4735, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.972489833831787, + "learning_rate": 2.833062550771731e-06, + "loss": 1.4415, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.9990909099578857, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.4203, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.849210023880005, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.4202, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.899545907974243, + "learning_rate": 2.826969943135662e-06, + "loss": 1.4541, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.7620913982391357, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.4672, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.718942642211914, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.8633880615234375, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.4548, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.809718370437622, + "learning_rate": 2.820877335499594e-06, + "loss": 1.4379, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.780914306640625, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.4244, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.6853432655334473, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.3999, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.8040759563446045, + "learning_rate": 2.814784727863526e-06, + "loss": 1.4151, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.974795341491699, + "learning_rate": 2.812753858651503e-06, + "loss": 1.4497, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.9964492321014404, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.4512, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 2.9635047912597656, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.4104, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 3.0267832279205322, + "learning_rate": 2.806661251015435e-06, + "loss": 1.4149, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 3.0778963565826416, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.5034, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.723001480102539, + "eval_runtime": 37.0513, + "eval_samples_per_second": 26.99, + "eval_steps_per_second": 1.134, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.9889729022979736, + "learning_rate": 2.802599512591389e-06, + "loss": 1.4851, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 3.111581325531006, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.5337, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.727510452270508, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.4647, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.848571300506592, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.4107, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 3.178166627883911, + "learning_rate": 2.794476035743298e-06, + "loss": 1.4427, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.9924423694610596, + "learning_rate": 2.792445166531276e-06, + "loss": 1.5024, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.825277328491211, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.4687, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 3.082688808441162, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.5057, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.8399462699890137, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.4274, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.8092541694641113, + "learning_rate": 2.784321689683185e-06, + "loss": 1.4157, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.7190473079681396, + "eval_runtime": 37.0662, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.8798582553863525, + "learning_rate": 2.782290820471162e-06, + "loss": 1.4213, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.7891805171966553, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.4181, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.865835666656494, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.4896, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.996018171310425, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.4829, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.868534803390503, + "learning_rate": 2.774167343623071e-06, + "loss": 1.4592, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.906359910964966, + "learning_rate": 2.772136474411048e-06, + "loss": 1.5205, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 3.099578857421875, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.4246, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.944490432739258, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.3856, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.870277166366577, + "learning_rate": 2.76604386677498e-06, + "loss": 1.4648, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.853956937789917, + "learning_rate": 2.764012997562957e-06, + "loss": 1.4317, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.7165414094924927, + "eval_runtime": 37.4056, + "eval_samples_per_second": 26.734, + "eval_steps_per_second": 1.123, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 3.0553836822509766, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.4649, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.973536729812622, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.4129, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 3.025920867919922, + "learning_rate": 2.757920389926889e-06, + "loss": 1.4974, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.853586196899414, + "learning_rate": 2.755889520714866e-06, + "loss": 1.423, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 3.0477190017700195, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.4149, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.8930039405822754, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.3938, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.8823812007904053, + "learning_rate": 2.749796913078798e-06, + "loss": 1.46, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.96528959274292, + "learning_rate": 2.747766043866775e-06, + "loss": 1.4818, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.892967939376831, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.4392, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.851961851119995, + "learning_rate": 2.7437043054427297e-06, + "loss": 1.4903, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.7118008136749268, + "eval_runtime": 37.4958, + "eval_samples_per_second": 26.67, + "eval_steps_per_second": 1.12, + "step": 1400 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 4.364233173106688e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/latest new file mode 100644 index 0000000000000000000000000000000000000000..cb24338af451be6d3ccddc18a950b58a70589761 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/latest @@ -0,0 +1 @@ +global_step1700 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..772eb5f12e0907e43a911d09cc57314758cbe6ad --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/trainer_state.json @@ -0,0 +1,13293 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.8630136986301369, + "eval_steps": 10, + "global_step": 1700, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6624906063079834, + "learning_rate": 4.77051177904143e-06, + "loss": 1.704, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 3.0075268745422363, + "learning_rate": 4.768480909829407e-06, + "loss": 1.6322, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.6070828437805176, + "learning_rate": 4.766450040617385e-06, + "loss": 1.6871, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 3.044222831726074, + "learning_rate": 4.764419171405362e-06, + "loss": 1.6652, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.773167610168457, + "learning_rate": 4.762388302193339e-06, + "loss": 1.7063, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.880176305770874, + "learning_rate": 4.760357432981316e-06, + "loss": 1.6775, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.6402764320373535, + "learning_rate": 4.758326563769294e-06, + "loss": 1.6694, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.9245846271514893, + "learning_rate": 4.756295694557271e-06, + "loss": 1.6688, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.623044967651367, + "learning_rate": 4.754264825345248e-06, + "loss": 1.6819, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.739896059036255, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.6401, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.8707164525985718, + "eval_runtime": 36.9513, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.603599786758423, + "learning_rate": 4.750203086921203e-06, + "loss": 1.6412, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.5571675300598145, + "learning_rate": 4.74817221770918e-06, + "loss": 1.5779, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.87091326713562, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.6878, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.8140673637390137, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7023, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.6550583839416504, + "learning_rate": 4.742079610073111e-06, + "loss": 1.5995, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.6259925365448, + "learning_rate": 4.740048740861089e-06, + "loss": 1.6194, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.971848726272583, + "learning_rate": 4.738017871649066e-06, + "loss": 1.6634, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6344778537750244, + "learning_rate": 4.735987002437044e-06, + "loss": 1.6502, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 2.9362456798553467, + "learning_rate": 4.73395613322502e-06, + "loss": 1.6948, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.8098835945129395, + "learning_rate": 4.731925264012998e-06, + "loss": 1.7161, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.8724393844604492, + "eval_runtime": 37.0876, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.933009386062622, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7014, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.8216991424560547, + "learning_rate": 4.727863525588953e-06, + "loss": 1.6192, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.7028493881225586, + "learning_rate": 4.725832656376929e-06, + "loss": 1.607, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.816818952560425, + "learning_rate": 4.723801787164907e-06, + "loss": 1.6183, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.7707865238189697, + "learning_rate": 4.721770917952885e-06, + "loss": 1.6795, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.779562473297119, + "learning_rate": 4.719740048740861e-06, + "loss": 1.63, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.907338857650757, + "learning_rate": 4.717709179528839e-06, + "loss": 1.644, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 3.073558807373047, + "learning_rate": 4.715678310316816e-06, + "loss": 1.6821, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.7556493282318115, + "learning_rate": 4.713647441104794e-06, + "loss": 1.6501, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6718344688415527, + "learning_rate": 4.71161657189277e-06, + "loss": 1.6225, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.854687213897705, + "eval_runtime": 36.9773, + "eval_samples_per_second": 27.044, + "eval_steps_per_second": 1.136, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.763335943222046, + "learning_rate": 4.709585702680748e-06, + "loss": 1.6311, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.7669479846954346, + "learning_rate": 4.707554833468725e-06, + "loss": 1.6809, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.711562395095825, + "learning_rate": 4.705523964256703e-06, + "loss": 1.741, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.799954652786255, + "learning_rate": 4.703493095044679e-06, + "loss": 1.6409, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.7278690338134766, + "learning_rate": 4.701462225832657e-06, + "loss": 1.6325, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.6802427768707275, + "learning_rate": 4.699431356620634e-06, + "loss": 1.5988, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.8826160430908203, + "learning_rate": 4.697400487408612e-06, + "loss": 1.7119, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.784343719482422, + "learning_rate": 4.695369618196588e-06, + "loss": 1.6855, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.683337688446045, + "learning_rate": 4.693338748984566e-06, + "loss": 1.6588, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.6646857261657715, + "learning_rate": 4.691307879772543e-06, + "loss": 1.623, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.8511513471603394, + "eval_runtime": 36.9852, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.8181183338165283, + "learning_rate": 4.68927701056052e-06, + "loss": 1.6171, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.8270857334136963, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.678, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.7685446739196777, + "learning_rate": 4.685215272136475e-06, + "loss": 1.6575, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.7003183364868164, + "learning_rate": 4.683184402924452e-06, + "loss": 1.6405, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.668750286102295, + "learning_rate": 4.681153533712429e-06, + "loss": 1.6541, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 3.065800905227661, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7288, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.615601062774658, + "learning_rate": 4.677091795288384e-06, + "loss": 1.6306, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.5948891639709473, + "learning_rate": 4.675060926076361e-06, + "loss": 1.6018, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.7550079822540283, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7009, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.868687868118286, + "learning_rate": 4.670999187652316e-06, + "loss": 1.6034, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.8473671674728394, + "eval_runtime": 37.0015, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.705962657928467, + "learning_rate": 4.668968318440293e-06, + "loss": 1.6084, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.7901813983917236, + "learning_rate": 4.66693744922827e-06, + "loss": 1.6171, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.688774824142456, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.6774, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.734617233276367, + "learning_rate": 4.662875710804225e-06, + "loss": 1.5796, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.742017984390259, + "learning_rate": 4.660844841592202e-06, + "loss": 1.6671, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.6438794136047363, + "learning_rate": 4.658813972380179e-06, + "loss": 1.5835, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6441051959991455, + "learning_rate": 4.656783103168157e-06, + "loss": 1.6909, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.824204206466675, + "learning_rate": 4.654752233956134e-06, + "loss": 1.5758, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.556638479232788, + "learning_rate": 4.652721364744111e-06, + "loss": 1.6504, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.897812843322754, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.6406, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.8537043333053589, + "eval_runtime": 37.257, + "eval_samples_per_second": 26.841, + "eval_steps_per_second": 1.127, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.6594088077545166, + "learning_rate": 4.648659626320066e-06, + "loss": 1.6966, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.79524302482605, + "learning_rate": 4.646628757108043e-06, + "loss": 1.6645, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.80291748046875, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.6281, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.7723848819732666, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.5959, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 3.0625650882720947, + "learning_rate": 4.640536149471974e-06, + "loss": 1.6313, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.775726079940796, + "learning_rate": 4.638505280259952e-06, + "loss": 1.6563, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.837758779525757, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.6727, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 3.041355609893799, + "learning_rate": 4.634443541835907e-06, + "loss": 1.6282, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.5775389671325684, + "learning_rate": 4.632412672623883e-06, + "loss": 1.651, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 3.026287794113159, + "learning_rate": 4.630381803411861e-06, + "loss": 1.6458, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.8461565971374512, + "eval_runtime": 37.3881, + "eval_samples_per_second": 26.747, + "eval_steps_per_second": 1.123, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.6781764030456543, + "learning_rate": 4.628350934199838e-06, + "loss": 1.5232, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.763350486755371, + "learning_rate": 4.626320064987816e-06, + "loss": 1.6186, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.7063372135162354, + "learning_rate": 4.624289195775792e-06, + "loss": 1.6784, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.824563980102539, + "learning_rate": 4.62225832656377e-06, + "loss": 1.6219, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.5335569381713867, + "learning_rate": 4.620227457351747e-06, + "loss": 1.6152, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.773644208908081, + "learning_rate": 4.618196588139724e-06, + "loss": 1.6179, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.9841721057891846, + "learning_rate": 4.616165718927701e-06, + "loss": 1.6258, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.684171676635742, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.6066, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.8817965984344482, + "learning_rate": 4.612103980503656e-06, + "loss": 1.6697, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.9846866130828857, + "learning_rate": 4.610073111291633e-06, + "loss": 1.6541, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.8485002517700195, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.904170513153076, + "learning_rate": 4.60804224207961e-06, + "loss": 1.6362, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.684476375579834, + "learning_rate": 4.606011372867588e-06, + "loss": 1.6303, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6692216396331787, + "learning_rate": 4.603980503655565e-06, + "loss": 1.638, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.73555588722229, + "learning_rate": 4.601949634443542e-06, + "loss": 1.6125, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.729525089263916, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.565, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.5972018241882324, + "learning_rate": 4.597887896019497e-06, + "loss": 1.5856, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7859137058258057, + "learning_rate": 4.595857026807474e-06, + "loss": 1.6325, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.662587881088257, + "learning_rate": 4.593826157595451e-06, + "loss": 1.6172, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.6380577087402344, + "learning_rate": 4.591795288383429e-06, + "loss": 1.6005, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8375349044799805, + "learning_rate": 4.589764419171406e-06, + "loss": 1.6507, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.833074688911438, + "eval_runtime": 37.0695, + "eval_samples_per_second": 26.976, + "eval_steps_per_second": 1.133, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.786713123321533, + "learning_rate": 4.587733549959383e-06, + "loss": 1.6283, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.6959030628204346, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.5994, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.6100692749023438, + "learning_rate": 4.583671811535338e-06, + "loss": 1.5884, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.7281863689422607, + "learning_rate": 4.581640942323315e-06, + "loss": 1.5737, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.7519376277923584, + "learning_rate": 4.579610073111292e-06, + "loss": 1.6531, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.9314866065979004, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.6755, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.9250576496124268, + "learning_rate": 4.575548334687247e-06, + "loss": 1.6604, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.6382815837860107, + "learning_rate": 4.573517465475224e-06, + "loss": 1.6773, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.6891751289367676, + "learning_rate": 4.571486596263201e-06, + "loss": 1.5849, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.7018136978149414, + "learning_rate": 4.569455727051178e-06, + "loss": 1.6358, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.838019847869873, + "eval_runtime": 37.0496, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.671952247619629, + "learning_rate": 4.567424857839156e-06, + "loss": 1.6207, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.882539749145508, + "learning_rate": 4.565393988627132e-06, + "loss": 1.6618, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.684070587158203, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.6502, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.5295751094818115, + "learning_rate": 4.561332250203087e-06, + "loss": 1.6078, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.0268194675445557, + "learning_rate": 4.559301380991065e-06, + "loss": 1.6312, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6895201206207275, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.5848, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.869950771331787, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.5951, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.909505605697632, + "learning_rate": 4.553208773354996e-06, + "loss": 1.6589, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.7604334354400635, + "learning_rate": 4.551177904142974e-06, + "loss": 1.5767, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.8792080879211426, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.6864, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.836847186088562, + "eval_runtime": 36.9142, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.607037305831909, + "learning_rate": 4.547116165718929e-06, + "loss": 1.5925, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.9338927268981934, + "learning_rate": 4.545085296506905e-06, + "loss": 1.6433, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.7723233699798584, + "learning_rate": 4.543054427294882e-06, + "loss": 1.5963, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.4122188091278076, + "learning_rate": 4.54102355808286e-06, + "loss": 1.5978, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.7582709789276123, + "learning_rate": 4.538992688870837e-06, + "loss": 1.581, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.6983039379119873, + "learning_rate": 4.536961819658814e-06, + "loss": 1.6253, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.6036365032196045, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.5957, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.759645700454712, + "learning_rate": 4.532900081234769e-06, + "loss": 1.6199, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.797744035720825, + "learning_rate": 4.530869212022746e-06, + "loss": 1.5854, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.752868890762329, + "learning_rate": 4.528838342810723e-06, + "loss": 1.619, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.8305871486663818, + "eval_runtime": 37.0307, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.7421889305114746, + "learning_rate": 4.526807473598701e-06, + "loss": 1.6182, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.6852707862854004, + "learning_rate": 4.524776604386678e-06, + "loss": 1.6135, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.861720323562622, + "learning_rate": 4.522745735174655e-06, + "loss": 1.6099, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.706851005554199, + "learning_rate": 4.520714865962632e-06, + "loss": 1.6597, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7992191314697266, + "learning_rate": 4.51868399675061e-06, + "loss": 1.6077, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.6248183250427246, + "learning_rate": 4.516653127538587e-06, + "loss": 1.6485, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.696040391921997, + "learning_rate": 4.514622258326564e-06, + "loss": 1.5853, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.5958261489868164, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.571, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.9715874195098877, + "learning_rate": 4.510560519902519e-06, + "loss": 1.6044, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.664834499359131, + "learning_rate": 4.508529650690496e-06, + "loss": 1.6568, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.8277279138565063, + "eval_runtime": 37.0489, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.869558811187744, + "learning_rate": 4.506498781478473e-06, + "loss": 1.5654, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.5814828872680664, + "learning_rate": 4.504467912266451e-06, + "loss": 1.5484, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.6158947944641113, + "learning_rate": 4.502437043054428e-06, + "loss": 1.6202, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.7424654960632324, + "learning_rate": 4.500406173842405e-06, + "loss": 1.6755, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.8670949935913086, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.6144, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.7914652824401855, + "learning_rate": 4.49634443541836e-06, + "loss": 1.6164, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.866380214691162, + "learning_rate": 4.494313566206336e-06, + "loss": 1.5961, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.966474771499634, + "learning_rate": 4.492282696994314e-06, + "loss": 1.6681, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.751291513442993, + "learning_rate": 4.490251827782291e-06, + "loss": 1.6338, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.58840012550354, + "learning_rate": 4.488220958570269e-06, + "loss": 1.5619, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.824889898300171, + "eval_runtime": 37.0821, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.786945104598999, + "learning_rate": 4.486190089358245e-06, + "loss": 1.5712, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4536845684051514, + "learning_rate": 4.484159220146223e-06, + "loss": 1.5487, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.780278205871582, + "learning_rate": 4.4821283509342e-06, + "loss": 1.6269, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.5427017211914062, + "learning_rate": 4.480097481722178e-06, + "loss": 1.6254, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.7540087699890137, + "learning_rate": 4.478066612510154e-06, + "loss": 1.6112, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.616352081298828, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.5637, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.686035633087158, + "learning_rate": 4.474004874086109e-06, + "loss": 1.5974, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.982365846633911, + "learning_rate": 4.471974004874087e-06, + "loss": 1.7034, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.6106085777282715, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.5482, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.7798519134521484, + "learning_rate": 4.467912266450041e-06, + "loss": 1.6475, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.820489764213562, + "eval_runtime": 37.0672, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.8634185791015625, + "learning_rate": 4.465881397238018e-06, + "loss": 1.6258, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.633626699447632, + "learning_rate": 4.463850528025995e-06, + "loss": 1.5806, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.719074249267578, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.6404, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.7545487880706787, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6404, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.6559669971466064, + "learning_rate": 4.457757920389927e-06, + "loss": 1.6104, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.9004898071289062, + "learning_rate": 4.455727051177904e-06, + "loss": 1.5513, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.6640021800994873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.5859, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.842026472091675, + "learning_rate": 4.451665312753859e-06, + "loss": 1.6241, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.784647226333618, + "learning_rate": 4.449634443541836e-06, + "loss": 1.627, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.7067618370056152, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.6276, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.8115602731704712, + "eval_runtime": 37.078, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.6582882404327393, + "learning_rate": 4.445572705117791e-06, + "loss": 1.5746, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7881102561950684, + "learning_rate": 4.443541835905768e-06, + "loss": 1.5749, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.9323439598083496, + "learning_rate": 4.441510966693745e-06, + "loss": 1.6421, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.63948655128479, + "learning_rate": 4.439480097481723e-06, + "loss": 1.5853, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.7746710777282715, + "learning_rate": 4.4374492282697e-06, + "loss": 1.6455, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.093895435333252, + "learning_rate": 4.435418359057677e-06, + "loss": 1.5311, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.526385545730591, + "learning_rate": 4.433387489845654e-06, + "loss": 1.6082, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.6604297161102295, + "learning_rate": 4.431356620633632e-06, + "loss": 1.61, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.6636247634887695, + "learning_rate": 4.429325751421609e-06, + "loss": 1.5815, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.5567514896392822, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6086, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.8170957565307617, + "eval_runtime": 37.4095, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.695563793182373, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.5836, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6009771823883057, + "learning_rate": 4.423233143785541e-06, + "loss": 1.5264, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.0490736961364746, + "learning_rate": 4.421202274573518e-06, + "loss": 1.5982, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.8609886169433594, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6265, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4415369033813477, + "learning_rate": 4.417140536149473e-06, + "loss": 1.4987, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.771657943725586, + "learning_rate": 4.415109666937449e-06, + "loss": 1.5494, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.739617347717285, + "learning_rate": 4.413078797725427e-06, + "loss": 1.6274, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.805098295211792, + "learning_rate": 4.411047928513404e-06, + "loss": 1.5686, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.7248735427856445, + "learning_rate": 4.409017059301382e-06, + "loss": 1.6027, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6653318405151367, + "learning_rate": 4.406986190089358e-06, + "loss": 1.5909, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.8076283931732178, + "eval_runtime": 37.5073, + "eval_samples_per_second": 26.661, + "eval_steps_per_second": 1.12, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.626783847808838, + "learning_rate": 4.404955320877336e-06, + "loss": 1.6205, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.6929268836975098, + "learning_rate": 4.402924451665313e-06, + "loss": 1.6146, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.6806800365448, + "learning_rate": 4.400893582453291e-06, + "loss": 1.5702, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.5407259464263916, + "learning_rate": 4.398862713241267e-06, + "loss": 1.5525, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.639122247695923, + "learning_rate": 4.396831844029245e-06, + "loss": 1.5826, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.669396162033081, + "learning_rate": 4.394800974817222e-06, + "loss": 1.5982, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.6377906799316406, + "learning_rate": 4.392770105605199e-06, + "loss": 1.5797, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6769638061523438, + "learning_rate": 4.390739236393176e-06, + "loss": 1.607, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.6347222328186035, + "learning_rate": 4.388708367181154e-06, + "loss": 1.6196, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.670680284500122, + "learning_rate": 4.386677497969131e-06, + "loss": 1.619, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.8124384880065918, + "eval_runtime": 37.4751, + "eval_samples_per_second": 26.684, + "eval_steps_per_second": 1.121, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.5886521339416504, + "learning_rate": 4.384646628757108e-06, + "loss": 1.6258, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.644566297531128, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6447, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.6432394981384277, + "learning_rate": 4.380584890333063e-06, + "loss": 1.5888, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.5602784156799316, + "learning_rate": 4.37855402112104e-06, + "loss": 1.5544, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.4874439239501953, + "learning_rate": 4.376523151909017e-06, + "loss": 1.5075, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.694748640060425, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.6498, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.582211494445801, + "learning_rate": 4.372461413484972e-06, + "loss": 1.5877, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.675713539123535, + "learning_rate": 4.370430544272949e-06, + "loss": 1.6045, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.7157161235809326, + "learning_rate": 4.368399675060926e-06, + "loss": 1.623, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.795076370239258, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6141, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.8062744140625, + "eval_runtime": 37.2435, + "eval_samples_per_second": 26.85, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6296374797821045, + "learning_rate": 4.364337936636881e-06, + "loss": 1.6087, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.362307067424858e-06, + "loss": 1.5182, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.7965915203094482, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.6025, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.807307481765747, + "learning_rate": 4.358245329000813e-06, + "loss": 1.574, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.584615468978882, + "learning_rate": 4.35621445978879e-06, + "loss": 1.5922, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.8472282886505127, + "learning_rate": 4.354183590576767e-06, + "loss": 1.5914, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.8471736907958984, + "learning_rate": 4.352152721364745e-06, + "loss": 1.5749, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.569225549697876, + "learning_rate": 4.350121852152722e-06, + "loss": 1.5591, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.688157796859741, + "learning_rate": 4.348090982940699e-06, + "loss": 1.6042, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.5865259170532227, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.6307, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.8040707111358643, + "eval_runtime": 36.9511, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.71850323677063, + "learning_rate": 4.344029244516654e-06, + "loss": 1.6043, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.7158093452453613, + "learning_rate": 4.341998375304631e-06, + "loss": 1.5907, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6879327297210693, + "learning_rate": 4.339967506092607e-06, + "loss": 1.6365, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.827277660369873, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.5928, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6709842681884766, + "learning_rate": 4.335905767668562e-06, + "loss": 1.5812, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.7071495056152344, + "learning_rate": 4.33387489845654e-06, + "loss": 1.5571, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6008732318878174, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.6048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.6238059997558594, + "learning_rate": 4.329813160032495e-06, + "loss": 1.5962, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.567065954208374, + "learning_rate": 4.327782290820471e-06, + "loss": 1.5713, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.573068857192993, + "learning_rate": 4.325751421608449e-06, + "loss": 1.536, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.8103471994400024, + "eval_runtime": 37.0956, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6245224475860596, + "learning_rate": 4.323720552396426e-06, + "loss": 1.5986, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.6243810653686523, + "learning_rate": 4.321689683184404e-06, + "loss": 1.589, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.5303313732147217, + "learning_rate": 4.31965881397238e-06, + "loss": 1.5793, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.5617806911468506, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.5704, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.7429656982421875, + "learning_rate": 4.315597075548335e-06, + "loss": 1.6174, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.6753251552581787, + "learning_rate": 4.313566206336312e-06, + "loss": 1.5879, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.5944488048553467, + "learning_rate": 4.311535337124289e-06, + "loss": 1.6082, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.9002904891967773, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.5521, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.5384247303009033, + "learning_rate": 4.307473598700244e-06, + "loss": 1.5751, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6015477180480957, + "learning_rate": 4.305442729488221e-06, + "loss": 1.5966, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8091154098510742, + "eval_runtime": 37.1075, + "eval_samples_per_second": 26.949, + "eval_steps_per_second": 1.132, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.6002886295318604, + "learning_rate": 4.303411860276198e-06, + "loss": 1.5403, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7755627632141113, + "learning_rate": 4.301380991064176e-06, + "loss": 1.6192, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.8351540565490723, + "learning_rate": 4.299350121852153e-06, + "loss": 1.6217, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.684007406234741, + "learning_rate": 4.29731925264013e-06, + "loss": 1.6242, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.8019614219665527, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6173, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.9122867584228516, + "learning_rate": 4.293257514216085e-06, + "loss": 1.5701, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.545732021331787, + "learning_rate": 4.291226645004062e-06, + "loss": 1.5494, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.9499566555023193, + "learning_rate": 4.289195775792039e-06, + "loss": 1.5848, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.8071892261505127, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6114, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.763864517211914, + "learning_rate": 4.285134037367994e-06, + "loss": 1.5768, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8057912588119507, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 3.001190423965454, + "learning_rate": 4.283103168155971e-06, + "loss": 1.5601, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.7640297412872314, + "learning_rate": 4.281072298943948e-06, + "loss": 1.5619, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.7372794151306152, + "learning_rate": 4.279041429731926e-06, + "loss": 1.562, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 3.0390751361846924, + "learning_rate": 4.277010560519903e-06, + "loss": 1.5519, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.758798599243164, + "learning_rate": 4.27497969130788e-06, + "loss": 1.527, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.96443772315979, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.5899, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.7941088676452637, + "learning_rate": 4.270917952883835e-06, + "loss": 1.617, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 3.0385100841522217, + "learning_rate": 4.268887083671812e-06, + "loss": 1.5996, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.6946637630462646, + "learning_rate": 4.266856214459789e-06, + "loss": 1.5641, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.776723861694336, + "learning_rate": 4.264825345247766e-06, + "loss": 1.6451, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.7911298274993896, + "eval_runtime": 37.0623, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.851224184036255, + "learning_rate": 4.262794476035744e-06, + "loss": 1.5302, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9045867919921875, + "learning_rate": 4.26076360682372e-06, + "loss": 1.5808, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.6568782329559326, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.5233, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.94191837310791, + "learning_rate": 4.256701868399675e-06, + "loss": 1.5912, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.726778984069824, + "learning_rate": 4.254670999187653e-06, + "loss": 1.5823, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.8704583644866943, + "learning_rate": 4.252640129975629e-06, + "loss": 1.5931, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.5721547603607178, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6128, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6883509159088135, + "learning_rate": 4.248578391551584e-06, + "loss": 1.5479, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.7186272144317627, + "learning_rate": 4.246547522339562e-06, + "loss": 1.5203, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.8326518535614014, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.5547, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.7948570251464844, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.7870523929595947, + "learning_rate": 4.242485783915517e-06, + "loss": 1.5891, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.866610288619995, + "learning_rate": 4.240454914703493e-06, + "loss": 1.5232, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.94161057472229, + "learning_rate": 4.23842404549147e-06, + "loss": 1.52, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.7331979274749756, + "learning_rate": 4.236393176279448e-06, + "loss": 1.5665, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.5375845432281494, + "learning_rate": 4.234362307067425e-06, + "loss": 1.5274, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.691927433013916, + "learning_rate": 4.232331437855402e-06, + "loss": 1.604, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7578394412994385, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.6, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.594538688659668, + "learning_rate": 4.228269699431357e-06, + "loss": 1.5928, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.5408968925476074, + "learning_rate": 4.226238830219334e-06, + "loss": 1.5908, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.689786672592163, + "learning_rate": 4.224207961007311e-06, + "loss": 1.5757, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.792494297027588, + "eval_runtime": 37.0822, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.735539197921753, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.5811, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.58634614944458, + "learning_rate": 4.220146222583266e-06, + "loss": 1.5808, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.7899138927459717, + "learning_rate": 4.218115353371243e-06, + "loss": 1.5801, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.6507744789123535, + "learning_rate": 4.21608448415922e-06, + "loss": 1.5585, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.6112217903137207, + "learning_rate": 4.214053614947198e-06, + "loss": 1.5329, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.729405164718628, + "learning_rate": 4.212022745735175e-06, + "loss": 1.6177, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.6431992053985596, + "learning_rate": 4.209991876523152e-06, + "loss": 1.5234, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.726243734359741, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.5926, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.6574180126190186, + "learning_rate": 4.205930138099107e-06, + "loss": 1.5424, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.6570956707000732, + "learning_rate": 4.203899268887084e-06, + "loss": 1.6119, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.781999945640564, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.528052806854248, + "learning_rate": 4.201868399675061e-06, + "loss": 1.598, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.6075727939605713, + "learning_rate": 4.199837530463039e-06, + "loss": 1.5886, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.588737964630127, + "learning_rate": 4.197806661251016e-06, + "loss": 1.507, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.773618221282959, + "learning_rate": 4.195775792038993e-06, + "loss": 1.5794, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.624725580215454, + "learning_rate": 4.19374492282697e-06, + "loss": 1.5495, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.662397861480713, + "learning_rate": 4.191714053614948e-06, + "loss": 1.5203, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.6617255210876465, + "learning_rate": 4.189683184402924e-06, + "loss": 1.5885, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.5784451961517334, + "learning_rate": 4.187652315190902e-06, + "loss": 1.5689, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.726011037826538, + "learning_rate": 4.185621445978879e-06, + "loss": 1.5913, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.6117324829101562, + "learning_rate": 4.183590576766857e-06, + "loss": 1.5453, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.7893352508544922, + "eval_runtime": 37.4096, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7025442123413086, + "learning_rate": 4.181559707554833e-06, + "loss": 1.5868, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.656102180480957, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6141, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.741962194442749, + "learning_rate": 4.177497969130788e-06, + "loss": 1.5516, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5887229442596436, + "learning_rate": 4.175467099918766e-06, + "loss": 1.5525, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.542985439300537, + "learning_rate": 4.173436230706742e-06, + "loss": 1.592, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.7797129154205322, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.5349, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.632887363433838, + "learning_rate": 4.169374492282697e-06, + "loss": 1.5548, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.691690444946289, + "learning_rate": 4.167343623070675e-06, + "loss": 1.5352, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.8922817707061768, + "learning_rate": 4.165312753858652e-06, + "loss": 1.5717, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.5517446994781494, + "learning_rate": 4.163281884646629e-06, + "loss": 1.5658, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.780922770500183, + "eval_runtime": 37.3755, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.718374490737915, + "learning_rate": 4.161251015434607e-06, + "loss": 1.5578, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.6249442100524902, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6018, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.7051141262054443, + "learning_rate": 4.157189277010561e-06, + "loss": 1.5182, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.677109956741333, + "learning_rate": 4.155158407798538e-06, + "loss": 1.5423, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.591059684753418, + "learning_rate": 4.153127538586516e-06, + "loss": 1.5694, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.699953556060791, + "learning_rate": 4.151096669374492e-06, + "loss": 1.5758, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.8837342262268066, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.5509, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.6566851139068604, + "learning_rate": 4.147034930950447e-06, + "loss": 1.5577, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.901057004928589, + "learning_rate": 4.145004061738425e-06, + "loss": 1.5797, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.796971321105957, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.5608, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.7797813415527344, + "eval_runtime": 36.8425, + "eval_samples_per_second": 27.143, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.605431318283081, + "learning_rate": 4.140942323314379e-06, + "loss": 1.491, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8084020614624023, + "learning_rate": 4.138911454102356e-06, + "loss": 1.5074, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.625511407852173, + "learning_rate": 4.136880584890333e-06, + "loss": 1.5404, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.6944286823272705, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.589, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.81792950630188, + "learning_rate": 4.132818846466288e-06, + "loss": 1.5646, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5813655853271484, + "learning_rate": 4.130787977254265e-06, + "loss": 1.5697, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.6733286380767822, + "learning_rate": 4.128757108042242e-06, + "loss": 1.5786, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.6972577571868896, + "learning_rate": 4.12672623883022e-06, + "loss": 1.5857, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7381093502044678, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6365, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.6162049770355225, + "learning_rate": 4.122664500406174e-06, + "loss": 1.5598, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.7869179248809814, + "eval_runtime": 36.974, + "eval_samples_per_second": 27.046, + "eval_steps_per_second": 1.136, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.7157270908355713, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.5533, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.560105085372925, + "learning_rate": 4.118602761982129e-06, + "loss": 1.5821, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.5261518955230713, + "learning_rate": 4.116571892770106e-06, + "loss": 1.5195, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7377054691314697, + "learning_rate": 4.114541023558083e-06, + "loss": 1.5066, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.632946252822876, + "learning_rate": 4.112510154346061e-06, + "loss": 1.5948, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.809671640396118, + "learning_rate": 4.110479285134038e-06, + "loss": 1.5435, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.843156099319458, + "learning_rate": 4.108448415922015e-06, + "loss": 1.5667, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.743614673614502, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5958, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.9596827030181885, + "learning_rate": 4.10438667749797e-06, + "loss": 1.5807, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.7321834564208984, + "learning_rate": 4.102355808285947e-06, + "loss": 1.4744, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.784859538078308, + "eval_runtime": 37.0146, + "eval_samples_per_second": 27.016, + "eval_steps_per_second": 1.135, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.010359764099121, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6146, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.744251251220703, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.5401, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.592690944671631, + "learning_rate": 4.096263200649879e-06, + "loss": 1.5508, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.786736011505127, + "learning_rate": 4.094232331437856e-06, + "loss": 1.5717, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.8773610591888428, + "learning_rate": 4.092201462225833e-06, + "loss": 1.5201, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.818610906600952, + "learning_rate": 4.090170593013811e-06, + "loss": 1.5282, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.9548542499542236, + "learning_rate": 4.088139723801787e-06, + "loss": 1.553, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.678842306137085, + "learning_rate": 4.086108854589765e-06, + "loss": 1.531, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.7757408618927, + "learning_rate": 4.084077985377742e-06, + "loss": 1.5371, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.6954638957977295, + "learning_rate": 4.08204711616572e-06, + "loss": 1.5462, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.781003475189209, + "eval_runtime": 37.0687, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6897518634796143, + "learning_rate": 4.080016246953696e-06, + "loss": 1.5979, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.581172466278076, + "learning_rate": 4.077985377741674e-06, + "loss": 1.5007, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.828234910964966, + "learning_rate": 4.075954508529651e-06, + "loss": 1.5857, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.6171951293945312, + "learning_rate": 4.073923639317629e-06, + "loss": 1.5937, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.6476316452026367, + "learning_rate": 4.071892770105605e-06, + "loss": 1.5864, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.7525634765625, + "learning_rate": 4.069861900893583e-06, + "loss": 1.5095, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.6881322860717773, + "learning_rate": 4.06783103168156e-06, + "loss": 1.4936, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.6104350090026855, + "learning_rate": 4.065800162469538e-06, + "loss": 1.5229, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.866706132888794, + "learning_rate": 4.063769293257514e-06, + "loss": 1.539, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.803999900817871, + "learning_rate": 4.061738424045492e-06, + "loss": 1.5792, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.7793465852737427, + "eval_runtime": 37.0749, + "eval_samples_per_second": 26.972, + "eval_steps_per_second": 1.133, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.670285940170288, + "learning_rate": 4.059707554833469e-06, + "loss": 1.5523, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.871267557144165, + "learning_rate": 4.057676685621446e-06, + "loss": 1.5623, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.6631205081939697, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6094, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.7580034732818604, + "learning_rate": 4.053614947197401e-06, + "loss": 1.5725, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.6412739753723145, + "learning_rate": 4.051584077985378e-06, + "loss": 1.4959, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.699176073074341, + "learning_rate": 4.049553208773355e-06, + "loss": 1.5586, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.622969150543213, + "learning_rate": 4.047522339561333e-06, + "loss": 1.5713, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.892338991165161, + "learning_rate": 4.04549147034931e-06, + "loss": 1.5471, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.77056622505188, + "learning_rate": 4.043460601137287e-06, + "loss": 1.5236, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.7122550010681152, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6562, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.7708193063735962, + "eval_runtime": 37.0743, + "eval_samples_per_second": 26.973, + "eval_steps_per_second": 1.133, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.702824592590332, + "learning_rate": 4.039398862713242e-06, + "loss": 1.534, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.6822307109832764, + "learning_rate": 4.037367993501219e-06, + "loss": 1.5368, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.7274515628814697, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6218, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.8114798069000244, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.532, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.8936192989349365, + "learning_rate": 4.031275385865151e-06, + "loss": 1.5124, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.7851672172546387, + "learning_rate": 4.029244516653128e-06, + "loss": 1.5596, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.815023183822632, + "learning_rate": 4.027213647441105e-06, + "loss": 1.5706, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5958921909332275, + "learning_rate": 4.025182778229083e-06, + "loss": 1.5161, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.779360771179199, + "learning_rate": 4.02315190901706e-06, + "loss": 1.5057, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.7411787509918213, + "learning_rate": 4.021121039805037e-06, + "loss": 1.5396, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.7778708934783936, + "eval_runtime": 37.0791, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.5709781646728516, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.5474, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8583388328552246, + "learning_rate": 4.017059301380992e-06, + "loss": 1.6149, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.6877572536468506, + "learning_rate": 4.015028432168969e-06, + "loss": 1.5292, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.680612087249756, + "learning_rate": 4.012997562956945e-06, + "loss": 1.596, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.653559446334839, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.4901, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.542787790298462, + "learning_rate": 4.0089358245329e-06, + "loss": 1.5295, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5476601123809814, + "learning_rate": 4.006904955320878e-06, + "loss": 1.5393, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.6367862224578857, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.5321, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.8112173080444336, + "learning_rate": 4.002843216896833e-06, + "loss": 1.5963, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.7136871814727783, + "learning_rate": 4.000812347684809e-06, + "loss": 1.5839, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.7763899564743042, + "eval_runtime": 37.2537, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.9095280170440674, + "learning_rate": 3.998781478472787e-06, + "loss": 1.5343, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.6196115016937256, + "learning_rate": 3.996750609260764e-06, + "loss": 1.5411, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6413371562957764, + "learning_rate": 3.994719740048742e-06, + "loss": 1.4862, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 3.0759036540985107, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6471, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.599147081375122, + "learning_rate": 3.990658001624696e-06, + "loss": 1.5556, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.726978302001953, + "learning_rate": 3.988627132412673e-06, + "loss": 1.5351, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.6760807037353516, + "learning_rate": 3.98659626320065e-06, + "loss": 1.5681, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.680483818054199, + "learning_rate": 3.984565393988627e-06, + "loss": 1.5225, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6649010181427, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.5325, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.638085126876831, + "learning_rate": 3.980503655564582e-06, + "loss": 1.564, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.7698904275894165, + "eval_runtime": 37.4492, + "eval_samples_per_second": 26.703, + "eval_steps_per_second": 1.122, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.647825241088867, + "learning_rate": 3.978472786352559e-06, + "loss": 1.5782, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.748222589492798, + "learning_rate": 3.976441917140536e-06, + "loss": 1.556, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.6319026947021484, + "learning_rate": 3.974411047928514e-06, + "loss": 1.5378, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.6759872436523438, + "learning_rate": 3.972380178716491e-06, + "loss": 1.6018, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.6760940551757812, + "learning_rate": 3.970349309504468e-06, + "loss": 1.5742, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.5565083026885986, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.565, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.5986931324005127, + "learning_rate": 3.966287571080423e-06, + "loss": 1.5124, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.713092803955078, + "learning_rate": 3.9642567018684e-06, + "loss": 1.523, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.6203417778015137, + "learning_rate": 3.962225832656377e-06, + "loss": 1.5601, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.604440927505493, + "learning_rate": 3.960194963444355e-06, + "loss": 1.5485, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.7633332014083862, + "eval_runtime": 37.49, + "eval_samples_per_second": 26.674, + "eval_steps_per_second": 1.12, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.5513081550598145, + "learning_rate": 3.958164094232332e-06, + "loss": 1.479, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.7990963459014893, + "learning_rate": 3.956133225020309e-06, + "loss": 1.5355, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.651113271713257, + "learning_rate": 3.954102355808286e-06, + "loss": 1.5406, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.7236671447753906, + "learning_rate": 3.952071486596264e-06, + "loss": 1.5314, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.697598695755005, + "learning_rate": 3.950040617384241e-06, + "loss": 1.5398, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.8271944522857666, + "learning_rate": 3.948009748172218e-06, + "loss": 1.54, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.733125925064087, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.5236, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.7555017471313477, + "learning_rate": 3.943948009748173e-06, + "loss": 1.5808, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.6592960357666016, + "learning_rate": 3.94191714053615e-06, + "loss": 1.5398, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.6870126724243164, + "learning_rate": 3.939886271324127e-06, + "loss": 1.5517, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.760675311088562, + "eval_runtime": 36.9372, + "eval_samples_per_second": 27.073, + "eval_steps_per_second": 1.137, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.7469022274017334, + "learning_rate": 3.937855402112104e-06, + "loss": 1.605, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.5255792140960693, + "learning_rate": 3.935824532900082e-06, + "loss": 1.5061, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.589054822921753, + "learning_rate": 3.933793663688058e-06, + "loss": 1.4971, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.7874889373779297, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.5643, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.819451093673706, + "learning_rate": 3.929731925264013e-06, + "loss": 1.5797, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.6516339778900146, + "learning_rate": 3.927701056051991e-06, + "loss": 1.5048, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.979091167449951, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.5174, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.743058204650879, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.5432, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.6864030361175537, + "learning_rate": 3.921608448415922e-06, + "loss": 1.5147, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.7751128673553467, + "learning_rate": 3.9195775792039e-06, + "loss": 1.4857, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.7609487771987915, + "eval_runtime": 37.09, + "eval_samples_per_second": 26.961, + "eval_steps_per_second": 1.132, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.626188278198242, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.5428, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.554067850112915, + "learning_rate": 3.915515840779855e-06, + "loss": 1.5153, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.8580875396728516, + "learning_rate": 3.913484971567831e-06, + "loss": 1.5562, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.7405097484588623, + "learning_rate": 3.911454102355808e-06, + "loss": 1.565, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.5650815963745117, + "learning_rate": 3.909423233143786e-06, + "loss": 1.4756, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.9405910968780518, + "learning_rate": 3.907392363931763e-06, + "loss": 1.5611, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.687448024749756, + "learning_rate": 3.90536149471974e-06, + "loss": 1.5405, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.7373857498168945, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.6055, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.875298261642456, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6215, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.713200330734253, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6493, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.7608598470687866, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.541935682296753, + "learning_rate": 3.897238017871649e-06, + "loss": 1.5257, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.582045316696167, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.5253, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.6884682178497314, + "learning_rate": 3.893176279447604e-06, + "loss": 1.5329, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.6456501483917236, + "learning_rate": 3.891145410235581e-06, + "loss": 1.5167, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.60788893699646, + "learning_rate": 3.889114541023558e-06, + "loss": 1.5945, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.708646774291992, + "learning_rate": 3.887083671811536e-06, + "loss": 1.5704, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.666922092437744, + "learning_rate": 3.885052802599513e-06, + "loss": 1.52, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.6623599529266357, + "learning_rate": 3.88302193338749e-06, + "loss": 1.4791, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.9031424522399902, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.5954, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.763303279876709, + "learning_rate": 3.878960194963445e-06, + "loss": 1.5618, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.7647819519042969, + "eval_runtime": 37.1392, + "eval_samples_per_second": 26.926, + "eval_steps_per_second": 1.131, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 3.0733320713043213, + "learning_rate": 3.876929325751422e-06, + "loss": 1.5191, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.7809090614318848, + "learning_rate": 3.874898456539399e-06, + "loss": 1.5804, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.6357381343841553, + "learning_rate": 3.872867587327377e-06, + "loss": 1.4843, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.680595874786377, + "learning_rate": 3.870836718115354e-06, + "loss": 1.5339, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.5908069610595703, + "learning_rate": 3.868805848903331e-06, + "loss": 1.5049, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.680586338043213, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.5188, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.763528823852539, + "learning_rate": 3.864744110479286e-06, + "loss": 1.5395, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.8522050380706787, + "learning_rate": 3.862713241267262e-06, + "loss": 1.562, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.806962490081787, + "learning_rate": 3.86068237205524e-06, + "loss": 1.5419, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.9716053009033203, + "learning_rate": 3.858651502843217e-06, + "loss": 1.5691, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.7608845233917236, + "eval_runtime": 37.1433, + "eval_samples_per_second": 26.923, + "eval_steps_per_second": 1.131, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.613098382949829, + "learning_rate": 3.856620633631195e-06, + "loss": 1.5297, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.5690557956695557, + "learning_rate": 3.854589764419171e-06, + "loss": 1.4964, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.8001089096069336, + "learning_rate": 3.852558895207149e-06, + "loss": 1.5253, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.6715309619903564, + "learning_rate": 3.850528025995126e-06, + "loss": 1.601, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.7044479846954346, + "learning_rate": 3.848497156783104e-06, + "loss": 1.5526, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.7368102073669434, + "learning_rate": 3.84646628757108e-06, + "loss": 1.5088, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.9358696937561035, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.5202, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.858172655105591, + "learning_rate": 3.842404549147035e-06, + "loss": 1.5378, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.6723833084106445, + "learning_rate": 3.840373679935013e-06, + "loss": 1.5134, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.748187780380249, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.4797, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.75319242477417, + "eval_runtime": 37.1302, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.4821436405181885, + "learning_rate": 3.836311941510967e-06, + "loss": 1.5382, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.5585227012634277, + "learning_rate": 3.834281072298944e-06, + "loss": 1.5373, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.927334785461426, + "learning_rate": 3.832250203086921e-06, + "loss": 1.5447, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.6096436977386475, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.5062, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.677304744720459, + "learning_rate": 3.828188464662876e-06, + "loss": 1.5671, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.9606077671051025, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6138, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.6971657276153564, + "learning_rate": 3.82412672623883e-06, + "loss": 1.5544, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.6380839347839355, + "learning_rate": 3.822095857026808e-06, + "loss": 1.4964, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.7537858486175537, + "learning_rate": 3.820064987814785e-06, + "loss": 1.5429, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.6189136505126953, + "learning_rate": 3.818034118602762e-06, + "loss": 1.4956, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.7546342611312866, + "eval_runtime": 37.0828, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5744292736053467, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.5256, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.5503833293914795, + "learning_rate": 3.813972380178717e-06, + "loss": 1.5211, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.979384422302246, + "learning_rate": 3.811941510966694e-06, + "loss": 1.55, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.867161512374878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5655, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.8006832599639893, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.4902, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.594712734222412, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.4451, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.6632349491119385, + "learning_rate": 3.803818034118603e-06, + "loss": 1.531, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.859894275665283, + "learning_rate": 3.80178716490658e-06, + "loss": 1.5018, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.6405653953552246, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.5765, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.7285192012786865, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.5537, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.7618242502212524, + "eval_runtime": 37.0898, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.6882379055023193, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.5143, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.706550121307373, + "learning_rate": 3.793663688058489e-06, + "loss": 1.4857, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.790907382965088, + "learning_rate": 3.791632818846467e-06, + "loss": 1.5982, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.631096363067627, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.5656, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.6981775760650635, + "learning_rate": 3.787571080422421e-06, + "loss": 1.567, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.612086296081543, + "learning_rate": 3.785540211210398e-06, + "loss": 1.5723, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.718630075454712, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.5565, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.6265041828155518, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.5326, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.578124523162842, + "learning_rate": 3.77944760357433e-06, + "loss": 1.533, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.832695484161377, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6133, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.7516050338745117, + "eval_runtime": 37.3173, + "eval_samples_per_second": 26.797, + "eval_steps_per_second": 1.125, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.851625442504883, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.5836, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.773354995938262e-06, + "loss": 1.5069, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.5738255977630615, + "learning_rate": 3.771324126726239e-06, + "loss": 1.5531, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.583575963973999, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.5589, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.7729389667510986, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.5901, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.6322245597839355, + "learning_rate": 3.765231519090171e-06, + "loss": 1.5423, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.6382699012756348, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.4574, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.7393112182617188, + "learning_rate": 3.761169780666125e-06, + "loss": 1.5351, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.684398651123047, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.4873, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.89401912689209, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.5358, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.7523635625839233, + "eval_runtime": 37.3763, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.629870891571045, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.5566, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.6936254501342773, + "learning_rate": 3.753046303818034e-06, + "loss": 1.5016, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.5889103412628174, + "learning_rate": 3.751015434606012e-06, + "loss": 1.5775, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.610703468322754, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.5508, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6318750381469727, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.5892, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.6650729179382324, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.5106, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.577174186706543, + "learning_rate": 3.742891957757921e-06, + "loss": 1.54, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.7274489402770996, + "learning_rate": 3.740861088545898e-06, + "loss": 1.5803, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6849985122680664, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.5393, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.811479091644287, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.5862, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.7490049600601196, + "eval_runtime": 36.9083, + "eval_samples_per_second": 27.094, + "eval_steps_per_second": 1.138, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.640002489089966, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.5149, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.6350319385528564, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5132, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.689814567565918, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.473, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.7181100845336914, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.4353, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.760629415512085, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.4859, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.9322562217712402, + "learning_rate": 3.724614134849716e-06, + "loss": 1.4733, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.280315399169922, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.5769, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 3.1343884468078613, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.4766, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.8537609577178955, + "learning_rate": 3.718521527213648e-06, + "loss": 1.5586, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.895299196243286, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5232, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.7473852634429932, + "eval_runtime": 37.006, + "eval_samples_per_second": 27.023, + "eval_steps_per_second": 1.135, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.82454776763916, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5057, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.711296558380127, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.4187, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.6281023025512695, + "learning_rate": 3.710398050365557e-06, + "loss": 1.4997, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.775632381439209, + "learning_rate": 3.708367181153534e-06, + "loss": 1.5087, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.8021788597106934, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.474, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 3.09647536277771, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5015, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.810450315475464, + "learning_rate": 3.702274573517466e-06, + "loss": 1.4681, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.65166974067688, + "learning_rate": 3.700243704305443e-06, + "loss": 1.4484, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.755690574645996, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5084, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.8223068714141846, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.5015, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.7516940832138062, + "eval_runtime": 37.0227, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.8401107788085938, + "learning_rate": 3.694151096669375e-06, + "loss": 1.4597, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 3.010011672973633, + "learning_rate": 3.692120227457352e-06, + "loss": 1.5625, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8575570583343506, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.4828, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.8527276515960693, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5074, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.81260085105896, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.4132, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.988429069519043, + "learning_rate": 3.683996750609261e-06, + "loss": 1.4803, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.964179515838623, + "learning_rate": 3.681965881397238e-06, + "loss": 1.4871, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.7151682376861572, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.5061, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.764187812805176, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.5188, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.810554265975952, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.4779, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.7422841787338257, + "eval_runtime": 37.0641, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.7640810012817383, + "learning_rate": 3.673842404549147e-06, + "loss": 1.4827, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.7683937549591064, + "learning_rate": 3.671811535337125e-06, + "loss": 1.4775, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.86434268951416, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.4752, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 3.0976617336273193, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.5099, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 3.107388496398926, + "learning_rate": 3.665718927701056e-06, + "loss": 1.5646, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 3.032921314239502, + "learning_rate": 3.663688058489034e-06, + "loss": 1.5048, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.9377830028533936, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5333, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.706542730331421, + "learning_rate": 3.659626320064988e-06, + "loss": 1.4903, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.819936513900757, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.4441, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.805360794067383, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.5122, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.747745156288147, + "eval_runtime": 37.0801, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.848580837249756, + "learning_rate": 3.65353371242892e-06, + "loss": 1.4984, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.772721290588379, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5026, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.760772943496704, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.4005, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.937507390975952, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.5131, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.923210859298706, + "learning_rate": 3.645410235580829e-06, + "loss": 1.4661, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8029167652130127, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.518, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.752716541290283, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.5109, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.7767865657806396, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.4831, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8483617305755615, + "learning_rate": 3.637286758732738e-06, + "loss": 1.4872, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.8714234828948975, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.5109, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.7455583810806274, + "eval_runtime": 37.0892, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7373011112213135, + "learning_rate": 3.633225020308692e-06, + "loss": 1.55, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.942373752593994, + "learning_rate": 3.63119415109667e-06, + "loss": 1.5079, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.8790931701660156, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.4669, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.9564976692199707, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5555, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.811699390411377, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.4457, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.926544427871704, + "learning_rate": 3.623070674248579e-06, + "loss": 1.4368, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.8694913387298584, + "learning_rate": 3.621039805036556e-06, + "loss": 1.4875, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 3.0294361114501953, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.4815, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8643198013305664, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.5031, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 3.0244719982147217, + "learning_rate": 3.614947197400488e-06, + "loss": 1.4914, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.7504756450653076, + "eval_runtime": 37.0238, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.8641672134399414, + "learning_rate": 3.612916328188465e-06, + "loss": 1.5122, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.8713624477386475, + "learning_rate": 3.610885458976442e-06, + "loss": 1.43, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.990912914276123, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.5124, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.963945150375366, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.533, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.802168846130371, + "learning_rate": 3.604792851340374e-06, + "loss": 1.4531, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 3.0556633472442627, + "learning_rate": 3.602761982128351e-06, + "loss": 1.4618, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.95744252204895, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5356, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.813145399093628, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.5134, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.8382728099823, + "learning_rate": 3.596669374492283e-06, + "loss": 1.4413, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.821100950241089, + "learning_rate": 3.59463850528026e-06, + "loss": 1.4748, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.7488586902618408, + "eval_runtime": 37.0361, + "eval_samples_per_second": 27.001, + "eval_steps_per_second": 1.134, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.930708408355713, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.5177, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.9312615394592285, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.4493, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 3.0051841735839844, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.5208, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.793515205383301, + "learning_rate": 3.586515028432169e-06, + "loss": 1.4813, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.7595956325531006, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5379, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.992871046066284, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.5153, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.7270801067352295, + "learning_rate": 3.580422420796101e-06, + "loss": 1.4392, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.797985076904297, + "learning_rate": 3.578391551584078e-06, + "loss": 1.4728, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9595260620117188, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.5158, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.8131442070007324, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.4214, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.7503563165664673, + "eval_runtime": 37.3831, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.124, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8242709636688232, + "learning_rate": 3.57229894394801e-06, + "loss": 1.5027, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.8973116874694824, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.5294, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 3.1207027435302734, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.4902, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.803910732269287, + "learning_rate": 3.566206336311942e-06, + "loss": 1.466, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.813541889190674, + "learning_rate": 3.564175467099919e-06, + "loss": 1.4957, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.9068171977996826, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.4435, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.8491995334625244, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.4432, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.8785598278045654, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.4791, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7701117992401123, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.4607, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.4821, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.7518969774246216, + "eval_runtime": 37.4808, + "eval_samples_per_second": 26.68, + "eval_steps_per_second": 1.121, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.5228, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.8513107299804688, + "learning_rate": 3.551990251827783e-06, + "loss": 1.4812, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.937448024749756, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.4538, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.8307411670684814, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.4807, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.7986767292022705, + "learning_rate": 3.545897644191714e-06, + "loss": 1.4706, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 3.1135106086730957, + "learning_rate": 3.543866774979692e-06, + "loss": 1.4574, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.7756283283233643, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5114, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.8114547729492188, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.4458, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 3.0544545650482178, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.4934, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8990671634674072, + "learning_rate": 3.535743298131601e-06, + "loss": 1.4808, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.7468736171722412, + "eval_runtime": 36.9135, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.976139545440674, + "learning_rate": 3.533712428919578e-06, + "loss": 1.5123, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.909146785736084, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.5204, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7225229740142822, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.4746, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.6145517826080322, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.5049, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.8660519123077393, + "learning_rate": 3.525588952071487e-06, + "loss": 1.5173, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.979801654815674, + "learning_rate": 3.523558082859464e-06, + "loss": 1.5146, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.7500879764556885, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.514, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.6923978328704834, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.455, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.6728951930999756, + "learning_rate": 3.517465475223396e-06, + "loss": 1.4712, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.769507884979248, + "learning_rate": 3.515434606011373e-06, + "loss": 1.444, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.7491599321365356, + "eval_runtime": 37.066, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.9768483638763428, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.5005, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.993196725845337, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.5347, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.886883497238159, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.4959, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.9152941703796387, + "learning_rate": 3.507311129163282e-06, + "loss": 1.4641, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.966872215270996, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.4232, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.9853367805480957, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.4883, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.9353625774383545, + "learning_rate": 3.501218521527214e-06, + "loss": 1.4826, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.7860453128814697, + "learning_rate": 3.499187652315191e-06, + "loss": 1.42, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.9269113540649414, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.5087, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 3.052962303161621, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5042, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.7421623468399048, + "eval_runtime": 37.103, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.7328579425811768, + "learning_rate": 3.493095044679123e-06, + "loss": 1.4375, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.876138925552368, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5559, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.7500345706939697, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.4043, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.836596965789795, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.4808, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.797060966491699, + "learning_rate": 3.484971567831032e-06, + "loss": 1.405, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.928368091583252, + "learning_rate": 3.482940698619009e-06, + "loss": 1.4765, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.923375368118286, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.5134, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.8624660968780518, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.4881, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.9593892097473145, + "learning_rate": 3.476848090982941e-06, + "loss": 1.4869, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.8019728660583496, + "learning_rate": 3.474817221770918e-06, + "loss": 1.468, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.7386831045150757, + "eval_runtime": 37.1034, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.8965771198272705, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.5604, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.8114798069000244, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.5092, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.9554526805877686, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.4869, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.981499195098877, + "learning_rate": 3.466693744922827e-06, + "loss": 1.4454, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.8626201152801514, + "learning_rate": 3.464662875710805e-06, + "loss": 1.4701, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.9752259254455566, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.5004, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 3.1553750038146973, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.4652, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.9458937644958496, + "learning_rate": 3.458570268074736e-06, + "loss": 1.51, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.9626898765563965, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.484, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.8093032836914062, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.468, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.740411639213562, + "eval_runtime": 37.1036, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.928344488143921, + "learning_rate": 3.452477660438668e-06, + "loss": 1.4474, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 3.0413129329681396, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.4759, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.7930357456207275, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.4406, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.8075618743896484, + "learning_rate": 3.4463850528026e-06, + "loss": 1.48, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 3.019402503967285, + "learning_rate": 3.444354183590577e-06, + "loss": 1.4713, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 3.009504795074463, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5124, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.9452531337738037, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.464, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.6980783939361572, + "learning_rate": 3.438261575954509e-06, + "loss": 1.4143, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 3.039795160293579, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.4712, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.285968542098999, + "learning_rate": 3.434199837530463e-06, + "loss": 1.5227, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.7402914762496948, + "eval_runtime": 37.1273, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 3.035813093185425, + "learning_rate": 3.432168968318441e-06, + "loss": 1.4772, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 3.1172401905059814, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.4389, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 3.0828707218170166, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.5098, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.922488212585449, + "learning_rate": 3.426076360682372e-06, + "loss": 1.5239, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 3.03352689743042, + "learning_rate": 3.42404549147035e-06, + "loss": 1.5296, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.933380603790283, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.4851, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7250027656555176, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.3966, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.991846799850464, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.4863, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 3.0704684257507324, + "learning_rate": 3.415922014622259e-06, + "loss": 1.5366, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 3.0659193992614746, + "learning_rate": 3.413891145410236e-06, + "loss": 1.516, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.7458266019821167, + "eval_runtime": 37.06, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.9695920944213867, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.508, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 3.004558801651001, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.5092, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.8377230167388916, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.4909, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.860020875930786, + "learning_rate": 3.405767668562145e-06, + "loss": 1.4415, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8683009147644043, + "learning_rate": 3.403736799350122e-06, + "loss": 1.4997, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.91894268989563, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.4886, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.8158814907073975, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.4295, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.9528515338897705, + "learning_rate": 3.397644191714054e-06, + "loss": 1.5422, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.994856119155884, + "learning_rate": 3.395613322502031e-06, + "loss": 1.4909, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.8976950645446777, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.4959, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.7350144386291504, + "eval_runtime": 37.0807, + "eval_samples_per_second": 26.968, + "eval_steps_per_second": 1.133, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.73504638671875, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.4454, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.816823720932007, + "learning_rate": 3.389520714865963e-06, + "loss": 1.4864, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.706554889678955, + "learning_rate": 3.38748984565394e-06, + "loss": 1.483, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.9004647731781006, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.492, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8677759170532227, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.4786, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.83447527885437, + "learning_rate": 3.381397238017872e-06, + "loss": 1.4575, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.9942550659179688, + "learning_rate": 3.379366368805849e-06, + "loss": 1.4746, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.124676465988159, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.5406, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.9151222705841064, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.4748, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.734571933746338, + "learning_rate": 3.373273761169781e-06, + "loss": 1.4773, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.7389248609542847, + "eval_runtime": 37.1519, + "eval_samples_per_second": 26.917, + "eval_steps_per_second": 1.13, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.8919990062713623, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.4698, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.873870611190796, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.4509, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 3.0072309970855713, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5024, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.784576177597046, + "learning_rate": 3.36515028432169e-06, + "loss": 1.4405, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.8316547870635986, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.4697, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.8758575916290283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.495, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.879115343093872, + "learning_rate": 3.359057676685622e-06, + "loss": 1.5248, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.9559383392333984, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5229, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.6783695220947266, + "learning_rate": 3.354995938261576e-06, + "loss": 1.4251, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.8372304439544678, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.4649, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.7335349321365356, + "eval_runtime": 37.3498, + "eval_samples_per_second": 26.774, + "eval_steps_per_second": 1.125, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.729576826095581, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.4263, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.7717669010162354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.4467, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.8166871070861816, + "learning_rate": 3.346872461413485e-06, + "loss": 1.4937, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.845677137374878, + "learning_rate": 3.344841592201463e-06, + "loss": 1.4303, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.9527740478515625, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.4651, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.810873031616211, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.4521, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.77546763420105, + "learning_rate": 3.338748984565394e-06, + "loss": 1.4857, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.9098589420318604, + "learning_rate": 3.336718115353372e-06, + "loss": 1.4988, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.797712802886963, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.5307, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.916668176651001, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.4828, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.7372808456420898, + "eval_runtime": 36.888, + "eval_samples_per_second": 27.109, + "eval_steps_per_second": 1.139, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.8220417499542236, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.4636, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.851010322570801, + "learning_rate": 3.32859463850528e-06, + "loss": 1.4579, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.983619213104248, + "learning_rate": 3.326563769293258e-06, + "loss": 1.4801, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.750732898712158, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.4438, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.631105422973633, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.4439, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.944300889968872, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.437, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8841664791107178, + "learning_rate": 3.318440292445167e-06, + "loss": 1.4201, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.917189121246338, + "learning_rate": 3.316409423233144e-06, + "loss": 1.4489, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.112077236175537, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.4449, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.9847209453582764, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.5153, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.7323338985443115, + "eval_runtime": 36.9935, + "eval_samples_per_second": 27.032, + "eval_steps_per_second": 1.135, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.948901414871216, + "learning_rate": 3.310316815597076e-06, + "loss": 1.473, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.84930419921875, + "learning_rate": 3.308285946385053e-06, + "loss": 1.4796, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.9120237827301025, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5392, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.7088463306427, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.4206, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.7495498657226562, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.4759, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.9146509170532227, + "learning_rate": 3.300162469536962e-06, + "loss": 1.51, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.781975030899048, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.4667, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.7524054050445557, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.5036, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.7672629356384277, + "learning_rate": 3.294069861900894e-06, + "loss": 1.4827, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.875802516937256, + "learning_rate": 3.292038992688871e-06, + "loss": 1.519, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.7273142337799072, + "eval_runtime": 37.0032, + "eval_samples_per_second": 27.025, + "eval_steps_per_second": 1.135, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.8699231147766113, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.474, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.8670589923858643, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.5039, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.9946320056915283, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5226, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.767071008682251, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.4221, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8807036876678467, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.486, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.687720537185669, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.3944, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.8170154094696045, + "learning_rate": 3.277822908204712e-06, + "loss": 1.5213, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.905557632446289, + "learning_rate": 3.275792038992689e-06, + "loss": 1.4598, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.643657922744751, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.4282, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.798712968826294, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.4752, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.7295492887496948, + "eval_runtime": 37.0432, + "eval_samples_per_second": 26.996, + "eval_steps_per_second": 1.134, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.8480567932128906, + "learning_rate": 3.269699431356621e-06, + "loss": 1.4322, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.8311381340026855, + "learning_rate": 3.267668562144598e-06, + "loss": 1.4744, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.83522891998291, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.4578, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.757307767868042, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.4469, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.9117610454559326, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.5034, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.8141143321990967, + "learning_rate": 3.259545085296507e-06, + "loss": 1.473, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 3.050138473510742, + "learning_rate": 3.257514216084484e-06, + "loss": 1.528, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.660510540008545, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.4335, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.898832082748413, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.5335, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.955589532852173, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.4832, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.7293461561203003, + "eval_runtime": 37.0096, + "eval_samples_per_second": 27.02, + "eval_steps_per_second": 1.135, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.8524296283721924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.4728, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.9741404056549072, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.4805, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.738152265548706, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.4532, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.717066764831543, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.4741, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.8367488384246826, + "learning_rate": 3.241267262388302e-06, + "loss": 1.4773, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.987780809402466, + "learning_rate": 3.23923639317628e-06, + "loss": 1.4214, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.705444097518921, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.4537, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.8966188430786133, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.4552, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.9027600288391113, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.4883, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.965813159942627, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5318, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.7236050367355347, + "eval_runtime": 37.0306, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.6390798091888428, + "learning_rate": 3.229082047116166e-06, + "loss": 1.4431, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.822848081588745, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5163, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7289068698883057, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.4551, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.766545057296753, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.4648, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.8657522201538086, + "learning_rate": 3.220958570268075e-06, + "loss": 1.4825, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.7456932067871094, + "learning_rate": 3.218927701056052e-06, + "loss": 1.3909, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 3.0274853706359863, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.4461, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.990029811859131, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.4057, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 3.086756706237793, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5251, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 3.0227134227752686, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.427, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.7248852252960205, + "eval_runtime": 37.0619, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.815480947494507, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.469, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.9171998500823975, + "learning_rate": 3.206742485783916e-06, + "loss": 1.4977, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 3.087686777114868, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.4233, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.7349331378936768, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.4187, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.9189705848693848, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.4949, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 3.0483345985412598, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5041, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.7958145141601562, + "learning_rate": 3.196588139723802e-06, + "loss": 1.4009, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 3.081714153289795, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.4718, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.7897303104400635, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.4926, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.899045944213867, + "learning_rate": 3.190495532087734e-06, + "loss": 1.4524, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.7273337841033936, + "eval_runtime": 37.0682, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.979063034057617, + "learning_rate": 3.188464662875711e-06, + "loss": 1.4706, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.8633804321289062, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5273, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.8997795581817627, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.4638, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.906151294708252, + "learning_rate": 3.182372055239643e-06, + "loss": 1.4187, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 3.0212671756744385, + "learning_rate": 3.18034118602762e-06, + "loss": 1.4213, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 3.0891034603118896, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.4921, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.9302260875701904, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.4787, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.8741140365600586, + "learning_rate": 3.174248578391552e-06, + "loss": 1.4591, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 3.0567023754119873, + "learning_rate": 3.172217709179529e-06, + "loss": 1.4636, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.80245304107666, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.4523, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.7270530462265015, + "eval_runtime": 37.325, + "eval_samples_per_second": 26.792, + "eval_steps_per_second": 1.125, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.816023349761963, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.4579, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.7405545711517334, + "learning_rate": 3.166125101543461e-06, + "loss": 1.4225, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.9569272994995117, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.483, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 3.0433337688446045, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.4846, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.856781244277954, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.426, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.8880672454833984, + "learning_rate": 3.15800162469537e-06, + "loss": 1.4764, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 3.0363011360168457, + "learning_rate": 3.155970755483347e-06, + "loss": 1.4916, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 3.0171539783477783, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.4596, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 3.183234930038452, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.4936, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.802027463912964, + "learning_rate": 3.149878147847279e-06, + "loss": 1.4733, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.7253632545471191, + "eval_runtime": 37.4737, + "eval_samples_per_second": 26.685, + "eval_steps_per_second": 1.121, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.9667463302612305, + "learning_rate": 3.147847278635256e-06, + "loss": 1.508, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 3.1557469367980957, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.5459, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.7861125469207764, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.4614, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.7431211471557617, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.4297, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 3.140911340713501, + "learning_rate": 3.139723801787165e-06, + "loss": 1.4497, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.970118522644043, + "learning_rate": 3.137692932575143e-06, + "loss": 1.4409, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.94400954246521, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.4889, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.925431966781616, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.4205, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.997279167175293, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.4714, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.855996608734131, + "learning_rate": 3.129569455727051e-06, + "loss": 1.4925, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.7247990369796753, + "eval_runtime": 36.9312, + "eval_samples_per_second": 27.077, + "eval_steps_per_second": 1.137, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 3.011199951171875, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.5132, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.8528099060058594, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.4791, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.8827919960021973, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.495, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.8150227069854736, + "learning_rate": 3.12144597887896e-06, + "loss": 1.4563, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.9341959953308105, + "learning_rate": 3.119415109666938e-06, + "loss": 1.4829, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.9161651134490967, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.492, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.906963348388672, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.445, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.992582321166992, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.4988, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 3.005469799041748, + "learning_rate": 3.111291632818847e-06, + "loss": 1.4999, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.7577736377716064, + "learning_rate": 3.109260763606824e-06, + "loss": 1.5152, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.724539041519165, + "eval_runtime": 37.0601, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.870109796524048, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.4746, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.7745094299316406, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.4891, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.6818361282348633, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.4575, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.823237180709839, + "learning_rate": 3.101137286758733e-06, + "loss": 1.492, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.9089932441711426, + "learning_rate": 3.09910641754671e-06, + "loss": 1.4943, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.85476016998291, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.5049, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 3.07242488861084, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.4709, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 3.117357015609741, + "learning_rate": 3.093013809910642e-06, + "loss": 1.52, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.959383964538574, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.4775, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8594462871551514, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.5113, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.721728801727295, + "eval_runtime": 37.1106, + "eval_samples_per_second": 26.946, + "eval_steps_per_second": 1.132, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.8418593406677246, + "learning_rate": 3.086921202274574e-06, + "loss": 1.482, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.762467384338379, + "learning_rate": 3.084890333062551e-06, + "loss": 1.4672, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.8023135662078857, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.431, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.7467143535614014, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.4837, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.6832900047302246, + "learning_rate": 3.078797725426483e-06, + "loss": 1.4765, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.863823175430298, + "learning_rate": 3.07676685621446e-06, + "loss": 1.4039, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.9055254459381104, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.4655, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.8665876388549805, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.449, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 3.029784917831421, + "learning_rate": 3.070674248578392e-06, + "loss": 1.4665, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.8582639694213867, + "learning_rate": 3.068643379366369e-06, + "loss": 1.4832, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.722991704940796, + "eval_runtime": 37.0827, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.8655312061309814, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.4759, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.892972469329834, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.4363, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.7701165676116943, + "learning_rate": 3.062550771730301e-06, + "loss": 1.4798, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7921078205108643, + "learning_rate": 3.060519902518278e-06, + "loss": 1.4557, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.8233699798583984, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.4598, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.8705663681030273, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.3898, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.9667348861694336, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.4865, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 3.0962018966674805, + "learning_rate": 3.052396425670187e-06, + "loss": 1.4512, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 3.06290340423584, + "learning_rate": 3.050365556458164e-06, + "loss": 1.5129, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.8010520935058594, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.4525, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.7228375673294067, + "eval_runtime": 37.1149, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.73380446434021, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.4537, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.9469099044799805, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.4624, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.8733043670654297, + "learning_rate": 3.042242079610073e-06, + "loss": 1.4981, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.867926597595215, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.4966, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.8584275245666504, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.4161, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.8059520721435547, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.4511, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.916750192642212, + "learning_rate": 3.034118602761982e-06, + "loss": 1.4694, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.979952335357666, + "learning_rate": 3.03208773354996e-06, + "loss": 1.4624, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.807194709777832, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.4227, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.880439519882202, + "learning_rate": 3.028025995125914e-06, + "loss": 1.5094, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.7204533815383911, + "eval_runtime": 37.1346, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.7696263790130615, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.492, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.840989351272583, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.4229, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.948915481567383, + "learning_rate": 3.021933387489846e-06, + "loss": 1.4601, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.9163503646850586, + "learning_rate": 3.019902518277823e-06, + "loss": 1.4218, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 3.0191619396209717, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.4648, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.9332194328308105, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.4791, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.8832461833953857, + "learning_rate": 3.013809910641755e-06, + "loss": 1.43, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.886837959289551, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.4455, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.8201355934143066, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.4417, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.9708259105682373, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.4657, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.7213743925094604, + "eval_runtime": 37.0993, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.927104949951172, + "learning_rate": 3.005686433793664e-06, + "loss": 1.4406, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.9701600074768066, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.4692, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 3.0241730213165283, + "learning_rate": 3.001624695369618e-06, + "loss": 1.4336, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.8802919387817383, + "learning_rate": 2.999593826157596e-06, + "loss": 1.4948, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.7666738033294678, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.4518, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 3.1206812858581543, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.4707, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 3.069323778152466, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.4534, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 3.155106544494629, + "learning_rate": 2.991470349309505e-06, + "loss": 1.5153, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.923816680908203, + "learning_rate": 2.989439480097482e-06, + "loss": 1.4606, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 3.2247867584228516, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.4564, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.7207131385803223, + "eval_runtime": 37.0684, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.8193411827087402, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.4721, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.865691661834717, + "learning_rate": 2.983346872461414e-06, + "loss": 1.5072, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 3.125004529953003, + "learning_rate": 2.981316003249391e-06, + "loss": 1.4913, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.873455286026001, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.4699, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.986186981201172, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.4808, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.8694097995758057, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.495, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.8555796146392822, + "learning_rate": 2.9731925264013e-06, + "loss": 1.4481, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.8808774948120117, + "learning_rate": 2.971161657189277e-06, + "loss": 1.4361, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.8831286430358887, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.4489, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.7533552646636963, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.4386, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.7256641387939453, + "eval_runtime": 37.1621, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 2.783586025238037, + "learning_rate": 2.965069049553209e-06, + "loss": 1.4476, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.9043776988983154, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5077, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.838284492492676, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.4287, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.94248366355896, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.4987, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.9397339820861816, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.4809, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.8922536373138428, + "learning_rate": 2.954914703493095e-06, + "loss": 1.426, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.787680149078369, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.4549, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.8116507530212402, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.4324, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.7330825328826904, + "learning_rate": 2.948822095857027e-06, + "loss": 1.4244, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.806539535522461, + "learning_rate": 2.946791226645004e-06, + "loss": 1.4604, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.7229852676391602, + "eval_runtime": 37.378, + "eval_samples_per_second": 26.754, + "eval_steps_per_second": 1.124, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.9335062503814697, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.4433, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.8057737350463867, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.4242, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 2.7581822872161865, + "learning_rate": 2.940698619008936e-06, + "loss": 1.4374, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.931168794631958, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.4588, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.865867853164673, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.4731, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.8826231956481934, + "learning_rate": 2.934606011372868e-06, + "loss": 1.4568, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.922809600830078, + "learning_rate": 2.932575142160845e-06, + "loss": 1.4747, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.9709854125976562, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.4631, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8259403705596924, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.449, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.850677490234375, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.4725, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.7200536727905273, + "eval_runtime": 36.8614, + "eval_samples_per_second": 27.129, + "eval_steps_per_second": 1.139, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.8449628353118896, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.5022, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 3.004880666732788, + "learning_rate": 2.922420796100731e-06, + "loss": 1.4606, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.737196445465088, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.3911, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.812662124633789, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.4873, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.7756738662719727, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.3975, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.8784940242767334, + "learning_rate": 2.91429731925264e-06, + "loss": 1.4726, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 2.946704626083374, + "learning_rate": 2.912266450040618e-06, + "loss": 1.4841, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.8813061714172363, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.4656, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.9221038818359375, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.4505, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 2.9353513717651367, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.4585, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.7200087308883667, + "eval_runtime": 36.9952, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.7565391063690186, + "learning_rate": 2.904142973192527e-06, + "loss": 1.4199, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 2.824780225753784, + "learning_rate": 2.902112103980504e-06, + "loss": 1.5003, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.963022232055664, + "learning_rate": 2.900081234768481e-06, + "loss": 1.4649, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.744490385055542, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.4677, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.8483076095581055, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.4291, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.878147602081299, + "learning_rate": 2.893988627132413e-06, + "loss": 1.4278, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.866380214691162, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.4491, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.802532911300659, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.397, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.863152265548706, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.4573, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.800173044204712, + "learning_rate": 2.885865150284322e-06, + "loss": 1.4466, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.7167619466781616, + "eval_runtime": 37.0136, + "eval_samples_per_second": 27.017, + "eval_steps_per_second": 1.135, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.975752830505371, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.4918, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.7768948078155518, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.4549, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.761004686355591, + "learning_rate": 2.879772542648254e-06, + "loss": 1.4039, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.9454665184020996, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.4542, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.85044264793396, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.3875, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.9699220657348633, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.5084, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.8547441959381104, + "learning_rate": 2.871649065800163e-06, + "loss": 1.4839, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 3.0871565341949463, + "learning_rate": 2.86961819658814e-06, + "loss": 1.5105, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 3.662945032119751, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.4483, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.916095495223999, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.455, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.7137683629989624, + "eval_runtime": 37.0176, + "eval_samples_per_second": 27.014, + "eval_steps_per_second": 1.135, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.857478141784668, + "learning_rate": 2.863525588952072e-06, + "loss": 1.4531, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.9542508125305176, + "learning_rate": 2.861494719740049e-06, + "loss": 1.488, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.677672863006592, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.3725, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.8790128231048584, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.4252, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.95095157623291, + "learning_rate": 2.855402112103981e-06, + "loss": 1.4532, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.997403383255005, + "learning_rate": 2.853371242891958e-06, + "loss": 1.4581, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.926973581314087, + "learning_rate": 2.851340373679935e-06, + "loss": 1.4368, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.8602488040924072, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.4427, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.875828504562378, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.4757, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.825498580932617, + "learning_rate": 2.845247766043867e-06, + "loss": 1.4535, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.7157984972000122, + "eval_runtime": 37.0329, + "eval_samples_per_second": 27.003, + "eval_steps_per_second": 1.134, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.8068923950195312, + "learning_rate": 2.843216896831844e-06, + "loss": 1.4391, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.8095016479492188, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.397, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.7768023014068604, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.4415, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.7868361473083496, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.4726, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 3.085806131362915, + "learning_rate": 2.835093419983753e-06, + "loss": 1.4735, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.972489833831787, + "learning_rate": 2.833062550771731e-06, + "loss": 1.4415, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.9990909099578857, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.4203, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.849210023880005, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.4202, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.899545907974243, + "learning_rate": 2.826969943135662e-06, + "loss": 1.4541, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.7620913982391357, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.4672, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.718942642211914, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.8633880615234375, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.4548, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.809718370437622, + "learning_rate": 2.820877335499594e-06, + "loss": 1.4379, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.780914306640625, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.4244, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.6853432655334473, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.3999, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.8040759563446045, + "learning_rate": 2.814784727863526e-06, + "loss": 1.4151, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.974795341491699, + "learning_rate": 2.812753858651503e-06, + "loss": 1.4497, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.9964492321014404, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.4512, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 2.9635047912597656, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.4104, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 3.0267832279205322, + "learning_rate": 2.806661251015435e-06, + "loss": 1.4149, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 3.0778963565826416, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.5034, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.723001480102539, + "eval_runtime": 37.0513, + "eval_samples_per_second": 26.99, + "eval_steps_per_second": 1.134, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.9889729022979736, + "learning_rate": 2.802599512591389e-06, + "loss": 1.4851, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 3.111581325531006, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.5337, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.727510452270508, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.4647, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.848571300506592, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.4107, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 3.178166627883911, + "learning_rate": 2.794476035743298e-06, + "loss": 1.4427, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.9924423694610596, + "learning_rate": 2.792445166531276e-06, + "loss": 1.5024, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.825277328491211, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.4687, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 3.082688808441162, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.5057, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.8399462699890137, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.4274, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.8092541694641113, + "learning_rate": 2.784321689683185e-06, + "loss": 1.4157, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.7190473079681396, + "eval_runtime": 37.0662, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.8798582553863525, + "learning_rate": 2.782290820471162e-06, + "loss": 1.4213, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.7891805171966553, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.4181, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.865835666656494, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.4896, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.996018171310425, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.4829, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.868534803390503, + "learning_rate": 2.774167343623071e-06, + "loss": 1.4592, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.906359910964966, + "learning_rate": 2.772136474411048e-06, + "loss": 1.5205, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 3.099578857421875, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.4246, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.944490432739258, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.3856, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.870277166366577, + "learning_rate": 2.76604386677498e-06, + "loss": 1.4648, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.853956937789917, + "learning_rate": 2.764012997562957e-06, + "loss": 1.4317, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.7165414094924927, + "eval_runtime": 37.4056, + "eval_samples_per_second": 26.734, + "eval_steps_per_second": 1.123, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 3.0553836822509766, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.4649, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.973536729812622, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.4129, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 3.025920867919922, + "learning_rate": 2.757920389926889e-06, + "loss": 1.4974, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.853586196899414, + "learning_rate": 2.755889520714866e-06, + "loss": 1.423, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 3.0477190017700195, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.4149, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.8930039405822754, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.3938, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.8823812007904053, + "learning_rate": 2.749796913078798e-06, + "loss": 1.46, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.96528959274292, + "learning_rate": 2.747766043866775e-06, + "loss": 1.4818, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.892967939376831, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.4392, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.851961851119995, + "learning_rate": 2.7437043054427297e-06, + "loss": 1.4903, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.7118008136749268, + "eval_runtime": 37.4958, + "eval_samples_per_second": 26.67, + "eval_steps_per_second": 1.12, + "step": 1400 + }, + { + "epoch": 1.5353424657534247, + "grad_norm": 2.9262213706970215, + "learning_rate": 2.741673436230707e-06, + "loss": 1.5253, + "step": 1401 + }, + { + "epoch": 1.5364383561643835, + "grad_norm": 3.0567355155944824, + "learning_rate": 2.739642567018684e-06, + "loss": 1.4442, + "step": 1402 + }, + { + "epoch": 1.5375342465753423, + "grad_norm": 3.0198726654052734, + "learning_rate": 2.7376116978066615e-06, + "loss": 1.4408, + "step": 1403 + }, + { + "epoch": 1.5386301369863014, + "grad_norm": 2.9126687049865723, + "learning_rate": 2.7355808285946388e-06, + "loss": 1.4421, + "step": 1404 + }, + { + "epoch": 1.5397260273972604, + "grad_norm": 2.9556353092193604, + "learning_rate": 2.733549959382616e-06, + "loss": 1.436, + "step": 1405 + }, + { + "epoch": 1.5408219178082192, + "grad_norm": 2.8351075649261475, + "learning_rate": 2.7315190901705933e-06, + "loss": 1.4444, + "step": 1406 + }, + { + "epoch": 1.541917808219178, + "grad_norm": 2.987910270690918, + "learning_rate": 2.7294882209585706e-06, + "loss": 1.4969, + "step": 1407 + }, + { + "epoch": 1.543013698630137, + "grad_norm": 2.734234571456909, + "learning_rate": 2.727457351746548e-06, + "loss": 1.4703, + "step": 1408 + }, + { + "epoch": 1.5441095890410959, + "grad_norm": 2.7853105068206787, + "learning_rate": 2.725426482534525e-06, + "loss": 1.4241, + "step": 1409 + }, + { + "epoch": 1.5452054794520547, + "grad_norm": 2.856630802154541, + "learning_rate": 2.723395613322502e-06, + "loss": 1.3985, + "step": 1410 + }, + { + "epoch": 1.5452054794520547, + "eval_loss": 1.7150108814239502, + "eval_runtime": 36.8966, + "eval_samples_per_second": 27.103, + "eval_steps_per_second": 1.138, + "step": 1410 + }, + { + "epoch": 1.5463013698630137, + "grad_norm": 2.952057123184204, + "learning_rate": 2.7213647441104797e-06, + "loss": 1.4453, + "step": 1411 + }, + { + "epoch": 1.5473972602739727, + "grad_norm": 2.855661153793335, + "learning_rate": 2.7193338748984565e-06, + "loss": 1.4288, + "step": 1412 + }, + { + "epoch": 1.5484931506849315, + "grad_norm": 2.9883601665496826, + "learning_rate": 2.7173030056864342e-06, + "loss": 1.436, + "step": 1413 + }, + { + "epoch": 1.5495890410958904, + "grad_norm": 3.175748109817505, + "learning_rate": 2.715272136474411e-06, + "loss": 1.5645, + "step": 1414 + }, + { + "epoch": 1.5506849315068494, + "grad_norm": 2.8751583099365234, + "learning_rate": 2.713241267262389e-06, + "loss": 1.4167, + "step": 1415 + }, + { + "epoch": 1.5517808219178082, + "grad_norm": 2.9523749351501465, + "learning_rate": 2.7112103980503656e-06, + "loss": 1.4379, + "step": 1416 + }, + { + "epoch": 1.552876712328767, + "grad_norm": 2.709944486618042, + "learning_rate": 2.7091795288383433e-06, + "loss": 1.4286, + "step": 1417 + }, + { + "epoch": 1.553972602739726, + "grad_norm": 2.782998561859131, + "learning_rate": 2.70714865962632e-06, + "loss": 1.4758, + "step": 1418 + }, + { + "epoch": 1.555068493150685, + "grad_norm": 2.803555965423584, + "learning_rate": 2.705117790414298e-06, + "loss": 1.4211, + "step": 1419 + }, + { + "epoch": 1.5561643835616439, + "grad_norm": 2.792894124984741, + "learning_rate": 2.7030869212022747e-06, + "loss": 1.4532, + "step": 1420 + }, + { + "epoch": 1.5561643835616439, + "eval_loss": 1.713323712348938, + "eval_runtime": 37.0626, + "eval_samples_per_second": 26.981, + "eval_steps_per_second": 1.133, + "step": 1420 + }, + { + "epoch": 1.5572602739726027, + "grad_norm": 2.9909141063690186, + "learning_rate": 2.7010560519902516e-06, + "loss": 1.4718, + "step": 1421 + }, + { + "epoch": 1.5583561643835617, + "grad_norm": 2.9359354972839355, + "learning_rate": 2.6990251827782293e-06, + "loss": 1.444, + "step": 1422 + }, + { + "epoch": 1.5594520547945205, + "grad_norm": 2.986454963684082, + "learning_rate": 2.696994313566206e-06, + "loss": 1.45, + "step": 1423 + }, + { + "epoch": 1.5605479452054793, + "grad_norm": 2.997868776321411, + "learning_rate": 2.694963444354184e-06, + "loss": 1.4309, + "step": 1424 + }, + { + "epoch": 1.5616438356164384, + "grad_norm": 2.922935962677002, + "learning_rate": 2.6929325751421607e-06, + "loss": 1.4961, + "step": 1425 + }, + { + "epoch": 1.5627397260273974, + "grad_norm": 2.7861900329589844, + "learning_rate": 2.6909017059301384e-06, + "loss": 1.4482, + "step": 1426 + }, + { + "epoch": 1.563835616438356, + "grad_norm": 2.848219394683838, + "learning_rate": 2.6888708367181153e-06, + "loss": 1.4637, + "step": 1427 + }, + { + "epoch": 1.564931506849315, + "grad_norm": 2.6985557079315186, + "learning_rate": 2.686839967506093e-06, + "loss": 1.4138, + "step": 1428 + }, + { + "epoch": 1.566027397260274, + "grad_norm": 2.835800886154175, + "learning_rate": 2.68480909829407e-06, + "loss": 1.4669, + "step": 1429 + }, + { + "epoch": 1.5671232876712329, + "grad_norm": 2.9267354011535645, + "learning_rate": 2.6827782290820475e-06, + "loss": 1.3823, + "step": 1430 + }, + { + "epoch": 1.5671232876712329, + "eval_loss": 1.7126407623291016, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 1430 + }, + { + "epoch": 1.5682191780821917, + "grad_norm": 2.8740406036376953, + "learning_rate": 2.6807473598700244e-06, + "loss": 1.3837, + "step": 1431 + }, + { + "epoch": 1.5693150684931507, + "grad_norm": 3.012890100479126, + "learning_rate": 2.678716490658002e-06, + "loss": 1.495, + "step": 1432 + }, + { + "epoch": 1.5704109589041098, + "grad_norm": 2.877326011657715, + "learning_rate": 2.676685621445979e-06, + "loss": 1.4622, + "step": 1433 + }, + { + "epoch": 1.5715068493150683, + "grad_norm": 2.852553606033325, + "learning_rate": 2.674654752233956e-06, + "loss": 1.4415, + "step": 1434 + }, + { + "epoch": 1.5726027397260274, + "grad_norm": 2.9420814514160156, + "learning_rate": 2.6726238830219335e-06, + "loss": 1.4463, + "step": 1435 + }, + { + "epoch": 1.5736986301369864, + "grad_norm": 2.91435170173645, + "learning_rate": 2.6705930138099107e-06, + "loss": 1.4482, + "step": 1436 + }, + { + "epoch": 1.5747945205479452, + "grad_norm": 2.851418972015381, + "learning_rate": 2.668562144597888e-06, + "loss": 1.4752, + "step": 1437 + }, + { + "epoch": 1.575890410958904, + "grad_norm": 2.7865328788757324, + "learning_rate": 2.6665312753858653e-06, + "loss": 1.3752, + "step": 1438 + }, + { + "epoch": 1.576986301369863, + "grad_norm": 2.8747165203094482, + "learning_rate": 2.6645004061738426e-06, + "loss": 1.4709, + "step": 1439 + }, + { + "epoch": 1.5780821917808219, + "grad_norm": 2.986900568008423, + "learning_rate": 2.66246953696182e-06, + "loss": 1.4524, + "step": 1440 + }, + { + "epoch": 1.5780821917808219, + "eval_loss": 1.7110822200775146, + "eval_runtime": 37.1062, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 1440 + }, + { + "epoch": 1.5791780821917807, + "grad_norm": 2.9051966667175293, + "learning_rate": 2.660438667749797e-06, + "loss": 1.393, + "step": 1441 + }, + { + "epoch": 1.5802739726027397, + "grad_norm": 2.8744468688964844, + "learning_rate": 2.6584077985377744e-06, + "loss": 1.4233, + "step": 1442 + }, + { + "epoch": 1.5813698630136988, + "grad_norm": 2.8655996322631836, + "learning_rate": 2.6563769293257517e-06, + "loss": 1.4501, + "step": 1443 + }, + { + "epoch": 1.5824657534246576, + "grad_norm": 2.8580846786499023, + "learning_rate": 2.654346060113729e-06, + "loss": 1.4612, + "step": 1444 + }, + { + "epoch": 1.5835616438356164, + "grad_norm": 2.824436902999878, + "learning_rate": 2.6523151909017062e-06, + "loss": 1.4263, + "step": 1445 + }, + { + "epoch": 1.5846575342465754, + "grad_norm": 2.8548355102539062, + "learning_rate": 2.6502843216896835e-06, + "loss": 1.4219, + "step": 1446 + }, + { + "epoch": 1.5857534246575342, + "grad_norm": 2.8495287895202637, + "learning_rate": 2.6482534524776603e-06, + "loss": 1.419, + "step": 1447 + }, + { + "epoch": 1.586849315068493, + "grad_norm": 2.7665646076202393, + "learning_rate": 2.646222583265638e-06, + "loss": 1.5077, + "step": 1448 + }, + { + "epoch": 1.587945205479452, + "grad_norm": 2.859886646270752, + "learning_rate": 2.644191714053615e-06, + "loss": 1.4687, + "step": 1449 + }, + { + "epoch": 1.589041095890411, + "grad_norm": 2.8246665000915527, + "learning_rate": 2.6421608448415926e-06, + "loss": 1.4557, + "step": 1450 + }, + { + "epoch": 1.589041095890411, + "eval_loss": 1.7074187994003296, + "eval_runtime": 37.0881, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 1450 + }, + { + "epoch": 1.59013698630137, + "grad_norm": 2.8395724296569824, + "learning_rate": 2.6401299756295694e-06, + "loss": 1.4351, + "step": 1451 + }, + { + "epoch": 1.5912328767123287, + "grad_norm": 2.7422566413879395, + "learning_rate": 2.638099106417547e-06, + "loss": 1.4056, + "step": 1452 + }, + { + "epoch": 1.5923287671232877, + "grad_norm": 2.911670446395874, + "learning_rate": 2.636068237205524e-06, + "loss": 1.4522, + "step": 1453 + }, + { + "epoch": 1.5934246575342466, + "grad_norm": 2.8709356784820557, + "learning_rate": 2.6340373679935017e-06, + "loss": 1.4466, + "step": 1454 + }, + { + "epoch": 1.5945205479452054, + "grad_norm": 2.9503116607666016, + "learning_rate": 2.6320064987814785e-06, + "loss": 1.5055, + "step": 1455 + }, + { + "epoch": 1.5956164383561644, + "grad_norm": 2.752567768096924, + "learning_rate": 2.6299756295694562e-06, + "loss": 1.3711, + "step": 1456 + }, + { + "epoch": 1.5967123287671234, + "grad_norm": 2.907646656036377, + "learning_rate": 2.627944760357433e-06, + "loss": 1.4455, + "step": 1457 + }, + { + "epoch": 1.5978082191780822, + "grad_norm": 2.7967607975006104, + "learning_rate": 2.6259138911454104e-06, + "loss": 1.4972, + "step": 1458 + }, + { + "epoch": 1.598904109589041, + "grad_norm": 2.854980707168579, + "learning_rate": 2.6238830219333877e-06, + "loss": 1.4957, + "step": 1459 + }, + { + "epoch": 1.6, + "grad_norm": 2.8644754886627197, + "learning_rate": 2.621852152721365e-06, + "loss": 1.478, + "step": 1460 + }, + { + "epoch": 1.6, + "eval_loss": 1.706020712852478, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 1460 + }, + { + "epoch": 1.601095890410959, + "grad_norm": 2.7806358337402344, + "learning_rate": 2.619821283509342e-06, + "loss": 1.4027, + "step": 1461 + }, + { + "epoch": 1.6021917808219177, + "grad_norm": 2.829030990600586, + "learning_rate": 2.6177904142973195e-06, + "loss": 1.4583, + "step": 1462 + }, + { + "epoch": 1.6032876712328767, + "grad_norm": 2.8990888595581055, + "learning_rate": 2.6157595450852968e-06, + "loss": 1.4071, + "step": 1463 + }, + { + "epoch": 1.6043835616438358, + "grad_norm": 3.0500295162200928, + "learning_rate": 2.613728675873274e-06, + "loss": 1.4581, + "step": 1464 + }, + { + "epoch": 1.6054794520547944, + "grad_norm": 3.0562543869018555, + "learning_rate": 2.6116978066612513e-06, + "loss": 1.4953, + "step": 1465 + }, + { + "epoch": 1.6065753424657534, + "grad_norm": 2.992666721343994, + "learning_rate": 2.6096669374492286e-06, + "loss": 1.4938, + "step": 1466 + }, + { + "epoch": 1.6076712328767124, + "grad_norm": 2.895270586013794, + "learning_rate": 2.607636068237206e-06, + "loss": 1.4485, + "step": 1467 + }, + { + "epoch": 1.6087671232876712, + "grad_norm": 2.942186117172241, + "learning_rate": 2.605605199025183e-06, + "loss": 1.4303, + "step": 1468 + }, + { + "epoch": 1.60986301369863, + "grad_norm": 2.8370566368103027, + "learning_rate": 2.6035743298131604e-06, + "loss": 1.4619, + "step": 1469 + }, + { + "epoch": 1.610958904109589, + "grad_norm": 2.885728359222412, + "learning_rate": 2.6015434606011377e-06, + "loss": 1.449, + "step": 1470 + }, + { + "epoch": 1.610958904109589, + "eval_loss": 1.708180546760559, + "eval_runtime": 37.0887, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 1470 + }, + { + "epoch": 1.612054794520548, + "grad_norm": 3.0368149280548096, + "learning_rate": 2.5995125913891145e-06, + "loss": 1.4799, + "step": 1471 + }, + { + "epoch": 1.6131506849315067, + "grad_norm": 3.136568546295166, + "learning_rate": 2.5974817221770922e-06, + "loss": 1.4801, + "step": 1472 + }, + { + "epoch": 1.6142465753424657, + "grad_norm": 2.9858787059783936, + "learning_rate": 2.595450852965069e-06, + "loss": 1.449, + "step": 1473 + }, + { + "epoch": 1.6153424657534248, + "grad_norm": 2.9646689891815186, + "learning_rate": 2.593419983753047e-06, + "loss": 1.4698, + "step": 1474 + }, + { + "epoch": 1.6164383561643836, + "grad_norm": 3.0527894496917725, + "learning_rate": 2.5913891145410236e-06, + "loss": 1.497, + "step": 1475 + }, + { + "epoch": 1.6175342465753424, + "grad_norm": 2.8184142112731934, + "learning_rate": 2.5893582453290013e-06, + "loss": 1.4421, + "step": 1476 + }, + { + "epoch": 1.6186301369863014, + "grad_norm": 2.7851176261901855, + "learning_rate": 2.587327376116978e-06, + "loss": 1.4666, + "step": 1477 + }, + { + "epoch": 1.6197260273972602, + "grad_norm": 2.7636115550994873, + "learning_rate": 2.585296506904956e-06, + "loss": 1.4368, + "step": 1478 + }, + { + "epoch": 1.620821917808219, + "grad_norm": 2.9213268756866455, + "learning_rate": 2.5832656376929327e-06, + "loss": 1.4408, + "step": 1479 + }, + { + "epoch": 1.621917808219178, + "grad_norm": 2.865529775619507, + "learning_rate": 2.5812347684809104e-06, + "loss": 1.4144, + "step": 1480 + }, + { + "epoch": 1.621917808219178, + "eval_loss": 1.7098593711853027, + "eval_runtime": 37.0613, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1480 + }, + { + "epoch": 1.6230136986301371, + "grad_norm": 2.840822458267212, + "learning_rate": 2.5792038992688873e-06, + "loss": 1.3843, + "step": 1481 + }, + { + "epoch": 1.624109589041096, + "grad_norm": 3.051424980163574, + "learning_rate": 2.577173030056865e-06, + "loss": 1.5298, + "step": 1482 + }, + { + "epoch": 1.6252054794520547, + "grad_norm": 2.995697498321533, + "learning_rate": 2.575142160844842e-06, + "loss": 1.3922, + "step": 1483 + }, + { + "epoch": 1.6263013698630138, + "grad_norm": 3.0062193870544434, + "learning_rate": 2.5731112916328187e-06, + "loss": 1.4704, + "step": 1484 + }, + { + "epoch": 1.6273972602739726, + "grad_norm": 2.9391777515411377, + "learning_rate": 2.5710804224207964e-06, + "loss": 1.481, + "step": 1485 + }, + { + "epoch": 1.6284931506849314, + "grad_norm": 2.9193878173828125, + "learning_rate": 2.5690495532087733e-06, + "loss": 1.4038, + "step": 1486 + }, + { + "epoch": 1.6295890410958904, + "grad_norm": 2.8676085472106934, + "learning_rate": 2.567018683996751e-06, + "loss": 1.3752, + "step": 1487 + }, + { + "epoch": 1.6306849315068495, + "grad_norm": 2.9337565898895264, + "learning_rate": 2.564987814784728e-06, + "loss": 1.4469, + "step": 1488 + }, + { + "epoch": 1.6317808219178083, + "grad_norm": 2.9485421180725098, + "learning_rate": 2.5629569455727055e-06, + "loss": 1.4224, + "step": 1489 + }, + { + "epoch": 1.632876712328767, + "grad_norm": 2.9687442779541016, + "learning_rate": 2.5609260763606824e-06, + "loss": 1.4347, + "step": 1490 + }, + { + "epoch": 1.632876712328767, + "eval_loss": 1.7070544958114624, + "eval_runtime": 37.3446, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 1490 + }, + { + "epoch": 1.633972602739726, + "grad_norm": 2.913801908493042, + "learning_rate": 2.55889520714866e-06, + "loss": 1.4478, + "step": 1491 + }, + { + "epoch": 1.635068493150685, + "grad_norm": 3.0276856422424316, + "learning_rate": 2.556864337936637e-06, + "loss": 1.5372, + "step": 1492 + }, + { + "epoch": 1.6361643835616437, + "grad_norm": 2.936163902282715, + "learning_rate": 2.5548334687246146e-06, + "loss": 1.4283, + "step": 1493 + }, + { + "epoch": 1.6372602739726028, + "grad_norm": 2.8936173915863037, + "learning_rate": 2.5528025995125915e-06, + "loss": 1.432, + "step": 1494 + }, + { + "epoch": 1.6383561643835618, + "grad_norm": 3.016876459121704, + "learning_rate": 2.550771730300569e-06, + "loss": 1.4966, + "step": 1495 + }, + { + "epoch": 1.6394520547945204, + "grad_norm": 2.8830440044403076, + "learning_rate": 2.548740861088546e-06, + "loss": 1.4408, + "step": 1496 + }, + { + "epoch": 1.6405479452054794, + "grad_norm": 2.9123940467834473, + "learning_rate": 2.5467099918765233e-06, + "loss": 1.4756, + "step": 1497 + }, + { + "epoch": 1.6416438356164385, + "grad_norm": 2.7645750045776367, + "learning_rate": 2.5446791226645006e-06, + "loss": 1.3705, + "step": 1498 + }, + { + "epoch": 1.6427397260273973, + "grad_norm": 2.8265559673309326, + "learning_rate": 2.542648253452478e-06, + "loss": 1.3738, + "step": 1499 + }, + { + "epoch": 1.643835616438356, + "grad_norm": 3.1146469116210938, + "learning_rate": 2.540617384240455e-06, + "loss": 1.4712, + "step": 1500 + }, + { + "epoch": 1.643835616438356, + "eval_loss": 1.714676022529602, + "eval_runtime": 37.4044, + "eval_samples_per_second": 26.735, + "eval_steps_per_second": 1.123, + "step": 1500 + }, + { + "epoch": 1.644931506849315, + "grad_norm": 3.0422415733337402, + "learning_rate": 2.5385865150284324e-06, + "loss": 1.4514, + "step": 1501 + }, + { + "epoch": 1.6460273972602741, + "grad_norm": 2.9049177169799805, + "learning_rate": 2.5365556458164097e-06, + "loss": 1.408, + "step": 1502 + }, + { + "epoch": 1.6471232876712327, + "grad_norm": 2.9806294441223145, + "learning_rate": 2.534524776604387e-06, + "loss": 1.4423, + "step": 1503 + }, + { + "epoch": 1.6482191780821918, + "grad_norm": 3.1372151374816895, + "learning_rate": 2.5324939073923642e-06, + "loss": 1.4368, + "step": 1504 + }, + { + "epoch": 1.6493150684931508, + "grad_norm": 2.778329849243164, + "learning_rate": 2.5304630381803415e-06, + "loss": 1.3559, + "step": 1505 + }, + { + "epoch": 1.6504109589041096, + "grad_norm": 2.935122013092041, + "learning_rate": 2.5284321689683188e-06, + "loss": 1.4299, + "step": 1506 + }, + { + "epoch": 1.6515068493150684, + "grad_norm": 2.8406238555908203, + "learning_rate": 2.526401299756296e-06, + "loss": 1.4496, + "step": 1507 + }, + { + "epoch": 1.6526027397260274, + "grad_norm": 2.919686794281006, + "learning_rate": 2.524370430544273e-06, + "loss": 1.446, + "step": 1508 + }, + { + "epoch": 1.6536986301369863, + "grad_norm": 2.8712282180786133, + "learning_rate": 2.5223395613322506e-06, + "loss": 1.4192, + "step": 1509 + }, + { + "epoch": 1.654794520547945, + "grad_norm": 2.984957695007324, + "learning_rate": 2.5203086921202274e-06, + "loss": 1.4234, + "step": 1510 + }, + { + "epoch": 1.654794520547945, + "eval_loss": 1.7065510749816895, + "eval_runtime": 36.8721, + "eval_samples_per_second": 27.121, + "eval_steps_per_second": 1.139, + "step": 1510 + }, + { + "epoch": 1.655890410958904, + "grad_norm": 3.0523569583892822, + "learning_rate": 2.518277822908205e-06, + "loss": 1.4488, + "step": 1511 + }, + { + "epoch": 1.6569863013698631, + "grad_norm": 3.1737515926361084, + "learning_rate": 2.516246953696182e-06, + "loss": 1.4311, + "step": 1512 + }, + { + "epoch": 1.658082191780822, + "grad_norm": 2.9754984378814697, + "learning_rate": 2.5142160844841597e-06, + "loss": 1.4363, + "step": 1513 + }, + { + "epoch": 1.6591780821917808, + "grad_norm": 3.0981898307800293, + "learning_rate": 2.5121852152721365e-06, + "loss": 1.4546, + "step": 1514 + }, + { + "epoch": 1.6602739726027398, + "grad_norm": 2.8706960678100586, + "learning_rate": 2.5101543460601142e-06, + "loss": 1.4599, + "step": 1515 + }, + { + "epoch": 1.6613698630136986, + "grad_norm": 2.9565646648406982, + "learning_rate": 2.508123476848091e-06, + "loss": 1.4471, + "step": 1516 + }, + { + "epoch": 1.6624657534246574, + "grad_norm": 2.687807083129883, + "learning_rate": 2.506092607636069e-06, + "loss": 1.3926, + "step": 1517 + }, + { + "epoch": 1.6635616438356164, + "grad_norm": 2.832486629486084, + "learning_rate": 2.5040617384240457e-06, + "loss": 1.4646, + "step": 1518 + }, + { + "epoch": 1.6646575342465755, + "grad_norm": 2.946368455886841, + "learning_rate": 2.5020308692120234e-06, + "loss": 1.4347, + "step": 1519 + }, + { + "epoch": 1.6657534246575343, + "grad_norm": 2.7646501064300537, + "learning_rate": 2.5e-06, + "loss": 1.4214, + "step": 1520 + }, + { + "epoch": 1.6657534246575343, + "eval_loss": 1.7065269947052002, + "eval_runtime": 36.9819, + "eval_samples_per_second": 27.04, + "eval_steps_per_second": 1.136, + "step": 1520 + }, + { + "epoch": 1.666849315068493, + "grad_norm": 2.995950698852539, + "learning_rate": 2.4979691307879775e-06, + "loss": 1.4986, + "step": 1521 + }, + { + "epoch": 1.6679452054794521, + "grad_norm": 2.886826515197754, + "learning_rate": 2.4959382615759548e-06, + "loss": 1.4319, + "step": 1522 + }, + { + "epoch": 1.669041095890411, + "grad_norm": 3.0105748176574707, + "learning_rate": 2.493907392363932e-06, + "loss": 1.4297, + "step": 1523 + }, + { + "epoch": 1.6701369863013698, + "grad_norm": 2.892512083053589, + "learning_rate": 2.4918765231519093e-06, + "loss": 1.4857, + "step": 1524 + }, + { + "epoch": 1.6712328767123288, + "grad_norm": 2.8092124462127686, + "learning_rate": 2.4898456539398866e-06, + "loss": 1.4304, + "step": 1525 + }, + { + "epoch": 1.6723287671232878, + "grad_norm": 2.8145992755889893, + "learning_rate": 2.487814784727864e-06, + "loss": 1.4559, + "step": 1526 + }, + { + "epoch": 1.6734246575342466, + "grad_norm": 2.894888401031494, + "learning_rate": 2.4857839155158407e-06, + "loss": 1.497, + "step": 1527 + }, + { + "epoch": 1.6745205479452054, + "grad_norm": 2.8759849071502686, + "learning_rate": 2.483753046303818e-06, + "loss": 1.4295, + "step": 1528 + }, + { + "epoch": 1.6756164383561645, + "grad_norm": 2.8624649047851562, + "learning_rate": 2.4817221770917953e-06, + "loss": 1.4369, + "step": 1529 + }, + { + "epoch": 1.6767123287671233, + "grad_norm": 2.860114574432373, + "learning_rate": 2.4796913078797725e-06, + "loss": 1.4389, + "step": 1530 + }, + { + "epoch": 1.6767123287671233, + "eval_loss": 1.708011269569397, + "eval_runtime": 37.0257, + "eval_samples_per_second": 27.008, + "eval_steps_per_second": 1.134, + "step": 1530 + }, + { + "epoch": 1.677808219178082, + "grad_norm": 3.062593460083008, + "learning_rate": 2.47766043866775e-06, + "loss": 1.4679, + "step": 1531 + }, + { + "epoch": 1.6789041095890411, + "grad_norm": 2.9316751956939697, + "learning_rate": 2.475629569455727e-06, + "loss": 1.4712, + "step": 1532 + }, + { + "epoch": 1.6800000000000002, + "grad_norm": 2.799980401992798, + "learning_rate": 2.4735987002437044e-06, + "loss": 1.4764, + "step": 1533 + }, + { + "epoch": 1.6810958904109587, + "grad_norm": 2.71172833442688, + "learning_rate": 2.4715678310316816e-06, + "loss": 1.4103, + "step": 1534 + }, + { + "epoch": 1.6821917808219178, + "grad_norm": 2.808029890060425, + "learning_rate": 2.469536961819659e-06, + "loss": 1.4134, + "step": 1535 + }, + { + "epoch": 1.6832876712328768, + "grad_norm": 2.8933329582214355, + "learning_rate": 2.467506092607636e-06, + "loss": 1.4301, + "step": 1536 + }, + { + "epoch": 1.6843835616438356, + "grad_norm": 2.866718053817749, + "learning_rate": 2.4654752233956135e-06, + "loss": 1.4438, + "step": 1537 + }, + { + "epoch": 1.6854794520547944, + "grad_norm": 2.8831021785736084, + "learning_rate": 2.4634443541835907e-06, + "loss": 1.4182, + "step": 1538 + }, + { + "epoch": 1.6865753424657535, + "grad_norm": 3.140631914138794, + "learning_rate": 2.461413484971568e-06, + "loss": 1.5008, + "step": 1539 + }, + { + "epoch": 1.6876712328767123, + "grad_norm": 3.0420989990234375, + "learning_rate": 2.4593826157595453e-06, + "loss": 1.4454, + "step": 1540 + }, + { + "epoch": 1.6876712328767123, + "eval_loss": 1.7109777927398682, + "eval_runtime": 37.0355, + "eval_samples_per_second": 27.001, + "eval_steps_per_second": 1.134, + "step": 1540 + }, + { + "epoch": 1.688767123287671, + "grad_norm": 3.0079169273376465, + "learning_rate": 2.4573517465475226e-06, + "loss": 1.5109, + "step": 1541 + }, + { + "epoch": 1.6898630136986301, + "grad_norm": 2.846114158630371, + "learning_rate": 2.4553208773355e-06, + "loss": 1.4035, + "step": 1542 + }, + { + "epoch": 1.6909589041095892, + "grad_norm": 2.828695297241211, + "learning_rate": 2.453290008123477e-06, + "loss": 1.4166, + "step": 1543 + }, + { + "epoch": 1.692054794520548, + "grad_norm": 2.824759006500244, + "learning_rate": 2.4512591389114544e-06, + "loss": 1.4421, + "step": 1544 + }, + { + "epoch": 1.6931506849315068, + "grad_norm": 2.8216183185577393, + "learning_rate": 2.4492282696994317e-06, + "loss": 1.4654, + "step": 1545 + }, + { + "epoch": 1.6942465753424658, + "grad_norm": 2.879293441772461, + "learning_rate": 2.447197400487409e-06, + "loss": 1.4511, + "step": 1546 + }, + { + "epoch": 1.6953424657534246, + "grad_norm": 2.98406720161438, + "learning_rate": 2.4451665312753862e-06, + "loss": 1.4164, + "step": 1547 + }, + { + "epoch": 1.6964383561643834, + "grad_norm": 3.0683135986328125, + "learning_rate": 2.4431356620633635e-06, + "loss": 1.4563, + "step": 1548 + }, + { + "epoch": 1.6975342465753425, + "grad_norm": 2.986192464828491, + "learning_rate": 2.4411047928513408e-06, + "loss": 1.4643, + "step": 1549 + }, + { + "epoch": 1.6986301369863015, + "grad_norm": 2.923180341720581, + "learning_rate": 2.439073923639318e-06, + "loss": 1.4546, + "step": 1550 + }, + { + "epoch": 1.6986301369863015, + "eval_loss": 1.7101587057113647, + "eval_runtime": 37.0292, + "eval_samples_per_second": 27.006, + "eval_steps_per_second": 1.134, + "step": 1550 + }, + { + "epoch": 1.6997260273972603, + "grad_norm": 2.928861141204834, + "learning_rate": 2.4370430544272953e-06, + "loss": 1.4424, + "step": 1551 + }, + { + "epoch": 1.7008219178082191, + "grad_norm": 2.8597733974456787, + "learning_rate": 2.435012185215272e-06, + "loss": 1.3872, + "step": 1552 + }, + { + "epoch": 1.7019178082191782, + "grad_norm": 3.054025650024414, + "learning_rate": 2.4329813160032495e-06, + "loss": 1.3912, + "step": 1553 + }, + { + "epoch": 1.703013698630137, + "grad_norm": 2.824483871459961, + "learning_rate": 2.4309504467912267e-06, + "loss": 1.4167, + "step": 1554 + }, + { + "epoch": 1.7041095890410958, + "grad_norm": 2.9558396339416504, + "learning_rate": 2.428919577579204e-06, + "loss": 1.4436, + "step": 1555 + }, + { + "epoch": 1.7052054794520548, + "grad_norm": 2.8712751865386963, + "learning_rate": 2.4268887083671813e-06, + "loss": 1.4198, + "step": 1556 + }, + { + "epoch": 1.7063013698630138, + "grad_norm": 2.9776816368103027, + "learning_rate": 2.4248578391551586e-06, + "loss": 1.5133, + "step": 1557 + }, + { + "epoch": 1.7073972602739727, + "grad_norm": 2.930847644805908, + "learning_rate": 2.422826969943136e-06, + "loss": 1.4239, + "step": 1558 + }, + { + "epoch": 1.7084931506849315, + "grad_norm": 2.9059464931488037, + "learning_rate": 2.420796100731113e-06, + "loss": 1.4071, + "step": 1559 + }, + { + "epoch": 1.7095890410958905, + "grad_norm": 2.8655083179473877, + "learning_rate": 2.4187652315190904e-06, + "loss": 1.3911, + "step": 1560 + }, + { + "epoch": 1.7095890410958905, + "eval_loss": 1.705108880996704, + "eval_runtime": 37.0068, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 1560 + }, + { + "epoch": 1.7106849315068493, + "grad_norm": 2.883193016052246, + "learning_rate": 2.4167343623070677e-06, + "loss": 1.444, + "step": 1561 + }, + { + "epoch": 1.7117808219178081, + "grad_norm": 2.8185808658599854, + "learning_rate": 2.414703493095045e-06, + "loss": 1.4444, + "step": 1562 + }, + { + "epoch": 1.7128767123287671, + "grad_norm": 2.8110158443450928, + "learning_rate": 2.4126726238830222e-06, + "loss": 1.4211, + "step": 1563 + }, + { + "epoch": 1.7139726027397262, + "grad_norm": 2.863987445831299, + "learning_rate": 2.410641754670999e-06, + "loss": 1.4319, + "step": 1564 + }, + { + "epoch": 1.7150684931506848, + "grad_norm": 2.977844715118408, + "learning_rate": 2.4086108854589763e-06, + "loss": 1.4112, + "step": 1565 + }, + { + "epoch": 1.7161643835616438, + "grad_norm": 2.8599488735198975, + "learning_rate": 2.4065800162469536e-06, + "loss": 1.4657, + "step": 1566 + }, + { + "epoch": 1.7172602739726028, + "grad_norm": 2.8669981956481934, + "learning_rate": 2.404549147034931e-06, + "loss": 1.4214, + "step": 1567 + }, + { + "epoch": 1.7183561643835616, + "grad_norm": 2.980116367340088, + "learning_rate": 2.402518277822908e-06, + "loss": 1.4189, + "step": 1568 + }, + { + "epoch": 1.7194520547945205, + "grad_norm": 2.936729907989502, + "learning_rate": 2.4004874086108854e-06, + "loss": 1.4689, + "step": 1569 + }, + { + "epoch": 1.7205479452054795, + "grad_norm": 2.8469579219818115, + "learning_rate": 2.3984565393988627e-06, + "loss": 1.4137, + "step": 1570 + }, + { + "epoch": 1.7205479452054795, + "eval_loss": 1.7018489837646484, + "eval_runtime": 37.0142, + "eval_samples_per_second": 27.017, + "eval_steps_per_second": 1.135, + "step": 1570 + }, + { + "epoch": 1.7216438356164385, + "grad_norm": 2.9605817794799805, + "learning_rate": 2.39642567018684e-06, + "loss": 1.4449, + "step": 1571 + }, + { + "epoch": 1.7227397260273971, + "grad_norm": 2.815717935562134, + "learning_rate": 2.3943948009748173e-06, + "loss": 1.4114, + "step": 1572 + }, + { + "epoch": 1.7238356164383561, + "grad_norm": 2.961744785308838, + "learning_rate": 2.3923639317627945e-06, + "loss": 1.4306, + "step": 1573 + }, + { + "epoch": 1.7249315068493152, + "grad_norm": 3.022645950317383, + "learning_rate": 2.390333062550772e-06, + "loss": 1.4619, + "step": 1574 + }, + { + "epoch": 1.726027397260274, + "grad_norm": 2.9433681964874268, + "learning_rate": 2.388302193338749e-06, + "loss": 1.401, + "step": 1575 + }, + { + "epoch": 1.7271232876712328, + "grad_norm": 2.794745683670044, + "learning_rate": 2.3862713241267264e-06, + "loss": 1.3441, + "step": 1576 + }, + { + "epoch": 1.7282191780821918, + "grad_norm": 3.041825294494629, + "learning_rate": 2.3842404549147037e-06, + "loss": 1.4938, + "step": 1577 + }, + { + "epoch": 1.7293150684931506, + "grad_norm": 2.9499473571777344, + "learning_rate": 2.382209585702681e-06, + "loss": 1.4295, + "step": 1578 + }, + { + "epoch": 1.7304109589041095, + "grad_norm": 2.8626675605773926, + "learning_rate": 2.380178716490658e-06, + "loss": 1.3911, + "step": 1579 + }, + { + "epoch": 1.7315068493150685, + "grad_norm": 2.842794179916382, + "learning_rate": 2.3781478472786355e-06, + "loss": 1.4737, + "step": 1580 + }, + { + "epoch": 1.7315068493150685, + "eval_loss": 1.7021467685699463, + "eval_runtime": 37.1037, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1580 + }, + { + "epoch": 1.7326027397260275, + "grad_norm": 2.8970463275909424, + "learning_rate": 2.3761169780666128e-06, + "loss": 1.4209, + "step": 1581 + }, + { + "epoch": 1.7336986301369863, + "grad_norm": 2.959874153137207, + "learning_rate": 2.37408610885459e-06, + "loss": 1.4662, + "step": 1582 + }, + { + "epoch": 1.7347945205479451, + "grad_norm": 3.0744926929473877, + "learning_rate": 2.3720552396425673e-06, + "loss": 1.4655, + "step": 1583 + }, + { + "epoch": 1.7358904109589042, + "grad_norm": 2.882037878036499, + "learning_rate": 2.3700243704305446e-06, + "loss": 1.4237, + "step": 1584 + }, + { + "epoch": 1.736986301369863, + "grad_norm": 2.956204652786255, + "learning_rate": 2.367993501218522e-06, + "loss": 1.4908, + "step": 1585 + }, + { + "epoch": 1.7380821917808218, + "grad_norm": 2.9728269577026367, + "learning_rate": 2.365962632006499e-06, + "loss": 1.4557, + "step": 1586 + }, + { + "epoch": 1.7391780821917808, + "grad_norm": 2.9332668781280518, + "learning_rate": 2.3639317627944764e-06, + "loss": 1.4286, + "step": 1587 + }, + { + "epoch": 1.7402739726027399, + "grad_norm": 2.8205621242523193, + "learning_rate": 2.3619008935824537e-06, + "loss": 1.4584, + "step": 1588 + }, + { + "epoch": 1.7413698630136987, + "grad_norm": 2.9190478324890137, + "learning_rate": 2.3598700243704305e-06, + "loss": 1.4657, + "step": 1589 + }, + { + "epoch": 1.7424657534246575, + "grad_norm": 2.8216135501861572, + "learning_rate": 2.357839155158408e-06, + "loss": 1.4655, + "step": 1590 + }, + { + "epoch": 1.7424657534246575, + "eval_loss": 1.7007001638412476, + "eval_runtime": 37.3826, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.124, + "step": 1590 + }, + { + "epoch": 1.7435616438356165, + "grad_norm": 2.827335834503174, + "learning_rate": 2.355808285946385e-06, + "loss": 1.4604, + "step": 1591 + }, + { + "epoch": 1.7446575342465753, + "grad_norm": 2.808551549911499, + "learning_rate": 2.3537774167343624e-06, + "loss": 1.448, + "step": 1592 + }, + { + "epoch": 1.7457534246575341, + "grad_norm": 2.958263635635376, + "learning_rate": 2.3517465475223396e-06, + "loss": 1.4612, + "step": 1593 + }, + { + "epoch": 1.7468493150684932, + "grad_norm": 2.900951623916626, + "learning_rate": 2.349715678310317e-06, + "loss": 1.4572, + "step": 1594 + }, + { + "epoch": 1.7479452054794522, + "grad_norm": 2.933318614959717, + "learning_rate": 2.347684809098294e-06, + "loss": 1.4439, + "step": 1595 + }, + { + "epoch": 1.749041095890411, + "grad_norm": 2.929802417755127, + "learning_rate": 2.3456539398862715e-06, + "loss": 1.4331, + "step": 1596 + }, + { + "epoch": 1.7501369863013698, + "grad_norm": 3.167396068572998, + "learning_rate": 2.3436230706742487e-06, + "loss": 1.4792, + "step": 1597 + }, + { + "epoch": 1.7512328767123289, + "grad_norm": 2.901446580886841, + "learning_rate": 2.341592201462226e-06, + "loss": 1.4143, + "step": 1598 + }, + { + "epoch": 1.7523287671232877, + "grad_norm": 2.976067066192627, + "learning_rate": 2.3395613322502033e-06, + "loss": 1.41, + "step": 1599 + }, + { + "epoch": 1.7534246575342465, + "grad_norm": 2.8032453060150146, + "learning_rate": 2.3375304630381806e-06, + "loss": 1.4304, + "step": 1600 + }, + { + "epoch": 1.7534246575342465, + "eval_loss": 1.700425386428833, + "eval_runtime": 37.4655, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 1600 + }, + { + "epoch": 1.7545205479452055, + "grad_norm": 2.9107625484466553, + "learning_rate": 2.335499593826158e-06, + "loss": 1.3964, + "step": 1601 + }, + { + "epoch": 1.7556164383561645, + "grad_norm": 2.7919020652770996, + "learning_rate": 2.333468724614135e-06, + "loss": 1.4178, + "step": 1602 + }, + { + "epoch": 1.7567123287671231, + "grad_norm": 2.958300828933716, + "learning_rate": 2.3314378554021124e-06, + "loss": 1.4646, + "step": 1603 + }, + { + "epoch": 1.7578082191780822, + "grad_norm": 2.8957359790802, + "learning_rate": 2.3294069861900897e-06, + "loss": 1.4565, + "step": 1604 + }, + { + "epoch": 1.7589041095890412, + "grad_norm": 2.8454794883728027, + "learning_rate": 2.327376116978067e-06, + "loss": 1.3941, + "step": 1605 + }, + { + "epoch": 1.76, + "grad_norm": 2.8161509037017822, + "learning_rate": 2.3253452477660442e-06, + "loss": 1.4012, + "step": 1606 + }, + { + "epoch": 1.7610958904109588, + "grad_norm": 3.020136594772339, + "learning_rate": 2.3233143785540215e-06, + "loss": 1.4629, + "step": 1607 + }, + { + "epoch": 1.7621917808219179, + "grad_norm": 2.901064395904541, + "learning_rate": 2.3212835093419988e-06, + "loss": 1.413, + "step": 1608 + }, + { + "epoch": 1.7632876712328767, + "grad_norm": 3.066730499267578, + "learning_rate": 2.319252640129976e-06, + "loss": 1.4513, + "step": 1609 + }, + { + "epoch": 1.7643835616438355, + "grad_norm": 2.859468460083008, + "learning_rate": 2.3172217709179533e-06, + "loss": 1.4024, + "step": 1610 + }, + { + "epoch": 1.7643835616438355, + "eval_loss": 1.704864501953125, + "eval_runtime": 36.9687, + "eval_samples_per_second": 27.05, + "eval_steps_per_second": 1.136, + "step": 1610 + }, + { + "epoch": 1.7654794520547945, + "grad_norm": 2.990548610687256, + "learning_rate": 2.3151909017059306e-06, + "loss": 1.5025, + "step": 1611 + }, + { + "epoch": 1.7665753424657535, + "grad_norm": 2.9470932483673096, + "learning_rate": 2.313160032493908e-06, + "loss": 1.4658, + "step": 1612 + }, + { + "epoch": 1.7676712328767124, + "grad_norm": 2.8431284427642822, + "learning_rate": 2.311129163281885e-06, + "loss": 1.3847, + "step": 1613 + }, + { + "epoch": 1.7687671232876712, + "grad_norm": 2.864870309829712, + "learning_rate": 2.309098294069862e-06, + "loss": 1.451, + "step": 1614 + }, + { + "epoch": 1.7698630136986302, + "grad_norm": 2.843156099319458, + "learning_rate": 2.3070674248578393e-06, + "loss": 1.4151, + "step": 1615 + }, + { + "epoch": 1.770958904109589, + "grad_norm": 2.9214529991149902, + "learning_rate": 2.3050365556458166e-06, + "loss": 1.3913, + "step": 1616 + }, + { + "epoch": 1.7720547945205478, + "grad_norm": 2.880232572555542, + "learning_rate": 2.303005686433794e-06, + "loss": 1.4301, + "step": 1617 + }, + { + "epoch": 1.7731506849315068, + "grad_norm": 2.8919169902801514, + "learning_rate": 2.300974817221771e-06, + "loss": 1.4661, + "step": 1618 + }, + { + "epoch": 1.7742465753424659, + "grad_norm": 2.950577974319458, + "learning_rate": 2.2989439480097484e-06, + "loss": 1.4316, + "step": 1619 + }, + { + "epoch": 1.7753424657534247, + "grad_norm": 2.8465826511383057, + "learning_rate": 2.2969130787977257e-06, + "loss": 1.3856, + "step": 1620 + }, + { + "epoch": 1.7753424657534247, + "eval_loss": 1.704519510269165, + "eval_runtime": 37.0769, + "eval_samples_per_second": 26.971, + "eval_steps_per_second": 1.133, + "step": 1620 + }, + { + "epoch": 1.7764383561643835, + "grad_norm": 2.91975474357605, + "learning_rate": 2.294882209585703e-06, + "loss": 1.4478, + "step": 1621 + }, + { + "epoch": 1.7775342465753425, + "grad_norm": 3.064791440963745, + "learning_rate": 2.2928513403736802e-06, + "loss": 1.4657, + "step": 1622 + }, + { + "epoch": 1.7786301369863013, + "grad_norm": 2.9099392890930176, + "learning_rate": 2.2908204711616575e-06, + "loss": 1.4162, + "step": 1623 + }, + { + "epoch": 1.7797260273972602, + "grad_norm": 2.9778053760528564, + "learning_rate": 2.2887896019496348e-06, + "loss": 1.4423, + "step": 1624 + }, + { + "epoch": 1.7808219178082192, + "grad_norm": 3.070389747619629, + "learning_rate": 2.286758732737612e-06, + "loss": 1.4845, + "step": 1625 + }, + { + "epoch": 1.7819178082191782, + "grad_norm": 2.860940456390381, + "learning_rate": 2.284727863525589e-06, + "loss": 1.4305, + "step": 1626 + }, + { + "epoch": 1.783013698630137, + "grad_norm": 2.9022462368011475, + "learning_rate": 2.282696994313566e-06, + "loss": 1.4282, + "step": 1627 + }, + { + "epoch": 1.7841095890410958, + "grad_norm": 2.899451971054077, + "learning_rate": 2.2806661251015434e-06, + "loss": 1.4595, + "step": 1628 + }, + { + "epoch": 1.7852054794520549, + "grad_norm": 2.748985528945923, + "learning_rate": 2.2786352558895207e-06, + "loss": 1.4368, + "step": 1629 + }, + { + "epoch": 1.7863013698630137, + "grad_norm": 2.7873692512512207, + "learning_rate": 2.276604386677498e-06, + "loss": 1.4207, + "step": 1630 + }, + { + "epoch": 1.7863013698630137, + "eval_loss": 1.7050743103027344, + "eval_runtime": 37.1442, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 1630 + }, + { + "epoch": 1.7873972602739725, + "grad_norm": 2.894057512283325, + "learning_rate": 2.2745735174654753e-06, + "loss": 1.3917, + "step": 1631 + }, + { + "epoch": 1.7884931506849315, + "grad_norm": 2.931969404220581, + "learning_rate": 2.2725426482534526e-06, + "loss": 1.4056, + "step": 1632 + }, + { + "epoch": 1.7895890410958906, + "grad_norm": 3.027864694595337, + "learning_rate": 2.27051177904143e-06, + "loss": 1.4772, + "step": 1633 + }, + { + "epoch": 1.7906849315068492, + "grad_norm": 2.8892524242401123, + "learning_rate": 2.268480909829407e-06, + "loss": 1.4464, + "step": 1634 + }, + { + "epoch": 1.7917808219178082, + "grad_norm": 3.1151480674743652, + "learning_rate": 2.2664500406173844e-06, + "loss": 1.4244, + "step": 1635 + }, + { + "epoch": 1.7928767123287672, + "grad_norm": 2.991565227508545, + "learning_rate": 2.2644191714053617e-06, + "loss": 1.4498, + "step": 1636 + }, + { + "epoch": 1.793972602739726, + "grad_norm": 3.0030806064605713, + "learning_rate": 2.262388302193339e-06, + "loss": 1.4723, + "step": 1637 + }, + { + "epoch": 1.7950684931506848, + "grad_norm": 2.8650996685028076, + "learning_rate": 2.260357432981316e-06, + "loss": 1.4313, + "step": 1638 + }, + { + "epoch": 1.7961643835616439, + "grad_norm": 2.7710604667663574, + "learning_rate": 2.2583265637692935e-06, + "loss": 1.4179, + "step": 1639 + }, + { + "epoch": 1.7972602739726027, + "grad_norm": 2.8966064453125, + "learning_rate": 2.2562956945572708e-06, + "loss": 1.4394, + "step": 1640 + }, + { + "epoch": 1.7972602739726027, + "eval_loss": 1.6987948417663574, + "eval_runtime": 37.1628, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1640 + }, + { + "epoch": 1.7983561643835615, + "grad_norm": 2.816455841064453, + "learning_rate": 2.254264825345248e-06, + "loss": 1.4477, + "step": 1641 + }, + { + "epoch": 1.7994520547945205, + "grad_norm": 2.8696861267089844, + "learning_rate": 2.2522339561332253e-06, + "loss": 1.4526, + "step": 1642 + }, + { + "epoch": 1.8005479452054796, + "grad_norm": 2.9743974208831787, + "learning_rate": 2.2502030869212026e-06, + "loss": 1.4408, + "step": 1643 + }, + { + "epoch": 1.8016438356164384, + "grad_norm": 2.8802881240844727, + "learning_rate": 2.24817221770918e-06, + "loss": 1.4448, + "step": 1644 + }, + { + "epoch": 1.8027397260273972, + "grad_norm": 3.029087781906128, + "learning_rate": 2.246141348497157e-06, + "loss": 1.4702, + "step": 1645 + }, + { + "epoch": 1.8038356164383562, + "grad_norm": 2.930821180343628, + "learning_rate": 2.2441104792851344e-06, + "loss": 1.4555, + "step": 1646 + }, + { + "epoch": 1.804931506849315, + "grad_norm": 2.888127326965332, + "learning_rate": 2.2420796100731117e-06, + "loss": 1.4297, + "step": 1647 + }, + { + "epoch": 1.8060273972602738, + "grad_norm": 2.8290679454803467, + "learning_rate": 2.240048740861089e-06, + "loss": 1.4313, + "step": 1648 + }, + { + "epoch": 1.8071232876712329, + "grad_norm": 2.8667287826538086, + "learning_rate": 2.2380178716490662e-06, + "loss": 1.4663, + "step": 1649 + }, + { + "epoch": 1.808219178082192, + "grad_norm": 2.8982157707214355, + "learning_rate": 2.2359870024370435e-06, + "loss": 1.3463, + "step": 1650 + }, + { + "epoch": 1.808219178082192, + "eval_loss": 1.698309302330017, + "eval_runtime": 37.1649, + "eval_samples_per_second": 26.907, + "eval_steps_per_second": 1.13, + "step": 1650 + }, + { + "epoch": 1.8093150684931507, + "grad_norm": 2.8749091625213623, + "learning_rate": 2.2339561332250204e-06, + "loss": 1.4489, + "step": 1651 + }, + { + "epoch": 1.8104109589041095, + "grad_norm": 2.9321630001068115, + "learning_rate": 2.2319252640129976e-06, + "loss": 1.4356, + "step": 1652 + }, + { + "epoch": 1.8115068493150686, + "grad_norm": 3.039285898208618, + "learning_rate": 2.229894394800975e-06, + "loss": 1.4332, + "step": 1653 + }, + { + "epoch": 1.8126027397260274, + "grad_norm": 2.9201481342315674, + "learning_rate": 2.227863525588952e-06, + "loss": 1.3828, + "step": 1654 + }, + { + "epoch": 1.8136986301369862, + "grad_norm": 2.8183014392852783, + "learning_rate": 2.2258326563769295e-06, + "loss": 1.3748, + "step": 1655 + }, + { + "epoch": 1.8147945205479452, + "grad_norm": 2.7358760833740234, + "learning_rate": 2.2238017871649067e-06, + "loss": 1.4064, + "step": 1656 + }, + { + "epoch": 1.8158904109589042, + "grad_norm": 2.8340682983398438, + "learning_rate": 2.221770917952884e-06, + "loss": 1.3918, + "step": 1657 + }, + { + "epoch": 1.816986301369863, + "grad_norm": 2.800960063934326, + "learning_rate": 2.2197400487408613e-06, + "loss": 1.4312, + "step": 1658 + }, + { + "epoch": 1.8180821917808219, + "grad_norm": 2.986204147338867, + "learning_rate": 2.2177091795288386e-06, + "loss": 1.457, + "step": 1659 + }, + { + "epoch": 1.819178082191781, + "grad_norm": 2.8145415782928467, + "learning_rate": 2.215678310316816e-06, + "loss": 1.4067, + "step": 1660 + }, + { + "epoch": 1.819178082191781, + "eval_loss": 1.7052077054977417, + "eval_runtime": 37.16, + "eval_samples_per_second": 26.911, + "eval_steps_per_second": 1.13, + "step": 1660 + }, + { + "epoch": 1.8202739726027397, + "grad_norm": 2.936354637145996, + "learning_rate": 2.213647441104793e-06, + "loss": 1.4159, + "step": 1661 + }, + { + "epoch": 1.8213698630136985, + "grad_norm": 2.9661855697631836, + "learning_rate": 2.2116165718927704e-06, + "loss": 1.4532, + "step": 1662 + }, + { + "epoch": 1.8224657534246576, + "grad_norm": 2.943347454071045, + "learning_rate": 2.2095857026807473e-06, + "loss": 1.4179, + "step": 1663 + }, + { + "epoch": 1.8235616438356166, + "grad_norm": 3.0526676177978516, + "learning_rate": 2.2075548334687245e-06, + "loss": 1.4274, + "step": 1664 + }, + { + "epoch": 1.8246575342465754, + "grad_norm": 2.943145751953125, + "learning_rate": 2.205523964256702e-06, + "loss": 1.4012, + "step": 1665 + }, + { + "epoch": 1.8257534246575342, + "grad_norm": 2.9686639308929443, + "learning_rate": 2.203493095044679e-06, + "loss": 1.4283, + "step": 1666 + }, + { + "epoch": 1.8268493150684932, + "grad_norm": 2.951856851577759, + "learning_rate": 2.2014622258326564e-06, + "loss": 1.4493, + "step": 1667 + }, + { + "epoch": 1.827945205479452, + "grad_norm": 2.9208593368530273, + "learning_rate": 2.1994313566206336e-06, + "loss": 1.4452, + "step": 1668 + }, + { + "epoch": 1.8290410958904109, + "grad_norm": 2.9517624378204346, + "learning_rate": 2.197400487408611e-06, + "loss": 1.4529, + "step": 1669 + }, + { + "epoch": 1.83013698630137, + "grad_norm": 2.9841079711914062, + "learning_rate": 2.195369618196588e-06, + "loss": 1.431, + "step": 1670 + }, + { + "epoch": 1.83013698630137, + "eval_loss": 1.6963521242141724, + "eval_runtime": 37.1536, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 1670 + }, + { + "epoch": 1.831232876712329, + "grad_norm": 2.9343044757843018, + "learning_rate": 2.1933387489845655e-06, + "loss": 1.3912, + "step": 1671 + }, + { + "epoch": 1.8323287671232875, + "grad_norm": 2.9684195518493652, + "learning_rate": 2.1913078797725427e-06, + "loss": 1.4222, + "step": 1672 + }, + { + "epoch": 1.8334246575342465, + "grad_norm": 2.929227113723755, + "learning_rate": 2.18927701056052e-06, + "loss": 1.4535, + "step": 1673 + }, + { + "epoch": 1.8345205479452056, + "grad_norm": 3.0814716815948486, + "learning_rate": 2.1872461413484973e-06, + "loss": 1.4984, + "step": 1674 + }, + { + "epoch": 1.8356164383561644, + "grad_norm": 3.0480618476867676, + "learning_rate": 2.1852152721364746e-06, + "loss": 1.4211, + "step": 1675 + }, + { + "epoch": 1.8367123287671232, + "grad_norm": 2.9622223377227783, + "learning_rate": 2.183184402924452e-06, + "loss": 1.4252, + "step": 1676 + }, + { + "epoch": 1.8378082191780822, + "grad_norm": 3.054903745651245, + "learning_rate": 2.181153533712429e-06, + "loss": 1.4253, + "step": 1677 + }, + { + "epoch": 1.838904109589041, + "grad_norm": 2.9990928173065186, + "learning_rate": 2.1791226645004064e-06, + "loss": 1.4515, + "step": 1678 + }, + { + "epoch": 1.8399999999999999, + "grad_norm": 2.914088249206543, + "learning_rate": 2.1770917952883837e-06, + "loss": 1.438, + "step": 1679 + }, + { + "epoch": 1.841095890410959, + "grad_norm": 2.874981641769409, + "learning_rate": 2.175060926076361e-06, + "loss": 1.4247, + "step": 1680 + }, + { + "epoch": 1.841095890410959, + "eval_loss": 1.6927348375320435, + "eval_runtime": 37.1382, + "eval_samples_per_second": 26.926, + "eval_steps_per_second": 1.131, + "step": 1680 + }, + { + "epoch": 1.842191780821918, + "grad_norm": 2.954643726348877, + "learning_rate": 2.1730300568643382e-06, + "loss": 1.4892, + "step": 1681 + }, + { + "epoch": 1.8432876712328767, + "grad_norm": 2.9708850383758545, + "learning_rate": 2.1709991876523155e-06, + "loss": 1.4884, + "step": 1682 + }, + { + "epoch": 1.8443835616438355, + "grad_norm": 2.8535971641540527, + "learning_rate": 2.1689683184402928e-06, + "loss": 1.4231, + "step": 1683 + }, + { + "epoch": 1.8454794520547946, + "grad_norm": 2.9040937423706055, + "learning_rate": 2.16693744922827e-06, + "loss": 1.3577, + "step": 1684 + }, + { + "epoch": 1.8465753424657534, + "grad_norm": 2.9993464946746826, + "learning_rate": 2.1649065800162473e-06, + "loss": 1.4381, + "step": 1685 + }, + { + "epoch": 1.8476712328767122, + "grad_norm": 2.842670202255249, + "learning_rate": 2.1628757108042246e-06, + "loss": 1.4023, + "step": 1686 + }, + { + "epoch": 1.8487671232876712, + "grad_norm": 2.9607810974121094, + "learning_rate": 2.160844841592202e-06, + "loss": 1.4179, + "step": 1687 + }, + { + "epoch": 1.8498630136986303, + "grad_norm": 2.862751007080078, + "learning_rate": 2.1588139723801787e-06, + "loss": 1.4048, + "step": 1688 + }, + { + "epoch": 1.850958904109589, + "grad_norm": 2.910599708557129, + "learning_rate": 2.156783103168156e-06, + "loss": 1.4188, + "step": 1689 + }, + { + "epoch": 1.8520547945205479, + "grad_norm": 2.980849027633667, + "learning_rate": 2.1547522339561333e-06, + "loss": 1.4159, + "step": 1690 + }, + { + "epoch": 1.8520547945205479, + "eval_loss": 1.6937642097473145, + "eval_runtime": 37.3409, + "eval_samples_per_second": 26.78, + "eval_steps_per_second": 1.125, + "step": 1690 + }, + { + "epoch": 1.853150684931507, + "grad_norm": 2.9203343391418457, + "learning_rate": 2.1527213647441106e-06, + "loss": 1.3949, + "step": 1691 + }, + { + "epoch": 1.8542465753424657, + "grad_norm": 2.928570032119751, + "learning_rate": 2.150690495532088e-06, + "loss": 1.4522, + "step": 1692 + }, + { + "epoch": 1.8553424657534245, + "grad_norm": 2.9594554901123047, + "learning_rate": 2.148659626320065e-06, + "loss": 1.4629, + "step": 1693 + }, + { + "epoch": 1.8564383561643836, + "grad_norm": 2.8420305252075195, + "learning_rate": 2.1466287571080424e-06, + "loss": 1.4525, + "step": 1694 + }, + { + "epoch": 1.8575342465753426, + "grad_norm": 2.8952643871307373, + "learning_rate": 2.1445978878960197e-06, + "loss": 1.4606, + "step": 1695 + }, + { + "epoch": 1.8586301369863014, + "grad_norm": 2.8206307888031006, + "learning_rate": 2.142567018683997e-06, + "loss": 1.402, + "step": 1696 + }, + { + "epoch": 1.8597260273972602, + "grad_norm": 2.9090628623962402, + "learning_rate": 2.140536149471974e-06, + "loss": 1.4146, + "step": 1697 + }, + { + "epoch": 1.8608219178082193, + "grad_norm": 3.018481731414795, + "learning_rate": 2.1385052802599515e-06, + "loss": 1.4749, + "step": 1698 + }, + { + "epoch": 1.861917808219178, + "grad_norm": 2.882465362548828, + "learning_rate": 2.1364744110479288e-06, + "loss": 1.4091, + "step": 1699 + }, + { + "epoch": 1.8630136986301369, + "grad_norm": 2.8599698543548584, + "learning_rate": 2.134443541835906e-06, + "loss": 1.4111, + "step": 1700 + }, + { + "epoch": 1.8630136986301369, + "eval_loss": 1.6989127397537231, + "eval_runtime": 37.4015, + "eval_samples_per_second": 26.737, + "eval_steps_per_second": 1.123, + "step": 1700 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 5.299425995915264e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-1700/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/latest new file mode 100644 index 0000000000000000000000000000000000000000..adbdf3f70e1e0db88a47bff3edcb0592ed0a2e61 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/latest @@ -0,0 +1 @@ +global_step2200 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..c6bb1015279fc5ccc10507f8c47643f937086871 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/trainer_state.json @@ -0,0 +1,17193 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.410958904109589, + "eval_steps": 10, + "global_step": 2200, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6624906063079834, + "learning_rate": 4.77051177904143e-06, + "loss": 1.704, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 3.0075268745422363, + "learning_rate": 4.768480909829407e-06, + "loss": 1.6322, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.6070828437805176, + "learning_rate": 4.766450040617385e-06, + "loss": 1.6871, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 3.044222831726074, + "learning_rate": 4.764419171405362e-06, + "loss": 1.6652, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.773167610168457, + "learning_rate": 4.762388302193339e-06, + "loss": 1.7063, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.880176305770874, + "learning_rate": 4.760357432981316e-06, + "loss": 1.6775, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.6402764320373535, + "learning_rate": 4.758326563769294e-06, + "loss": 1.6694, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.9245846271514893, + "learning_rate": 4.756295694557271e-06, + "loss": 1.6688, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.623044967651367, + "learning_rate": 4.754264825345248e-06, + "loss": 1.6819, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.739896059036255, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.6401, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.8707164525985718, + "eval_runtime": 36.9513, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.603599786758423, + "learning_rate": 4.750203086921203e-06, + "loss": 1.6412, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.5571675300598145, + "learning_rate": 4.74817221770918e-06, + "loss": 1.5779, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.87091326713562, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.6878, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.8140673637390137, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7023, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.6550583839416504, + "learning_rate": 4.742079610073111e-06, + "loss": 1.5995, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.6259925365448, + "learning_rate": 4.740048740861089e-06, + "loss": 1.6194, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.971848726272583, + "learning_rate": 4.738017871649066e-06, + "loss": 1.6634, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6344778537750244, + "learning_rate": 4.735987002437044e-06, + "loss": 1.6502, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 2.9362456798553467, + "learning_rate": 4.73395613322502e-06, + "loss": 1.6948, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.8098835945129395, + "learning_rate": 4.731925264012998e-06, + "loss": 1.7161, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.8724393844604492, + "eval_runtime": 37.0876, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.933009386062622, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7014, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.8216991424560547, + "learning_rate": 4.727863525588953e-06, + "loss": 1.6192, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.7028493881225586, + "learning_rate": 4.725832656376929e-06, + "loss": 1.607, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.816818952560425, + "learning_rate": 4.723801787164907e-06, + "loss": 1.6183, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.7707865238189697, + "learning_rate": 4.721770917952885e-06, + "loss": 1.6795, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.779562473297119, + "learning_rate": 4.719740048740861e-06, + "loss": 1.63, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.907338857650757, + "learning_rate": 4.717709179528839e-06, + "loss": 1.644, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 3.073558807373047, + "learning_rate": 4.715678310316816e-06, + "loss": 1.6821, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.7556493282318115, + "learning_rate": 4.713647441104794e-06, + "loss": 1.6501, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6718344688415527, + "learning_rate": 4.71161657189277e-06, + "loss": 1.6225, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.854687213897705, + "eval_runtime": 36.9773, + "eval_samples_per_second": 27.044, + "eval_steps_per_second": 1.136, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.763335943222046, + "learning_rate": 4.709585702680748e-06, + "loss": 1.6311, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.7669479846954346, + "learning_rate": 4.707554833468725e-06, + "loss": 1.6809, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.711562395095825, + "learning_rate": 4.705523964256703e-06, + "loss": 1.741, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.799954652786255, + "learning_rate": 4.703493095044679e-06, + "loss": 1.6409, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.7278690338134766, + "learning_rate": 4.701462225832657e-06, + "loss": 1.6325, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.6802427768707275, + "learning_rate": 4.699431356620634e-06, + "loss": 1.5988, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.8826160430908203, + "learning_rate": 4.697400487408612e-06, + "loss": 1.7119, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.784343719482422, + "learning_rate": 4.695369618196588e-06, + "loss": 1.6855, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.683337688446045, + "learning_rate": 4.693338748984566e-06, + "loss": 1.6588, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.6646857261657715, + "learning_rate": 4.691307879772543e-06, + "loss": 1.623, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.8511513471603394, + "eval_runtime": 36.9852, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.8181183338165283, + "learning_rate": 4.68927701056052e-06, + "loss": 1.6171, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.8270857334136963, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.678, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.7685446739196777, + "learning_rate": 4.685215272136475e-06, + "loss": 1.6575, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.7003183364868164, + "learning_rate": 4.683184402924452e-06, + "loss": 1.6405, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.668750286102295, + "learning_rate": 4.681153533712429e-06, + "loss": 1.6541, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 3.065800905227661, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7288, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.615601062774658, + "learning_rate": 4.677091795288384e-06, + "loss": 1.6306, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.5948891639709473, + "learning_rate": 4.675060926076361e-06, + "loss": 1.6018, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.7550079822540283, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7009, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.868687868118286, + "learning_rate": 4.670999187652316e-06, + "loss": 1.6034, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.8473671674728394, + "eval_runtime": 37.0015, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.705962657928467, + "learning_rate": 4.668968318440293e-06, + "loss": 1.6084, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.7901813983917236, + "learning_rate": 4.66693744922827e-06, + "loss": 1.6171, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.688774824142456, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.6774, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.734617233276367, + "learning_rate": 4.662875710804225e-06, + "loss": 1.5796, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.742017984390259, + "learning_rate": 4.660844841592202e-06, + "loss": 1.6671, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.6438794136047363, + "learning_rate": 4.658813972380179e-06, + "loss": 1.5835, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6441051959991455, + "learning_rate": 4.656783103168157e-06, + "loss": 1.6909, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.824204206466675, + "learning_rate": 4.654752233956134e-06, + "loss": 1.5758, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.556638479232788, + "learning_rate": 4.652721364744111e-06, + "loss": 1.6504, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.897812843322754, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.6406, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.8537043333053589, + "eval_runtime": 37.257, + "eval_samples_per_second": 26.841, + "eval_steps_per_second": 1.127, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.6594088077545166, + "learning_rate": 4.648659626320066e-06, + "loss": 1.6966, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.79524302482605, + "learning_rate": 4.646628757108043e-06, + "loss": 1.6645, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.80291748046875, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.6281, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.7723848819732666, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.5959, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 3.0625650882720947, + "learning_rate": 4.640536149471974e-06, + "loss": 1.6313, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.775726079940796, + "learning_rate": 4.638505280259952e-06, + "loss": 1.6563, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.837758779525757, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.6727, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 3.041355609893799, + "learning_rate": 4.634443541835907e-06, + "loss": 1.6282, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.5775389671325684, + "learning_rate": 4.632412672623883e-06, + "loss": 1.651, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 3.026287794113159, + "learning_rate": 4.630381803411861e-06, + "loss": 1.6458, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.8461565971374512, + "eval_runtime": 37.3881, + "eval_samples_per_second": 26.747, + "eval_steps_per_second": 1.123, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.6781764030456543, + "learning_rate": 4.628350934199838e-06, + "loss": 1.5232, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.763350486755371, + "learning_rate": 4.626320064987816e-06, + "loss": 1.6186, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.7063372135162354, + "learning_rate": 4.624289195775792e-06, + "loss": 1.6784, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.824563980102539, + "learning_rate": 4.62225832656377e-06, + "loss": 1.6219, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.5335569381713867, + "learning_rate": 4.620227457351747e-06, + "loss": 1.6152, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.773644208908081, + "learning_rate": 4.618196588139724e-06, + "loss": 1.6179, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.9841721057891846, + "learning_rate": 4.616165718927701e-06, + "loss": 1.6258, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.684171676635742, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.6066, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.8817965984344482, + "learning_rate": 4.612103980503656e-06, + "loss": 1.6697, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.9846866130828857, + "learning_rate": 4.610073111291633e-06, + "loss": 1.6541, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.8485002517700195, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.904170513153076, + "learning_rate": 4.60804224207961e-06, + "loss": 1.6362, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.684476375579834, + "learning_rate": 4.606011372867588e-06, + "loss": 1.6303, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6692216396331787, + "learning_rate": 4.603980503655565e-06, + "loss": 1.638, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.73555588722229, + "learning_rate": 4.601949634443542e-06, + "loss": 1.6125, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.729525089263916, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.565, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.5972018241882324, + "learning_rate": 4.597887896019497e-06, + "loss": 1.5856, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7859137058258057, + "learning_rate": 4.595857026807474e-06, + "loss": 1.6325, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.662587881088257, + "learning_rate": 4.593826157595451e-06, + "loss": 1.6172, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.6380577087402344, + "learning_rate": 4.591795288383429e-06, + "loss": 1.6005, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8375349044799805, + "learning_rate": 4.589764419171406e-06, + "loss": 1.6507, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.833074688911438, + "eval_runtime": 37.0695, + "eval_samples_per_second": 26.976, + "eval_steps_per_second": 1.133, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.786713123321533, + "learning_rate": 4.587733549959383e-06, + "loss": 1.6283, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.6959030628204346, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.5994, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.6100692749023438, + "learning_rate": 4.583671811535338e-06, + "loss": 1.5884, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.7281863689422607, + "learning_rate": 4.581640942323315e-06, + "loss": 1.5737, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.7519376277923584, + "learning_rate": 4.579610073111292e-06, + "loss": 1.6531, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.9314866065979004, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.6755, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.9250576496124268, + "learning_rate": 4.575548334687247e-06, + "loss": 1.6604, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.6382815837860107, + "learning_rate": 4.573517465475224e-06, + "loss": 1.6773, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.6891751289367676, + "learning_rate": 4.571486596263201e-06, + "loss": 1.5849, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.7018136978149414, + "learning_rate": 4.569455727051178e-06, + "loss": 1.6358, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.838019847869873, + "eval_runtime": 37.0496, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.671952247619629, + "learning_rate": 4.567424857839156e-06, + "loss": 1.6207, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.882539749145508, + "learning_rate": 4.565393988627132e-06, + "loss": 1.6618, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.684070587158203, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.6502, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.5295751094818115, + "learning_rate": 4.561332250203087e-06, + "loss": 1.6078, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.0268194675445557, + "learning_rate": 4.559301380991065e-06, + "loss": 1.6312, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6895201206207275, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.5848, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.869950771331787, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.5951, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.909505605697632, + "learning_rate": 4.553208773354996e-06, + "loss": 1.6589, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.7604334354400635, + "learning_rate": 4.551177904142974e-06, + "loss": 1.5767, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.8792080879211426, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.6864, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.836847186088562, + "eval_runtime": 36.9142, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.607037305831909, + "learning_rate": 4.547116165718929e-06, + "loss": 1.5925, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.9338927268981934, + "learning_rate": 4.545085296506905e-06, + "loss": 1.6433, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.7723233699798584, + "learning_rate": 4.543054427294882e-06, + "loss": 1.5963, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.4122188091278076, + "learning_rate": 4.54102355808286e-06, + "loss": 1.5978, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.7582709789276123, + "learning_rate": 4.538992688870837e-06, + "loss": 1.581, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.6983039379119873, + "learning_rate": 4.536961819658814e-06, + "loss": 1.6253, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.6036365032196045, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.5957, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.759645700454712, + "learning_rate": 4.532900081234769e-06, + "loss": 1.6199, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.797744035720825, + "learning_rate": 4.530869212022746e-06, + "loss": 1.5854, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.752868890762329, + "learning_rate": 4.528838342810723e-06, + "loss": 1.619, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.8305871486663818, + "eval_runtime": 37.0307, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.7421889305114746, + "learning_rate": 4.526807473598701e-06, + "loss": 1.6182, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.6852707862854004, + "learning_rate": 4.524776604386678e-06, + "loss": 1.6135, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.861720323562622, + "learning_rate": 4.522745735174655e-06, + "loss": 1.6099, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.706851005554199, + "learning_rate": 4.520714865962632e-06, + "loss": 1.6597, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7992191314697266, + "learning_rate": 4.51868399675061e-06, + "loss": 1.6077, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.6248183250427246, + "learning_rate": 4.516653127538587e-06, + "loss": 1.6485, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.696040391921997, + "learning_rate": 4.514622258326564e-06, + "loss": 1.5853, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.5958261489868164, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.571, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.9715874195098877, + "learning_rate": 4.510560519902519e-06, + "loss": 1.6044, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.664834499359131, + "learning_rate": 4.508529650690496e-06, + "loss": 1.6568, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.8277279138565063, + "eval_runtime": 37.0489, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.869558811187744, + "learning_rate": 4.506498781478473e-06, + "loss": 1.5654, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.5814828872680664, + "learning_rate": 4.504467912266451e-06, + "loss": 1.5484, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.6158947944641113, + "learning_rate": 4.502437043054428e-06, + "loss": 1.6202, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.7424654960632324, + "learning_rate": 4.500406173842405e-06, + "loss": 1.6755, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.8670949935913086, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.6144, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.7914652824401855, + "learning_rate": 4.49634443541836e-06, + "loss": 1.6164, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.866380214691162, + "learning_rate": 4.494313566206336e-06, + "loss": 1.5961, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.966474771499634, + "learning_rate": 4.492282696994314e-06, + "loss": 1.6681, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.751291513442993, + "learning_rate": 4.490251827782291e-06, + "loss": 1.6338, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.58840012550354, + "learning_rate": 4.488220958570269e-06, + "loss": 1.5619, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.824889898300171, + "eval_runtime": 37.0821, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.786945104598999, + "learning_rate": 4.486190089358245e-06, + "loss": 1.5712, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4536845684051514, + "learning_rate": 4.484159220146223e-06, + "loss": 1.5487, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.780278205871582, + "learning_rate": 4.4821283509342e-06, + "loss": 1.6269, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.5427017211914062, + "learning_rate": 4.480097481722178e-06, + "loss": 1.6254, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.7540087699890137, + "learning_rate": 4.478066612510154e-06, + "loss": 1.6112, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.616352081298828, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.5637, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.686035633087158, + "learning_rate": 4.474004874086109e-06, + "loss": 1.5974, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.982365846633911, + "learning_rate": 4.471974004874087e-06, + "loss": 1.7034, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.6106085777282715, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.5482, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.7798519134521484, + "learning_rate": 4.467912266450041e-06, + "loss": 1.6475, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.820489764213562, + "eval_runtime": 37.0672, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.8634185791015625, + "learning_rate": 4.465881397238018e-06, + "loss": 1.6258, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.633626699447632, + "learning_rate": 4.463850528025995e-06, + "loss": 1.5806, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.719074249267578, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.6404, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.7545487880706787, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6404, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.6559669971466064, + "learning_rate": 4.457757920389927e-06, + "loss": 1.6104, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.9004898071289062, + "learning_rate": 4.455727051177904e-06, + "loss": 1.5513, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.6640021800994873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.5859, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.842026472091675, + "learning_rate": 4.451665312753859e-06, + "loss": 1.6241, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.784647226333618, + "learning_rate": 4.449634443541836e-06, + "loss": 1.627, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.7067618370056152, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.6276, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.8115602731704712, + "eval_runtime": 37.078, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.6582882404327393, + "learning_rate": 4.445572705117791e-06, + "loss": 1.5746, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7881102561950684, + "learning_rate": 4.443541835905768e-06, + "loss": 1.5749, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.9323439598083496, + "learning_rate": 4.441510966693745e-06, + "loss": 1.6421, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.63948655128479, + "learning_rate": 4.439480097481723e-06, + "loss": 1.5853, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.7746710777282715, + "learning_rate": 4.4374492282697e-06, + "loss": 1.6455, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.093895435333252, + "learning_rate": 4.435418359057677e-06, + "loss": 1.5311, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.526385545730591, + "learning_rate": 4.433387489845654e-06, + "loss": 1.6082, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.6604297161102295, + "learning_rate": 4.431356620633632e-06, + "loss": 1.61, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.6636247634887695, + "learning_rate": 4.429325751421609e-06, + "loss": 1.5815, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.5567514896392822, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6086, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.8170957565307617, + "eval_runtime": 37.4095, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.695563793182373, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.5836, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6009771823883057, + "learning_rate": 4.423233143785541e-06, + "loss": 1.5264, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.0490736961364746, + "learning_rate": 4.421202274573518e-06, + "loss": 1.5982, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.8609886169433594, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6265, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4415369033813477, + "learning_rate": 4.417140536149473e-06, + "loss": 1.4987, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.771657943725586, + "learning_rate": 4.415109666937449e-06, + "loss": 1.5494, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.739617347717285, + "learning_rate": 4.413078797725427e-06, + "loss": 1.6274, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.805098295211792, + "learning_rate": 4.411047928513404e-06, + "loss": 1.5686, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.7248735427856445, + "learning_rate": 4.409017059301382e-06, + "loss": 1.6027, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6653318405151367, + "learning_rate": 4.406986190089358e-06, + "loss": 1.5909, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.8076283931732178, + "eval_runtime": 37.5073, + "eval_samples_per_second": 26.661, + "eval_steps_per_second": 1.12, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.626783847808838, + "learning_rate": 4.404955320877336e-06, + "loss": 1.6205, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.6929268836975098, + "learning_rate": 4.402924451665313e-06, + "loss": 1.6146, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.6806800365448, + "learning_rate": 4.400893582453291e-06, + "loss": 1.5702, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.5407259464263916, + "learning_rate": 4.398862713241267e-06, + "loss": 1.5525, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.639122247695923, + "learning_rate": 4.396831844029245e-06, + "loss": 1.5826, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.669396162033081, + "learning_rate": 4.394800974817222e-06, + "loss": 1.5982, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.6377906799316406, + "learning_rate": 4.392770105605199e-06, + "loss": 1.5797, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6769638061523438, + "learning_rate": 4.390739236393176e-06, + "loss": 1.607, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.6347222328186035, + "learning_rate": 4.388708367181154e-06, + "loss": 1.6196, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.670680284500122, + "learning_rate": 4.386677497969131e-06, + "loss": 1.619, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.8124384880065918, + "eval_runtime": 37.4751, + "eval_samples_per_second": 26.684, + "eval_steps_per_second": 1.121, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.5886521339416504, + "learning_rate": 4.384646628757108e-06, + "loss": 1.6258, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.644566297531128, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6447, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.6432394981384277, + "learning_rate": 4.380584890333063e-06, + "loss": 1.5888, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.5602784156799316, + "learning_rate": 4.37855402112104e-06, + "loss": 1.5544, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.4874439239501953, + "learning_rate": 4.376523151909017e-06, + "loss": 1.5075, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.694748640060425, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.6498, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.582211494445801, + "learning_rate": 4.372461413484972e-06, + "loss": 1.5877, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.675713539123535, + "learning_rate": 4.370430544272949e-06, + "loss": 1.6045, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.7157161235809326, + "learning_rate": 4.368399675060926e-06, + "loss": 1.623, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.795076370239258, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6141, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.8062744140625, + "eval_runtime": 37.2435, + "eval_samples_per_second": 26.85, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6296374797821045, + "learning_rate": 4.364337936636881e-06, + "loss": 1.6087, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.362307067424858e-06, + "loss": 1.5182, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.7965915203094482, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.6025, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.807307481765747, + "learning_rate": 4.358245329000813e-06, + "loss": 1.574, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.584615468978882, + "learning_rate": 4.35621445978879e-06, + "loss": 1.5922, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.8472282886505127, + "learning_rate": 4.354183590576767e-06, + "loss": 1.5914, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.8471736907958984, + "learning_rate": 4.352152721364745e-06, + "loss": 1.5749, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.569225549697876, + "learning_rate": 4.350121852152722e-06, + "loss": 1.5591, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.688157796859741, + "learning_rate": 4.348090982940699e-06, + "loss": 1.6042, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.5865259170532227, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.6307, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.8040707111358643, + "eval_runtime": 36.9511, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.71850323677063, + "learning_rate": 4.344029244516654e-06, + "loss": 1.6043, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.7158093452453613, + "learning_rate": 4.341998375304631e-06, + "loss": 1.5907, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6879327297210693, + "learning_rate": 4.339967506092607e-06, + "loss": 1.6365, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.827277660369873, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.5928, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6709842681884766, + "learning_rate": 4.335905767668562e-06, + "loss": 1.5812, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.7071495056152344, + "learning_rate": 4.33387489845654e-06, + "loss": 1.5571, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6008732318878174, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.6048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.6238059997558594, + "learning_rate": 4.329813160032495e-06, + "loss": 1.5962, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.567065954208374, + "learning_rate": 4.327782290820471e-06, + "loss": 1.5713, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.573068857192993, + "learning_rate": 4.325751421608449e-06, + "loss": 1.536, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.8103471994400024, + "eval_runtime": 37.0956, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6245224475860596, + "learning_rate": 4.323720552396426e-06, + "loss": 1.5986, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.6243810653686523, + "learning_rate": 4.321689683184404e-06, + "loss": 1.589, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.5303313732147217, + "learning_rate": 4.31965881397238e-06, + "loss": 1.5793, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.5617806911468506, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.5704, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.7429656982421875, + "learning_rate": 4.315597075548335e-06, + "loss": 1.6174, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.6753251552581787, + "learning_rate": 4.313566206336312e-06, + "loss": 1.5879, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.5944488048553467, + "learning_rate": 4.311535337124289e-06, + "loss": 1.6082, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.9002904891967773, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.5521, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.5384247303009033, + "learning_rate": 4.307473598700244e-06, + "loss": 1.5751, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6015477180480957, + "learning_rate": 4.305442729488221e-06, + "loss": 1.5966, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8091154098510742, + "eval_runtime": 37.1075, + "eval_samples_per_second": 26.949, + "eval_steps_per_second": 1.132, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.6002886295318604, + "learning_rate": 4.303411860276198e-06, + "loss": 1.5403, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7755627632141113, + "learning_rate": 4.301380991064176e-06, + "loss": 1.6192, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.8351540565490723, + "learning_rate": 4.299350121852153e-06, + "loss": 1.6217, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.684007406234741, + "learning_rate": 4.29731925264013e-06, + "loss": 1.6242, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.8019614219665527, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6173, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.9122867584228516, + "learning_rate": 4.293257514216085e-06, + "loss": 1.5701, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.545732021331787, + "learning_rate": 4.291226645004062e-06, + "loss": 1.5494, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.9499566555023193, + "learning_rate": 4.289195775792039e-06, + "loss": 1.5848, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.8071892261505127, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6114, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.763864517211914, + "learning_rate": 4.285134037367994e-06, + "loss": 1.5768, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8057912588119507, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 3.001190423965454, + "learning_rate": 4.283103168155971e-06, + "loss": 1.5601, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.7640297412872314, + "learning_rate": 4.281072298943948e-06, + "loss": 1.5619, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.7372794151306152, + "learning_rate": 4.279041429731926e-06, + "loss": 1.562, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 3.0390751361846924, + "learning_rate": 4.277010560519903e-06, + "loss": 1.5519, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.758798599243164, + "learning_rate": 4.27497969130788e-06, + "loss": 1.527, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.96443772315979, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.5899, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.7941088676452637, + "learning_rate": 4.270917952883835e-06, + "loss": 1.617, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 3.0385100841522217, + "learning_rate": 4.268887083671812e-06, + "loss": 1.5996, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.6946637630462646, + "learning_rate": 4.266856214459789e-06, + "loss": 1.5641, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.776723861694336, + "learning_rate": 4.264825345247766e-06, + "loss": 1.6451, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.7911298274993896, + "eval_runtime": 37.0623, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.851224184036255, + "learning_rate": 4.262794476035744e-06, + "loss": 1.5302, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9045867919921875, + "learning_rate": 4.26076360682372e-06, + "loss": 1.5808, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.6568782329559326, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.5233, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.94191837310791, + "learning_rate": 4.256701868399675e-06, + "loss": 1.5912, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.726778984069824, + "learning_rate": 4.254670999187653e-06, + "loss": 1.5823, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.8704583644866943, + "learning_rate": 4.252640129975629e-06, + "loss": 1.5931, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.5721547603607178, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6128, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6883509159088135, + "learning_rate": 4.248578391551584e-06, + "loss": 1.5479, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.7186272144317627, + "learning_rate": 4.246547522339562e-06, + "loss": 1.5203, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.8326518535614014, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.5547, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.7948570251464844, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.7870523929595947, + "learning_rate": 4.242485783915517e-06, + "loss": 1.5891, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.866610288619995, + "learning_rate": 4.240454914703493e-06, + "loss": 1.5232, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.94161057472229, + "learning_rate": 4.23842404549147e-06, + "loss": 1.52, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.7331979274749756, + "learning_rate": 4.236393176279448e-06, + "loss": 1.5665, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.5375845432281494, + "learning_rate": 4.234362307067425e-06, + "loss": 1.5274, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.691927433013916, + "learning_rate": 4.232331437855402e-06, + "loss": 1.604, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7578394412994385, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.6, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.594538688659668, + "learning_rate": 4.228269699431357e-06, + "loss": 1.5928, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.5408968925476074, + "learning_rate": 4.226238830219334e-06, + "loss": 1.5908, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.689786672592163, + "learning_rate": 4.224207961007311e-06, + "loss": 1.5757, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.792494297027588, + "eval_runtime": 37.0822, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.735539197921753, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.5811, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.58634614944458, + "learning_rate": 4.220146222583266e-06, + "loss": 1.5808, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.7899138927459717, + "learning_rate": 4.218115353371243e-06, + "loss": 1.5801, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.6507744789123535, + "learning_rate": 4.21608448415922e-06, + "loss": 1.5585, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.6112217903137207, + "learning_rate": 4.214053614947198e-06, + "loss": 1.5329, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.729405164718628, + "learning_rate": 4.212022745735175e-06, + "loss": 1.6177, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.6431992053985596, + "learning_rate": 4.209991876523152e-06, + "loss": 1.5234, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.726243734359741, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.5926, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.6574180126190186, + "learning_rate": 4.205930138099107e-06, + "loss": 1.5424, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.6570956707000732, + "learning_rate": 4.203899268887084e-06, + "loss": 1.6119, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.781999945640564, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.528052806854248, + "learning_rate": 4.201868399675061e-06, + "loss": 1.598, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.6075727939605713, + "learning_rate": 4.199837530463039e-06, + "loss": 1.5886, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.588737964630127, + "learning_rate": 4.197806661251016e-06, + "loss": 1.507, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.773618221282959, + "learning_rate": 4.195775792038993e-06, + "loss": 1.5794, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.624725580215454, + "learning_rate": 4.19374492282697e-06, + "loss": 1.5495, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.662397861480713, + "learning_rate": 4.191714053614948e-06, + "loss": 1.5203, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.6617255210876465, + "learning_rate": 4.189683184402924e-06, + "loss": 1.5885, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.5784451961517334, + "learning_rate": 4.187652315190902e-06, + "loss": 1.5689, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.726011037826538, + "learning_rate": 4.185621445978879e-06, + "loss": 1.5913, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.6117324829101562, + "learning_rate": 4.183590576766857e-06, + "loss": 1.5453, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.7893352508544922, + "eval_runtime": 37.4096, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7025442123413086, + "learning_rate": 4.181559707554833e-06, + "loss": 1.5868, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.656102180480957, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6141, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.741962194442749, + "learning_rate": 4.177497969130788e-06, + "loss": 1.5516, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5887229442596436, + "learning_rate": 4.175467099918766e-06, + "loss": 1.5525, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.542985439300537, + "learning_rate": 4.173436230706742e-06, + "loss": 1.592, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.7797129154205322, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.5349, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.632887363433838, + "learning_rate": 4.169374492282697e-06, + "loss": 1.5548, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.691690444946289, + "learning_rate": 4.167343623070675e-06, + "loss": 1.5352, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.8922817707061768, + "learning_rate": 4.165312753858652e-06, + "loss": 1.5717, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.5517446994781494, + "learning_rate": 4.163281884646629e-06, + "loss": 1.5658, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.780922770500183, + "eval_runtime": 37.3755, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.718374490737915, + "learning_rate": 4.161251015434607e-06, + "loss": 1.5578, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.6249442100524902, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6018, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.7051141262054443, + "learning_rate": 4.157189277010561e-06, + "loss": 1.5182, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.677109956741333, + "learning_rate": 4.155158407798538e-06, + "loss": 1.5423, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.591059684753418, + "learning_rate": 4.153127538586516e-06, + "loss": 1.5694, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.699953556060791, + "learning_rate": 4.151096669374492e-06, + "loss": 1.5758, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.8837342262268066, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.5509, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.6566851139068604, + "learning_rate": 4.147034930950447e-06, + "loss": 1.5577, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.901057004928589, + "learning_rate": 4.145004061738425e-06, + "loss": 1.5797, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.796971321105957, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.5608, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.7797813415527344, + "eval_runtime": 36.8425, + "eval_samples_per_second": 27.143, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.605431318283081, + "learning_rate": 4.140942323314379e-06, + "loss": 1.491, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8084020614624023, + "learning_rate": 4.138911454102356e-06, + "loss": 1.5074, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.625511407852173, + "learning_rate": 4.136880584890333e-06, + "loss": 1.5404, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.6944286823272705, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.589, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.81792950630188, + "learning_rate": 4.132818846466288e-06, + "loss": 1.5646, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5813655853271484, + "learning_rate": 4.130787977254265e-06, + "loss": 1.5697, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.6733286380767822, + "learning_rate": 4.128757108042242e-06, + "loss": 1.5786, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.6972577571868896, + "learning_rate": 4.12672623883022e-06, + "loss": 1.5857, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7381093502044678, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6365, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.6162049770355225, + "learning_rate": 4.122664500406174e-06, + "loss": 1.5598, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.7869179248809814, + "eval_runtime": 36.974, + "eval_samples_per_second": 27.046, + "eval_steps_per_second": 1.136, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.7157270908355713, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.5533, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.560105085372925, + "learning_rate": 4.118602761982129e-06, + "loss": 1.5821, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.5261518955230713, + "learning_rate": 4.116571892770106e-06, + "loss": 1.5195, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7377054691314697, + "learning_rate": 4.114541023558083e-06, + "loss": 1.5066, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.632946252822876, + "learning_rate": 4.112510154346061e-06, + "loss": 1.5948, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.809671640396118, + "learning_rate": 4.110479285134038e-06, + "loss": 1.5435, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.843156099319458, + "learning_rate": 4.108448415922015e-06, + "loss": 1.5667, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.743614673614502, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5958, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.9596827030181885, + "learning_rate": 4.10438667749797e-06, + "loss": 1.5807, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.7321834564208984, + "learning_rate": 4.102355808285947e-06, + "loss": 1.4744, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.784859538078308, + "eval_runtime": 37.0146, + "eval_samples_per_second": 27.016, + "eval_steps_per_second": 1.135, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.010359764099121, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6146, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.744251251220703, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.5401, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.592690944671631, + "learning_rate": 4.096263200649879e-06, + "loss": 1.5508, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.786736011505127, + "learning_rate": 4.094232331437856e-06, + "loss": 1.5717, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.8773610591888428, + "learning_rate": 4.092201462225833e-06, + "loss": 1.5201, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.818610906600952, + "learning_rate": 4.090170593013811e-06, + "loss": 1.5282, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.9548542499542236, + "learning_rate": 4.088139723801787e-06, + "loss": 1.553, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.678842306137085, + "learning_rate": 4.086108854589765e-06, + "loss": 1.531, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.7757408618927, + "learning_rate": 4.084077985377742e-06, + "loss": 1.5371, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.6954638957977295, + "learning_rate": 4.08204711616572e-06, + "loss": 1.5462, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.781003475189209, + "eval_runtime": 37.0687, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6897518634796143, + "learning_rate": 4.080016246953696e-06, + "loss": 1.5979, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.581172466278076, + "learning_rate": 4.077985377741674e-06, + "loss": 1.5007, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.828234910964966, + "learning_rate": 4.075954508529651e-06, + "loss": 1.5857, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.6171951293945312, + "learning_rate": 4.073923639317629e-06, + "loss": 1.5937, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.6476316452026367, + "learning_rate": 4.071892770105605e-06, + "loss": 1.5864, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.7525634765625, + "learning_rate": 4.069861900893583e-06, + "loss": 1.5095, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.6881322860717773, + "learning_rate": 4.06783103168156e-06, + "loss": 1.4936, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.6104350090026855, + "learning_rate": 4.065800162469538e-06, + "loss": 1.5229, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.866706132888794, + "learning_rate": 4.063769293257514e-06, + "loss": 1.539, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.803999900817871, + "learning_rate": 4.061738424045492e-06, + "loss": 1.5792, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.7793465852737427, + "eval_runtime": 37.0749, + "eval_samples_per_second": 26.972, + "eval_steps_per_second": 1.133, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.670285940170288, + "learning_rate": 4.059707554833469e-06, + "loss": 1.5523, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.871267557144165, + "learning_rate": 4.057676685621446e-06, + "loss": 1.5623, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.6631205081939697, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6094, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.7580034732818604, + "learning_rate": 4.053614947197401e-06, + "loss": 1.5725, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.6412739753723145, + "learning_rate": 4.051584077985378e-06, + "loss": 1.4959, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.699176073074341, + "learning_rate": 4.049553208773355e-06, + "loss": 1.5586, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.622969150543213, + "learning_rate": 4.047522339561333e-06, + "loss": 1.5713, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.892338991165161, + "learning_rate": 4.04549147034931e-06, + "loss": 1.5471, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.77056622505188, + "learning_rate": 4.043460601137287e-06, + "loss": 1.5236, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.7122550010681152, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6562, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.7708193063735962, + "eval_runtime": 37.0743, + "eval_samples_per_second": 26.973, + "eval_steps_per_second": 1.133, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.702824592590332, + "learning_rate": 4.039398862713242e-06, + "loss": 1.534, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.6822307109832764, + "learning_rate": 4.037367993501219e-06, + "loss": 1.5368, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.7274515628814697, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6218, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.8114798069000244, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.532, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.8936192989349365, + "learning_rate": 4.031275385865151e-06, + "loss": 1.5124, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.7851672172546387, + "learning_rate": 4.029244516653128e-06, + "loss": 1.5596, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.815023183822632, + "learning_rate": 4.027213647441105e-06, + "loss": 1.5706, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5958921909332275, + "learning_rate": 4.025182778229083e-06, + "loss": 1.5161, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.779360771179199, + "learning_rate": 4.02315190901706e-06, + "loss": 1.5057, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.7411787509918213, + "learning_rate": 4.021121039805037e-06, + "loss": 1.5396, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.7778708934783936, + "eval_runtime": 37.0791, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.5709781646728516, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.5474, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8583388328552246, + "learning_rate": 4.017059301380992e-06, + "loss": 1.6149, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.6877572536468506, + "learning_rate": 4.015028432168969e-06, + "loss": 1.5292, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.680612087249756, + "learning_rate": 4.012997562956945e-06, + "loss": 1.596, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.653559446334839, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.4901, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.542787790298462, + "learning_rate": 4.0089358245329e-06, + "loss": 1.5295, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5476601123809814, + "learning_rate": 4.006904955320878e-06, + "loss": 1.5393, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.6367862224578857, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.5321, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.8112173080444336, + "learning_rate": 4.002843216896833e-06, + "loss": 1.5963, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.7136871814727783, + "learning_rate": 4.000812347684809e-06, + "loss": 1.5839, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.7763899564743042, + "eval_runtime": 37.2537, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.9095280170440674, + "learning_rate": 3.998781478472787e-06, + "loss": 1.5343, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.6196115016937256, + "learning_rate": 3.996750609260764e-06, + "loss": 1.5411, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6413371562957764, + "learning_rate": 3.994719740048742e-06, + "loss": 1.4862, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 3.0759036540985107, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6471, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.599147081375122, + "learning_rate": 3.990658001624696e-06, + "loss": 1.5556, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.726978302001953, + "learning_rate": 3.988627132412673e-06, + "loss": 1.5351, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.6760807037353516, + "learning_rate": 3.98659626320065e-06, + "loss": 1.5681, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.680483818054199, + "learning_rate": 3.984565393988627e-06, + "loss": 1.5225, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6649010181427, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.5325, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.638085126876831, + "learning_rate": 3.980503655564582e-06, + "loss": 1.564, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.7698904275894165, + "eval_runtime": 37.4492, + "eval_samples_per_second": 26.703, + "eval_steps_per_second": 1.122, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.647825241088867, + "learning_rate": 3.978472786352559e-06, + "loss": 1.5782, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.748222589492798, + "learning_rate": 3.976441917140536e-06, + "loss": 1.556, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.6319026947021484, + "learning_rate": 3.974411047928514e-06, + "loss": 1.5378, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.6759872436523438, + "learning_rate": 3.972380178716491e-06, + "loss": 1.6018, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.6760940551757812, + "learning_rate": 3.970349309504468e-06, + "loss": 1.5742, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.5565083026885986, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.565, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.5986931324005127, + "learning_rate": 3.966287571080423e-06, + "loss": 1.5124, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.713092803955078, + "learning_rate": 3.9642567018684e-06, + "loss": 1.523, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.6203417778015137, + "learning_rate": 3.962225832656377e-06, + "loss": 1.5601, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.604440927505493, + "learning_rate": 3.960194963444355e-06, + "loss": 1.5485, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.7633332014083862, + "eval_runtime": 37.49, + "eval_samples_per_second": 26.674, + "eval_steps_per_second": 1.12, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.5513081550598145, + "learning_rate": 3.958164094232332e-06, + "loss": 1.479, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.7990963459014893, + "learning_rate": 3.956133225020309e-06, + "loss": 1.5355, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.651113271713257, + "learning_rate": 3.954102355808286e-06, + "loss": 1.5406, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.7236671447753906, + "learning_rate": 3.952071486596264e-06, + "loss": 1.5314, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.697598695755005, + "learning_rate": 3.950040617384241e-06, + "loss": 1.5398, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.8271944522857666, + "learning_rate": 3.948009748172218e-06, + "loss": 1.54, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.733125925064087, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.5236, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.7555017471313477, + "learning_rate": 3.943948009748173e-06, + "loss": 1.5808, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.6592960357666016, + "learning_rate": 3.94191714053615e-06, + "loss": 1.5398, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.6870126724243164, + "learning_rate": 3.939886271324127e-06, + "loss": 1.5517, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.760675311088562, + "eval_runtime": 36.9372, + "eval_samples_per_second": 27.073, + "eval_steps_per_second": 1.137, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.7469022274017334, + "learning_rate": 3.937855402112104e-06, + "loss": 1.605, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.5255792140960693, + "learning_rate": 3.935824532900082e-06, + "loss": 1.5061, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.589054822921753, + "learning_rate": 3.933793663688058e-06, + "loss": 1.4971, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.7874889373779297, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.5643, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.819451093673706, + "learning_rate": 3.929731925264013e-06, + "loss": 1.5797, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.6516339778900146, + "learning_rate": 3.927701056051991e-06, + "loss": 1.5048, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.979091167449951, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.5174, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.743058204650879, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.5432, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.6864030361175537, + "learning_rate": 3.921608448415922e-06, + "loss": 1.5147, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.7751128673553467, + "learning_rate": 3.9195775792039e-06, + "loss": 1.4857, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.7609487771987915, + "eval_runtime": 37.09, + "eval_samples_per_second": 26.961, + "eval_steps_per_second": 1.132, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.626188278198242, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.5428, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.554067850112915, + "learning_rate": 3.915515840779855e-06, + "loss": 1.5153, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.8580875396728516, + "learning_rate": 3.913484971567831e-06, + "loss": 1.5562, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.7405097484588623, + "learning_rate": 3.911454102355808e-06, + "loss": 1.565, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.5650815963745117, + "learning_rate": 3.909423233143786e-06, + "loss": 1.4756, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.9405910968780518, + "learning_rate": 3.907392363931763e-06, + "loss": 1.5611, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.687448024749756, + "learning_rate": 3.90536149471974e-06, + "loss": 1.5405, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.7373857498168945, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.6055, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.875298261642456, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6215, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.713200330734253, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6493, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.7608598470687866, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.541935682296753, + "learning_rate": 3.897238017871649e-06, + "loss": 1.5257, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.582045316696167, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.5253, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.6884682178497314, + "learning_rate": 3.893176279447604e-06, + "loss": 1.5329, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.6456501483917236, + "learning_rate": 3.891145410235581e-06, + "loss": 1.5167, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.60788893699646, + "learning_rate": 3.889114541023558e-06, + "loss": 1.5945, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.708646774291992, + "learning_rate": 3.887083671811536e-06, + "loss": 1.5704, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.666922092437744, + "learning_rate": 3.885052802599513e-06, + "loss": 1.52, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.6623599529266357, + "learning_rate": 3.88302193338749e-06, + "loss": 1.4791, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.9031424522399902, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.5954, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.763303279876709, + "learning_rate": 3.878960194963445e-06, + "loss": 1.5618, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.7647819519042969, + "eval_runtime": 37.1392, + "eval_samples_per_second": 26.926, + "eval_steps_per_second": 1.131, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 3.0733320713043213, + "learning_rate": 3.876929325751422e-06, + "loss": 1.5191, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.7809090614318848, + "learning_rate": 3.874898456539399e-06, + "loss": 1.5804, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.6357381343841553, + "learning_rate": 3.872867587327377e-06, + "loss": 1.4843, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.680595874786377, + "learning_rate": 3.870836718115354e-06, + "loss": 1.5339, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.5908069610595703, + "learning_rate": 3.868805848903331e-06, + "loss": 1.5049, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.680586338043213, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.5188, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.763528823852539, + "learning_rate": 3.864744110479286e-06, + "loss": 1.5395, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.8522050380706787, + "learning_rate": 3.862713241267262e-06, + "loss": 1.562, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.806962490081787, + "learning_rate": 3.86068237205524e-06, + "loss": 1.5419, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.9716053009033203, + "learning_rate": 3.858651502843217e-06, + "loss": 1.5691, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.7608845233917236, + "eval_runtime": 37.1433, + "eval_samples_per_second": 26.923, + "eval_steps_per_second": 1.131, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.613098382949829, + "learning_rate": 3.856620633631195e-06, + "loss": 1.5297, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.5690557956695557, + "learning_rate": 3.854589764419171e-06, + "loss": 1.4964, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.8001089096069336, + "learning_rate": 3.852558895207149e-06, + "loss": 1.5253, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.6715309619903564, + "learning_rate": 3.850528025995126e-06, + "loss": 1.601, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.7044479846954346, + "learning_rate": 3.848497156783104e-06, + "loss": 1.5526, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.7368102073669434, + "learning_rate": 3.84646628757108e-06, + "loss": 1.5088, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.9358696937561035, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.5202, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.858172655105591, + "learning_rate": 3.842404549147035e-06, + "loss": 1.5378, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.6723833084106445, + "learning_rate": 3.840373679935013e-06, + "loss": 1.5134, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.748187780380249, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.4797, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.75319242477417, + "eval_runtime": 37.1302, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.4821436405181885, + "learning_rate": 3.836311941510967e-06, + "loss": 1.5382, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.5585227012634277, + "learning_rate": 3.834281072298944e-06, + "loss": 1.5373, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.927334785461426, + "learning_rate": 3.832250203086921e-06, + "loss": 1.5447, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.6096436977386475, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.5062, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.677304744720459, + "learning_rate": 3.828188464662876e-06, + "loss": 1.5671, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.9606077671051025, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6138, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.6971657276153564, + "learning_rate": 3.82412672623883e-06, + "loss": 1.5544, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.6380839347839355, + "learning_rate": 3.822095857026808e-06, + "loss": 1.4964, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.7537858486175537, + "learning_rate": 3.820064987814785e-06, + "loss": 1.5429, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.6189136505126953, + "learning_rate": 3.818034118602762e-06, + "loss": 1.4956, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.7546342611312866, + "eval_runtime": 37.0828, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5744292736053467, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.5256, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.5503833293914795, + "learning_rate": 3.813972380178717e-06, + "loss": 1.5211, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.979384422302246, + "learning_rate": 3.811941510966694e-06, + "loss": 1.55, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.867161512374878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5655, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.8006832599639893, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.4902, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.594712734222412, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.4451, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.6632349491119385, + "learning_rate": 3.803818034118603e-06, + "loss": 1.531, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.859894275665283, + "learning_rate": 3.80178716490658e-06, + "loss": 1.5018, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.6405653953552246, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.5765, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.7285192012786865, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.5537, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.7618242502212524, + "eval_runtime": 37.0898, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.6882379055023193, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.5143, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.706550121307373, + "learning_rate": 3.793663688058489e-06, + "loss": 1.4857, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.790907382965088, + "learning_rate": 3.791632818846467e-06, + "loss": 1.5982, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.631096363067627, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.5656, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.6981775760650635, + "learning_rate": 3.787571080422421e-06, + "loss": 1.567, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.612086296081543, + "learning_rate": 3.785540211210398e-06, + "loss": 1.5723, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.718630075454712, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.5565, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.6265041828155518, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.5326, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.578124523162842, + "learning_rate": 3.77944760357433e-06, + "loss": 1.533, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.832695484161377, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6133, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.7516050338745117, + "eval_runtime": 37.3173, + "eval_samples_per_second": 26.797, + "eval_steps_per_second": 1.125, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.851625442504883, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.5836, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.773354995938262e-06, + "loss": 1.5069, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.5738255977630615, + "learning_rate": 3.771324126726239e-06, + "loss": 1.5531, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.583575963973999, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.5589, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.7729389667510986, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.5901, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.6322245597839355, + "learning_rate": 3.765231519090171e-06, + "loss": 1.5423, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.6382699012756348, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.4574, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.7393112182617188, + "learning_rate": 3.761169780666125e-06, + "loss": 1.5351, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.684398651123047, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.4873, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.89401912689209, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.5358, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.7523635625839233, + "eval_runtime": 37.3763, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.629870891571045, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.5566, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.6936254501342773, + "learning_rate": 3.753046303818034e-06, + "loss": 1.5016, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.5889103412628174, + "learning_rate": 3.751015434606012e-06, + "loss": 1.5775, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.610703468322754, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.5508, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6318750381469727, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.5892, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.6650729179382324, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.5106, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.577174186706543, + "learning_rate": 3.742891957757921e-06, + "loss": 1.54, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.7274489402770996, + "learning_rate": 3.740861088545898e-06, + "loss": 1.5803, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6849985122680664, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.5393, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.811479091644287, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.5862, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.7490049600601196, + "eval_runtime": 36.9083, + "eval_samples_per_second": 27.094, + "eval_steps_per_second": 1.138, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.640002489089966, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.5149, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.6350319385528564, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5132, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.689814567565918, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.473, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.7181100845336914, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.4353, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.760629415512085, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.4859, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.9322562217712402, + "learning_rate": 3.724614134849716e-06, + "loss": 1.4733, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.280315399169922, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.5769, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 3.1343884468078613, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.4766, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.8537609577178955, + "learning_rate": 3.718521527213648e-06, + "loss": 1.5586, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.895299196243286, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5232, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.7473852634429932, + "eval_runtime": 37.006, + "eval_samples_per_second": 27.023, + "eval_steps_per_second": 1.135, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.82454776763916, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5057, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.711296558380127, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.4187, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.6281023025512695, + "learning_rate": 3.710398050365557e-06, + "loss": 1.4997, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.775632381439209, + "learning_rate": 3.708367181153534e-06, + "loss": 1.5087, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.8021788597106934, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.474, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 3.09647536277771, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5015, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.810450315475464, + "learning_rate": 3.702274573517466e-06, + "loss": 1.4681, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.65166974067688, + "learning_rate": 3.700243704305443e-06, + "loss": 1.4484, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.755690574645996, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5084, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.8223068714141846, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.5015, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.7516940832138062, + "eval_runtime": 37.0227, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.8401107788085938, + "learning_rate": 3.694151096669375e-06, + "loss": 1.4597, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 3.010011672973633, + "learning_rate": 3.692120227457352e-06, + "loss": 1.5625, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8575570583343506, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.4828, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.8527276515960693, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5074, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.81260085105896, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.4132, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.988429069519043, + "learning_rate": 3.683996750609261e-06, + "loss": 1.4803, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.964179515838623, + "learning_rate": 3.681965881397238e-06, + "loss": 1.4871, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.7151682376861572, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.5061, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.764187812805176, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.5188, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.810554265975952, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.4779, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.7422841787338257, + "eval_runtime": 37.0641, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.7640810012817383, + "learning_rate": 3.673842404549147e-06, + "loss": 1.4827, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.7683937549591064, + "learning_rate": 3.671811535337125e-06, + "loss": 1.4775, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.86434268951416, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.4752, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 3.0976617336273193, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.5099, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 3.107388496398926, + "learning_rate": 3.665718927701056e-06, + "loss": 1.5646, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 3.032921314239502, + "learning_rate": 3.663688058489034e-06, + "loss": 1.5048, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.9377830028533936, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5333, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.706542730331421, + "learning_rate": 3.659626320064988e-06, + "loss": 1.4903, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.819936513900757, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.4441, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.805360794067383, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.5122, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.747745156288147, + "eval_runtime": 37.0801, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.848580837249756, + "learning_rate": 3.65353371242892e-06, + "loss": 1.4984, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.772721290588379, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5026, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.760772943496704, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.4005, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.937507390975952, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.5131, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.923210859298706, + "learning_rate": 3.645410235580829e-06, + "loss": 1.4661, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8029167652130127, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.518, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.752716541290283, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.5109, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.7767865657806396, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.4831, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8483617305755615, + "learning_rate": 3.637286758732738e-06, + "loss": 1.4872, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.8714234828948975, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.5109, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.7455583810806274, + "eval_runtime": 37.0892, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7373011112213135, + "learning_rate": 3.633225020308692e-06, + "loss": 1.55, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.942373752593994, + "learning_rate": 3.63119415109667e-06, + "loss": 1.5079, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.8790931701660156, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.4669, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.9564976692199707, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5555, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.811699390411377, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.4457, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.926544427871704, + "learning_rate": 3.623070674248579e-06, + "loss": 1.4368, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.8694913387298584, + "learning_rate": 3.621039805036556e-06, + "loss": 1.4875, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 3.0294361114501953, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.4815, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8643198013305664, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.5031, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 3.0244719982147217, + "learning_rate": 3.614947197400488e-06, + "loss": 1.4914, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.7504756450653076, + "eval_runtime": 37.0238, + "eval_samples_per_second": 27.01, + "eval_steps_per_second": 1.134, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.8641672134399414, + "learning_rate": 3.612916328188465e-06, + "loss": 1.5122, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.8713624477386475, + "learning_rate": 3.610885458976442e-06, + "loss": 1.43, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.990912914276123, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.5124, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.963945150375366, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.533, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.802168846130371, + "learning_rate": 3.604792851340374e-06, + "loss": 1.4531, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 3.0556633472442627, + "learning_rate": 3.602761982128351e-06, + "loss": 1.4618, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.95744252204895, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5356, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.813145399093628, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.5134, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.8382728099823, + "learning_rate": 3.596669374492283e-06, + "loss": 1.4413, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.821100950241089, + "learning_rate": 3.59463850528026e-06, + "loss": 1.4748, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.7488586902618408, + "eval_runtime": 37.0361, + "eval_samples_per_second": 27.001, + "eval_steps_per_second": 1.134, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.930708408355713, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.5177, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.9312615394592285, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.4493, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 3.0051841735839844, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.5208, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.793515205383301, + "learning_rate": 3.586515028432169e-06, + "loss": 1.4813, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.7595956325531006, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5379, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.992871046066284, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.5153, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.7270801067352295, + "learning_rate": 3.580422420796101e-06, + "loss": 1.4392, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.797985076904297, + "learning_rate": 3.578391551584078e-06, + "loss": 1.4728, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9595260620117188, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.5158, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.8131442070007324, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.4214, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.7503563165664673, + "eval_runtime": 37.3831, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.124, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8242709636688232, + "learning_rate": 3.57229894394801e-06, + "loss": 1.5027, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.8973116874694824, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.5294, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 3.1207027435302734, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.4902, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.803910732269287, + "learning_rate": 3.566206336311942e-06, + "loss": 1.466, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.813541889190674, + "learning_rate": 3.564175467099919e-06, + "loss": 1.4957, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.9068171977996826, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.4435, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.8491995334625244, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.4432, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.8785598278045654, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.4791, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7701117992401123, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.4607, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.4821, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.7518969774246216, + "eval_runtime": 37.4808, + "eval_samples_per_second": 26.68, + "eval_steps_per_second": 1.121, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 3.0595574378967285, + "learning_rate": 3.554021121039805e-06, + "loss": 1.5228, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.8513107299804688, + "learning_rate": 3.551990251827783e-06, + "loss": 1.4812, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.937448024749756, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.4538, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.8307411670684814, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.4807, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.7986767292022705, + "learning_rate": 3.545897644191714e-06, + "loss": 1.4706, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 3.1135106086730957, + "learning_rate": 3.543866774979692e-06, + "loss": 1.4574, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.7756283283233643, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5114, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.8114547729492188, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.4458, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 3.0544545650482178, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.4934, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8990671634674072, + "learning_rate": 3.535743298131601e-06, + "loss": 1.4808, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.7468736171722412, + "eval_runtime": 36.9135, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.976139545440674, + "learning_rate": 3.533712428919578e-06, + "loss": 1.5123, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.909146785736084, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.5204, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7225229740142822, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.4746, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.6145517826080322, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.5049, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.8660519123077393, + "learning_rate": 3.525588952071487e-06, + "loss": 1.5173, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.979801654815674, + "learning_rate": 3.523558082859464e-06, + "loss": 1.5146, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.7500879764556885, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.514, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.6923978328704834, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.455, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.6728951930999756, + "learning_rate": 3.517465475223396e-06, + "loss": 1.4712, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.769507884979248, + "learning_rate": 3.515434606011373e-06, + "loss": 1.444, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.7491599321365356, + "eval_runtime": 37.066, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.9768483638763428, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.5005, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.993196725845337, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.5347, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.886883497238159, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.4959, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.9152941703796387, + "learning_rate": 3.507311129163282e-06, + "loss": 1.4641, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.966872215270996, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.4232, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.9853367805480957, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.4883, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.9353625774383545, + "learning_rate": 3.501218521527214e-06, + "loss": 1.4826, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.7860453128814697, + "learning_rate": 3.499187652315191e-06, + "loss": 1.42, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.9269113540649414, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.5087, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 3.052962303161621, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5042, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.7421623468399048, + "eval_runtime": 37.103, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.7328579425811768, + "learning_rate": 3.493095044679123e-06, + "loss": 1.4375, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.876138925552368, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5559, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.7500345706939697, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.4043, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.836596965789795, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.4808, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.797060966491699, + "learning_rate": 3.484971567831032e-06, + "loss": 1.405, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.928368091583252, + "learning_rate": 3.482940698619009e-06, + "loss": 1.4765, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.923375368118286, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.5134, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.8624660968780518, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.4881, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.9593892097473145, + "learning_rate": 3.476848090982941e-06, + "loss": 1.4869, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.8019728660583496, + "learning_rate": 3.474817221770918e-06, + "loss": 1.468, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.7386831045150757, + "eval_runtime": 37.1034, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.8965771198272705, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.5604, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.8114798069000244, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.5092, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.9554526805877686, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.4869, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.981499195098877, + "learning_rate": 3.466693744922827e-06, + "loss": 1.4454, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.8626201152801514, + "learning_rate": 3.464662875710805e-06, + "loss": 1.4701, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.9752259254455566, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.5004, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 3.1553750038146973, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.4652, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.9458937644958496, + "learning_rate": 3.458570268074736e-06, + "loss": 1.51, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.9626898765563965, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.484, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.8093032836914062, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.468, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.740411639213562, + "eval_runtime": 37.1036, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.928344488143921, + "learning_rate": 3.452477660438668e-06, + "loss": 1.4474, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 3.0413129329681396, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.4759, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.7930357456207275, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.4406, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.8075618743896484, + "learning_rate": 3.4463850528026e-06, + "loss": 1.48, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 3.019402503967285, + "learning_rate": 3.444354183590577e-06, + "loss": 1.4713, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 3.009504795074463, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5124, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.9452531337738037, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.464, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.6980783939361572, + "learning_rate": 3.438261575954509e-06, + "loss": 1.4143, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 3.039795160293579, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.4712, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.285968542098999, + "learning_rate": 3.434199837530463e-06, + "loss": 1.5227, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.7402914762496948, + "eval_runtime": 37.1273, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 3.035813093185425, + "learning_rate": 3.432168968318441e-06, + "loss": 1.4772, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 3.1172401905059814, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.4389, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 3.0828707218170166, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.5098, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.922488212585449, + "learning_rate": 3.426076360682372e-06, + "loss": 1.5239, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 3.03352689743042, + "learning_rate": 3.42404549147035e-06, + "loss": 1.5296, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.933380603790283, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.4851, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7250027656555176, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.3966, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.991846799850464, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.4863, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 3.0704684257507324, + "learning_rate": 3.415922014622259e-06, + "loss": 1.5366, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 3.0659193992614746, + "learning_rate": 3.413891145410236e-06, + "loss": 1.516, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.7458266019821167, + "eval_runtime": 37.06, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.9695920944213867, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.508, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 3.004558801651001, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.5092, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.8377230167388916, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.4909, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.860020875930786, + "learning_rate": 3.405767668562145e-06, + "loss": 1.4415, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8683009147644043, + "learning_rate": 3.403736799350122e-06, + "loss": 1.4997, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.91894268989563, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.4886, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.8158814907073975, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.4295, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.9528515338897705, + "learning_rate": 3.397644191714054e-06, + "loss": 1.5422, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.994856119155884, + "learning_rate": 3.395613322502031e-06, + "loss": 1.4909, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.8976950645446777, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.4959, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.7350144386291504, + "eval_runtime": 37.0807, + "eval_samples_per_second": 26.968, + "eval_steps_per_second": 1.133, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.73504638671875, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.4454, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.816823720932007, + "learning_rate": 3.389520714865963e-06, + "loss": 1.4864, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.706554889678955, + "learning_rate": 3.38748984565394e-06, + "loss": 1.483, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.9004647731781006, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.492, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8677759170532227, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.4786, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.83447527885437, + "learning_rate": 3.381397238017872e-06, + "loss": 1.4575, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.9942550659179688, + "learning_rate": 3.379366368805849e-06, + "loss": 1.4746, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.124676465988159, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.5406, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.9151222705841064, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.4748, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.734571933746338, + "learning_rate": 3.373273761169781e-06, + "loss": 1.4773, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.7389248609542847, + "eval_runtime": 37.1519, + "eval_samples_per_second": 26.917, + "eval_steps_per_second": 1.13, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.8919990062713623, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.4698, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.873870611190796, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.4509, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 3.0072309970855713, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5024, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.784576177597046, + "learning_rate": 3.36515028432169e-06, + "loss": 1.4405, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.8316547870635986, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.4697, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.8758575916290283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.495, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.879115343093872, + "learning_rate": 3.359057676685622e-06, + "loss": 1.5248, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.9559383392333984, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5229, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.6783695220947266, + "learning_rate": 3.354995938261576e-06, + "loss": 1.4251, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.8372304439544678, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.4649, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.7335349321365356, + "eval_runtime": 37.3498, + "eval_samples_per_second": 26.774, + "eval_steps_per_second": 1.125, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.729576826095581, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.4263, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.7717669010162354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.4467, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.8166871070861816, + "learning_rate": 3.346872461413485e-06, + "loss": 1.4937, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.845677137374878, + "learning_rate": 3.344841592201463e-06, + "loss": 1.4303, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.9527740478515625, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.4651, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.810873031616211, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.4521, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.77546763420105, + "learning_rate": 3.338748984565394e-06, + "loss": 1.4857, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.9098589420318604, + "learning_rate": 3.336718115353372e-06, + "loss": 1.4988, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.797712802886963, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.5307, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.916668176651001, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.4828, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.7372808456420898, + "eval_runtime": 36.888, + "eval_samples_per_second": 27.109, + "eval_steps_per_second": 1.139, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.8220417499542236, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.4636, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.851010322570801, + "learning_rate": 3.32859463850528e-06, + "loss": 1.4579, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.983619213104248, + "learning_rate": 3.326563769293258e-06, + "loss": 1.4801, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.750732898712158, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.4438, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.631105422973633, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.4439, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.944300889968872, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.437, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8841664791107178, + "learning_rate": 3.318440292445167e-06, + "loss": 1.4201, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.917189121246338, + "learning_rate": 3.316409423233144e-06, + "loss": 1.4489, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.112077236175537, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.4449, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.9847209453582764, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.5153, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.7323338985443115, + "eval_runtime": 36.9935, + "eval_samples_per_second": 27.032, + "eval_steps_per_second": 1.135, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.948901414871216, + "learning_rate": 3.310316815597076e-06, + "loss": 1.473, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.84930419921875, + "learning_rate": 3.308285946385053e-06, + "loss": 1.4796, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.9120237827301025, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5392, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.7088463306427, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.4206, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.7495498657226562, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.4759, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.9146509170532227, + "learning_rate": 3.300162469536962e-06, + "loss": 1.51, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.781975030899048, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.4667, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.7524054050445557, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.5036, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.7672629356384277, + "learning_rate": 3.294069861900894e-06, + "loss": 1.4827, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.875802516937256, + "learning_rate": 3.292038992688871e-06, + "loss": 1.519, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.7273142337799072, + "eval_runtime": 37.0032, + "eval_samples_per_second": 27.025, + "eval_steps_per_second": 1.135, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.8699231147766113, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.474, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.8670589923858643, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.5039, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.9946320056915283, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5226, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.767071008682251, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.4221, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8807036876678467, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.486, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.687720537185669, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.3944, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.8170154094696045, + "learning_rate": 3.277822908204712e-06, + "loss": 1.5213, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.905557632446289, + "learning_rate": 3.275792038992689e-06, + "loss": 1.4598, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.643657922744751, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.4282, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.798712968826294, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.4752, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.7295492887496948, + "eval_runtime": 37.0432, + "eval_samples_per_second": 26.996, + "eval_steps_per_second": 1.134, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.8480567932128906, + "learning_rate": 3.269699431356621e-06, + "loss": 1.4322, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.8311381340026855, + "learning_rate": 3.267668562144598e-06, + "loss": 1.4744, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.83522891998291, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.4578, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.757307767868042, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.4469, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.9117610454559326, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.5034, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.8141143321990967, + "learning_rate": 3.259545085296507e-06, + "loss": 1.473, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 3.050138473510742, + "learning_rate": 3.257514216084484e-06, + "loss": 1.528, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.660510540008545, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.4335, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.898832082748413, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.5335, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.955589532852173, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.4832, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.7293461561203003, + "eval_runtime": 37.0096, + "eval_samples_per_second": 27.02, + "eval_steps_per_second": 1.135, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.8524296283721924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.4728, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.9741404056549072, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.4805, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.738152265548706, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.4532, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.717066764831543, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.4741, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.8367488384246826, + "learning_rate": 3.241267262388302e-06, + "loss": 1.4773, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.987780809402466, + "learning_rate": 3.23923639317628e-06, + "loss": 1.4214, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.705444097518921, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.4537, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.8966188430786133, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.4552, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.9027600288391113, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.4883, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.965813159942627, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5318, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.7236050367355347, + "eval_runtime": 37.0306, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.6390798091888428, + "learning_rate": 3.229082047116166e-06, + "loss": 1.4431, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.822848081588745, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5163, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7289068698883057, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.4551, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.766545057296753, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.4648, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.8657522201538086, + "learning_rate": 3.220958570268075e-06, + "loss": 1.4825, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.7456932067871094, + "learning_rate": 3.218927701056052e-06, + "loss": 1.3909, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 3.0274853706359863, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.4461, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.990029811859131, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.4057, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 3.086756706237793, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5251, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 3.0227134227752686, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.427, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.7248852252960205, + "eval_runtime": 37.0619, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.815480947494507, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.469, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.9171998500823975, + "learning_rate": 3.206742485783916e-06, + "loss": 1.4977, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 3.087686777114868, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.4233, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.7349331378936768, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.4187, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.9189705848693848, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.4949, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 3.0483345985412598, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5041, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.7958145141601562, + "learning_rate": 3.196588139723802e-06, + "loss": 1.4009, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 3.081714153289795, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.4718, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.7897303104400635, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.4926, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.899045944213867, + "learning_rate": 3.190495532087734e-06, + "loss": 1.4524, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.7273337841033936, + "eval_runtime": 37.0682, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.979063034057617, + "learning_rate": 3.188464662875711e-06, + "loss": 1.4706, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.8633804321289062, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5273, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.8997795581817627, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.4638, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.906151294708252, + "learning_rate": 3.182372055239643e-06, + "loss": 1.4187, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 3.0212671756744385, + "learning_rate": 3.18034118602762e-06, + "loss": 1.4213, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 3.0891034603118896, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.4921, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.9302260875701904, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.4787, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.8741140365600586, + "learning_rate": 3.174248578391552e-06, + "loss": 1.4591, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 3.0567023754119873, + "learning_rate": 3.172217709179529e-06, + "loss": 1.4636, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.80245304107666, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.4523, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.7270530462265015, + "eval_runtime": 37.325, + "eval_samples_per_second": 26.792, + "eval_steps_per_second": 1.125, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.816023349761963, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.4579, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.7405545711517334, + "learning_rate": 3.166125101543461e-06, + "loss": 1.4225, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.9569272994995117, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.483, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 3.0433337688446045, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.4846, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.856781244277954, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.426, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.8880672454833984, + "learning_rate": 3.15800162469537e-06, + "loss": 1.4764, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 3.0363011360168457, + "learning_rate": 3.155970755483347e-06, + "loss": 1.4916, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 3.0171539783477783, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.4596, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 3.183234930038452, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.4936, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.802027463912964, + "learning_rate": 3.149878147847279e-06, + "loss": 1.4733, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.7253632545471191, + "eval_runtime": 37.4737, + "eval_samples_per_second": 26.685, + "eval_steps_per_second": 1.121, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.9667463302612305, + "learning_rate": 3.147847278635256e-06, + "loss": 1.508, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 3.1557469367980957, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.5459, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.7861125469207764, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.4614, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.7431211471557617, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.4297, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 3.140911340713501, + "learning_rate": 3.139723801787165e-06, + "loss": 1.4497, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.970118522644043, + "learning_rate": 3.137692932575143e-06, + "loss": 1.4409, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.94400954246521, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.4889, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.925431966781616, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.4205, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.997279167175293, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.4714, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.855996608734131, + "learning_rate": 3.129569455727051e-06, + "loss": 1.4925, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.7247990369796753, + "eval_runtime": 36.9312, + "eval_samples_per_second": 27.077, + "eval_steps_per_second": 1.137, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 3.011199951171875, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.5132, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.8528099060058594, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.4791, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.8827919960021973, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.495, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.8150227069854736, + "learning_rate": 3.12144597887896e-06, + "loss": 1.4563, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.9341959953308105, + "learning_rate": 3.119415109666938e-06, + "loss": 1.4829, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.9161651134490967, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.492, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.906963348388672, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.445, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.992582321166992, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.4988, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 3.005469799041748, + "learning_rate": 3.111291632818847e-06, + "loss": 1.4999, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.7577736377716064, + "learning_rate": 3.109260763606824e-06, + "loss": 1.5152, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.724539041519165, + "eval_runtime": 37.0601, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.870109796524048, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.4746, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.7745094299316406, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.4891, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.6818361282348633, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.4575, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.823237180709839, + "learning_rate": 3.101137286758733e-06, + "loss": 1.492, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.9089932441711426, + "learning_rate": 3.09910641754671e-06, + "loss": 1.4943, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.85476016998291, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.5049, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 3.07242488861084, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.4709, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 3.117357015609741, + "learning_rate": 3.093013809910642e-06, + "loss": 1.52, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.959383964538574, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.4775, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8594462871551514, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.5113, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.721728801727295, + "eval_runtime": 37.1106, + "eval_samples_per_second": 26.946, + "eval_steps_per_second": 1.132, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.8418593406677246, + "learning_rate": 3.086921202274574e-06, + "loss": 1.482, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.762467384338379, + "learning_rate": 3.084890333062551e-06, + "loss": 1.4672, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.8023135662078857, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.431, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.7467143535614014, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.4837, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.6832900047302246, + "learning_rate": 3.078797725426483e-06, + "loss": 1.4765, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.863823175430298, + "learning_rate": 3.07676685621446e-06, + "loss": 1.4039, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.9055254459381104, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.4655, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.8665876388549805, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.449, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 3.029784917831421, + "learning_rate": 3.070674248578392e-06, + "loss": 1.4665, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.8582639694213867, + "learning_rate": 3.068643379366369e-06, + "loss": 1.4832, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.722991704940796, + "eval_runtime": 37.0827, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.8655312061309814, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.4759, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.892972469329834, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.4363, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.7701165676116943, + "learning_rate": 3.062550771730301e-06, + "loss": 1.4798, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7921078205108643, + "learning_rate": 3.060519902518278e-06, + "loss": 1.4557, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.8233699798583984, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.4598, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.8705663681030273, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.3898, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.9667348861694336, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.4865, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 3.0962018966674805, + "learning_rate": 3.052396425670187e-06, + "loss": 1.4512, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 3.06290340423584, + "learning_rate": 3.050365556458164e-06, + "loss": 1.5129, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.8010520935058594, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.4525, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.7228375673294067, + "eval_runtime": 37.1149, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.73380446434021, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.4537, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.9469099044799805, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.4624, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.8733043670654297, + "learning_rate": 3.042242079610073e-06, + "loss": 1.4981, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.867926597595215, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.4966, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.8584275245666504, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.4161, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.8059520721435547, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.4511, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.916750192642212, + "learning_rate": 3.034118602761982e-06, + "loss": 1.4694, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.979952335357666, + "learning_rate": 3.03208773354996e-06, + "loss": 1.4624, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.807194709777832, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.4227, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.880439519882202, + "learning_rate": 3.028025995125914e-06, + "loss": 1.5094, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.7204533815383911, + "eval_runtime": 37.1346, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.7696263790130615, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.492, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.840989351272583, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.4229, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.948915481567383, + "learning_rate": 3.021933387489846e-06, + "loss": 1.4601, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.9163503646850586, + "learning_rate": 3.019902518277823e-06, + "loss": 1.4218, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 3.0191619396209717, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.4648, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.9332194328308105, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.4791, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.8832461833953857, + "learning_rate": 3.013809910641755e-06, + "loss": 1.43, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.886837959289551, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.4455, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.8201355934143066, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.4417, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.9708259105682373, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.4657, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.7213743925094604, + "eval_runtime": 37.0993, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.927104949951172, + "learning_rate": 3.005686433793664e-06, + "loss": 1.4406, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.9701600074768066, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.4692, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 3.0241730213165283, + "learning_rate": 3.001624695369618e-06, + "loss": 1.4336, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.8802919387817383, + "learning_rate": 2.999593826157596e-06, + "loss": 1.4948, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.7666738033294678, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.4518, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 3.1206812858581543, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.4707, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 3.069323778152466, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.4534, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 3.155106544494629, + "learning_rate": 2.991470349309505e-06, + "loss": 1.5153, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.923816680908203, + "learning_rate": 2.989439480097482e-06, + "loss": 1.4606, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 3.2247867584228516, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.4564, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.7207131385803223, + "eval_runtime": 37.0684, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.8193411827087402, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.4721, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.865691661834717, + "learning_rate": 2.983346872461414e-06, + "loss": 1.5072, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 3.125004529953003, + "learning_rate": 2.981316003249391e-06, + "loss": 1.4913, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.873455286026001, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.4699, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.986186981201172, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.4808, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.8694097995758057, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.495, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.8555796146392822, + "learning_rate": 2.9731925264013e-06, + "loss": 1.4481, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.8808774948120117, + "learning_rate": 2.971161657189277e-06, + "loss": 1.4361, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.8831286430358887, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.4489, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.7533552646636963, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.4386, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.7256641387939453, + "eval_runtime": 37.1621, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 2.783586025238037, + "learning_rate": 2.965069049553209e-06, + "loss": 1.4476, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.9043776988983154, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5077, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.838284492492676, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.4287, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.94248366355896, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.4987, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.9397339820861816, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.4809, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.8922536373138428, + "learning_rate": 2.954914703493095e-06, + "loss": 1.426, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.787680149078369, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.4549, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.8116507530212402, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.4324, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.7330825328826904, + "learning_rate": 2.948822095857027e-06, + "loss": 1.4244, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.806539535522461, + "learning_rate": 2.946791226645004e-06, + "loss": 1.4604, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.7229852676391602, + "eval_runtime": 37.378, + "eval_samples_per_second": 26.754, + "eval_steps_per_second": 1.124, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.9335062503814697, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.4433, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.8057737350463867, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.4242, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 2.7581822872161865, + "learning_rate": 2.940698619008936e-06, + "loss": 1.4374, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.931168794631958, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.4588, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.865867853164673, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.4731, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.8826231956481934, + "learning_rate": 2.934606011372868e-06, + "loss": 1.4568, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.922809600830078, + "learning_rate": 2.932575142160845e-06, + "loss": 1.4747, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.9709854125976562, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.4631, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8259403705596924, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.449, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.850677490234375, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.4725, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.7200536727905273, + "eval_runtime": 36.8614, + "eval_samples_per_second": 27.129, + "eval_steps_per_second": 1.139, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.8449628353118896, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.5022, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 3.004880666732788, + "learning_rate": 2.922420796100731e-06, + "loss": 1.4606, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.737196445465088, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.3911, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.812662124633789, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.4873, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.7756738662719727, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.3975, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.8784940242767334, + "learning_rate": 2.91429731925264e-06, + "loss": 1.4726, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 2.946704626083374, + "learning_rate": 2.912266450040618e-06, + "loss": 1.4841, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.8813061714172363, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.4656, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.9221038818359375, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.4505, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 2.9353513717651367, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.4585, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.7200087308883667, + "eval_runtime": 36.9952, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.7565391063690186, + "learning_rate": 2.904142973192527e-06, + "loss": 1.4199, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 2.824780225753784, + "learning_rate": 2.902112103980504e-06, + "loss": 1.5003, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.963022232055664, + "learning_rate": 2.900081234768481e-06, + "loss": 1.4649, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.744490385055542, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.4677, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.8483076095581055, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.4291, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.878147602081299, + "learning_rate": 2.893988627132413e-06, + "loss": 1.4278, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.866380214691162, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.4491, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.802532911300659, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.397, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.863152265548706, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.4573, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.800173044204712, + "learning_rate": 2.885865150284322e-06, + "loss": 1.4466, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.7167619466781616, + "eval_runtime": 37.0136, + "eval_samples_per_second": 27.017, + "eval_steps_per_second": 1.135, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.975752830505371, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.4918, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.7768948078155518, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.4549, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.761004686355591, + "learning_rate": 2.879772542648254e-06, + "loss": 1.4039, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.9454665184020996, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.4542, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.85044264793396, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.3875, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.9699220657348633, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.5084, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.8547441959381104, + "learning_rate": 2.871649065800163e-06, + "loss": 1.4839, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 3.0871565341949463, + "learning_rate": 2.86961819658814e-06, + "loss": 1.5105, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 3.662945032119751, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.4483, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.916095495223999, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.455, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.7137683629989624, + "eval_runtime": 37.0176, + "eval_samples_per_second": 27.014, + "eval_steps_per_second": 1.135, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.857478141784668, + "learning_rate": 2.863525588952072e-06, + "loss": 1.4531, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.9542508125305176, + "learning_rate": 2.861494719740049e-06, + "loss": 1.488, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.677672863006592, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.3725, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.8790128231048584, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.4252, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.95095157623291, + "learning_rate": 2.855402112103981e-06, + "loss": 1.4532, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.997403383255005, + "learning_rate": 2.853371242891958e-06, + "loss": 1.4581, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.926973581314087, + "learning_rate": 2.851340373679935e-06, + "loss": 1.4368, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.8602488040924072, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.4427, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.875828504562378, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.4757, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.825498580932617, + "learning_rate": 2.845247766043867e-06, + "loss": 1.4535, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.7157984972000122, + "eval_runtime": 37.0329, + "eval_samples_per_second": 27.003, + "eval_steps_per_second": 1.134, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.8068923950195312, + "learning_rate": 2.843216896831844e-06, + "loss": 1.4391, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.8095016479492188, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.397, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.7768023014068604, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.4415, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.7868361473083496, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.4726, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 3.085806131362915, + "learning_rate": 2.835093419983753e-06, + "loss": 1.4735, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.972489833831787, + "learning_rate": 2.833062550771731e-06, + "loss": 1.4415, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.9990909099578857, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.4203, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.849210023880005, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.4202, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.899545907974243, + "learning_rate": 2.826969943135662e-06, + "loss": 1.4541, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.7620913982391357, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.4672, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.718942642211914, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.8633880615234375, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.4548, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.809718370437622, + "learning_rate": 2.820877335499594e-06, + "loss": 1.4379, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.780914306640625, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.4244, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.6853432655334473, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.3999, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.8040759563446045, + "learning_rate": 2.814784727863526e-06, + "loss": 1.4151, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.974795341491699, + "learning_rate": 2.812753858651503e-06, + "loss": 1.4497, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.9964492321014404, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.4512, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 2.9635047912597656, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.4104, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 3.0267832279205322, + "learning_rate": 2.806661251015435e-06, + "loss": 1.4149, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 3.0778963565826416, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.5034, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.723001480102539, + "eval_runtime": 37.0513, + "eval_samples_per_second": 26.99, + "eval_steps_per_second": 1.134, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.9889729022979736, + "learning_rate": 2.802599512591389e-06, + "loss": 1.4851, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 3.111581325531006, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.5337, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.727510452270508, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.4647, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.848571300506592, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.4107, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 3.178166627883911, + "learning_rate": 2.794476035743298e-06, + "loss": 1.4427, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.9924423694610596, + "learning_rate": 2.792445166531276e-06, + "loss": 1.5024, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.825277328491211, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.4687, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 3.082688808441162, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.5057, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.8399462699890137, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.4274, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.8092541694641113, + "learning_rate": 2.784321689683185e-06, + "loss": 1.4157, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.7190473079681396, + "eval_runtime": 37.0662, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.8798582553863525, + "learning_rate": 2.782290820471162e-06, + "loss": 1.4213, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.7891805171966553, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.4181, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.865835666656494, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.4896, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.996018171310425, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.4829, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.868534803390503, + "learning_rate": 2.774167343623071e-06, + "loss": 1.4592, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.906359910964966, + "learning_rate": 2.772136474411048e-06, + "loss": 1.5205, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 3.099578857421875, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.4246, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.944490432739258, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.3856, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.870277166366577, + "learning_rate": 2.76604386677498e-06, + "loss": 1.4648, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.853956937789917, + "learning_rate": 2.764012997562957e-06, + "loss": 1.4317, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.7165414094924927, + "eval_runtime": 37.4056, + "eval_samples_per_second": 26.734, + "eval_steps_per_second": 1.123, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 3.0553836822509766, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.4649, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.973536729812622, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.4129, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 3.025920867919922, + "learning_rate": 2.757920389926889e-06, + "loss": 1.4974, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.853586196899414, + "learning_rate": 2.755889520714866e-06, + "loss": 1.423, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 3.0477190017700195, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.4149, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.8930039405822754, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.3938, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.8823812007904053, + "learning_rate": 2.749796913078798e-06, + "loss": 1.46, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.96528959274292, + "learning_rate": 2.747766043866775e-06, + "loss": 1.4818, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.892967939376831, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.4392, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.851961851119995, + "learning_rate": 2.7437043054427297e-06, + "loss": 1.4903, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.7118008136749268, + "eval_runtime": 37.4958, + "eval_samples_per_second": 26.67, + "eval_steps_per_second": 1.12, + "step": 1400 + }, + { + "epoch": 1.5353424657534247, + "grad_norm": 2.9262213706970215, + "learning_rate": 2.741673436230707e-06, + "loss": 1.5253, + "step": 1401 + }, + { + "epoch": 1.5364383561643835, + "grad_norm": 3.0567355155944824, + "learning_rate": 2.739642567018684e-06, + "loss": 1.4442, + "step": 1402 + }, + { + "epoch": 1.5375342465753423, + "grad_norm": 3.0198726654052734, + "learning_rate": 2.7376116978066615e-06, + "loss": 1.4408, + "step": 1403 + }, + { + "epoch": 1.5386301369863014, + "grad_norm": 2.9126687049865723, + "learning_rate": 2.7355808285946388e-06, + "loss": 1.4421, + "step": 1404 + }, + { + "epoch": 1.5397260273972604, + "grad_norm": 2.9556353092193604, + "learning_rate": 2.733549959382616e-06, + "loss": 1.436, + "step": 1405 + }, + { + "epoch": 1.5408219178082192, + "grad_norm": 2.8351075649261475, + "learning_rate": 2.7315190901705933e-06, + "loss": 1.4444, + "step": 1406 + }, + { + "epoch": 1.541917808219178, + "grad_norm": 2.987910270690918, + "learning_rate": 2.7294882209585706e-06, + "loss": 1.4969, + "step": 1407 + }, + { + "epoch": 1.543013698630137, + "grad_norm": 2.734234571456909, + "learning_rate": 2.727457351746548e-06, + "loss": 1.4703, + "step": 1408 + }, + { + "epoch": 1.5441095890410959, + "grad_norm": 2.7853105068206787, + "learning_rate": 2.725426482534525e-06, + "loss": 1.4241, + "step": 1409 + }, + { + "epoch": 1.5452054794520547, + "grad_norm": 2.856630802154541, + "learning_rate": 2.723395613322502e-06, + "loss": 1.3985, + "step": 1410 + }, + { + "epoch": 1.5452054794520547, + "eval_loss": 1.7150108814239502, + "eval_runtime": 36.8966, + "eval_samples_per_second": 27.103, + "eval_steps_per_second": 1.138, + "step": 1410 + }, + { + "epoch": 1.5463013698630137, + "grad_norm": 2.952057123184204, + "learning_rate": 2.7213647441104797e-06, + "loss": 1.4453, + "step": 1411 + }, + { + "epoch": 1.5473972602739727, + "grad_norm": 2.855661153793335, + "learning_rate": 2.7193338748984565e-06, + "loss": 1.4288, + "step": 1412 + }, + { + "epoch": 1.5484931506849315, + "grad_norm": 2.9883601665496826, + "learning_rate": 2.7173030056864342e-06, + "loss": 1.436, + "step": 1413 + }, + { + "epoch": 1.5495890410958904, + "grad_norm": 3.175748109817505, + "learning_rate": 2.715272136474411e-06, + "loss": 1.5645, + "step": 1414 + }, + { + "epoch": 1.5506849315068494, + "grad_norm": 2.8751583099365234, + "learning_rate": 2.713241267262389e-06, + "loss": 1.4167, + "step": 1415 + }, + { + "epoch": 1.5517808219178082, + "grad_norm": 2.9523749351501465, + "learning_rate": 2.7112103980503656e-06, + "loss": 1.4379, + "step": 1416 + }, + { + "epoch": 1.552876712328767, + "grad_norm": 2.709944486618042, + "learning_rate": 2.7091795288383433e-06, + "loss": 1.4286, + "step": 1417 + }, + { + "epoch": 1.553972602739726, + "grad_norm": 2.782998561859131, + "learning_rate": 2.70714865962632e-06, + "loss": 1.4758, + "step": 1418 + }, + { + "epoch": 1.555068493150685, + "grad_norm": 2.803555965423584, + "learning_rate": 2.705117790414298e-06, + "loss": 1.4211, + "step": 1419 + }, + { + "epoch": 1.5561643835616439, + "grad_norm": 2.792894124984741, + "learning_rate": 2.7030869212022747e-06, + "loss": 1.4532, + "step": 1420 + }, + { + "epoch": 1.5561643835616439, + "eval_loss": 1.713323712348938, + "eval_runtime": 37.0626, + "eval_samples_per_second": 26.981, + "eval_steps_per_second": 1.133, + "step": 1420 + }, + { + "epoch": 1.5572602739726027, + "grad_norm": 2.9909141063690186, + "learning_rate": 2.7010560519902516e-06, + "loss": 1.4718, + "step": 1421 + }, + { + "epoch": 1.5583561643835617, + "grad_norm": 2.9359354972839355, + "learning_rate": 2.6990251827782293e-06, + "loss": 1.444, + "step": 1422 + }, + { + "epoch": 1.5594520547945205, + "grad_norm": 2.986454963684082, + "learning_rate": 2.696994313566206e-06, + "loss": 1.45, + "step": 1423 + }, + { + "epoch": 1.5605479452054793, + "grad_norm": 2.997868776321411, + "learning_rate": 2.694963444354184e-06, + "loss": 1.4309, + "step": 1424 + }, + { + "epoch": 1.5616438356164384, + "grad_norm": 2.922935962677002, + "learning_rate": 2.6929325751421607e-06, + "loss": 1.4961, + "step": 1425 + }, + { + "epoch": 1.5627397260273974, + "grad_norm": 2.7861900329589844, + "learning_rate": 2.6909017059301384e-06, + "loss": 1.4482, + "step": 1426 + }, + { + "epoch": 1.563835616438356, + "grad_norm": 2.848219394683838, + "learning_rate": 2.6888708367181153e-06, + "loss": 1.4637, + "step": 1427 + }, + { + "epoch": 1.564931506849315, + "grad_norm": 2.6985557079315186, + "learning_rate": 2.686839967506093e-06, + "loss": 1.4138, + "step": 1428 + }, + { + "epoch": 1.566027397260274, + "grad_norm": 2.835800886154175, + "learning_rate": 2.68480909829407e-06, + "loss": 1.4669, + "step": 1429 + }, + { + "epoch": 1.5671232876712329, + "grad_norm": 2.9267354011535645, + "learning_rate": 2.6827782290820475e-06, + "loss": 1.3823, + "step": 1430 + }, + { + "epoch": 1.5671232876712329, + "eval_loss": 1.7126407623291016, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 1430 + }, + { + "epoch": 1.5682191780821917, + "grad_norm": 2.8740406036376953, + "learning_rate": 2.6807473598700244e-06, + "loss": 1.3837, + "step": 1431 + }, + { + "epoch": 1.5693150684931507, + "grad_norm": 3.012890100479126, + "learning_rate": 2.678716490658002e-06, + "loss": 1.495, + "step": 1432 + }, + { + "epoch": 1.5704109589041098, + "grad_norm": 2.877326011657715, + "learning_rate": 2.676685621445979e-06, + "loss": 1.4622, + "step": 1433 + }, + { + "epoch": 1.5715068493150683, + "grad_norm": 2.852553606033325, + "learning_rate": 2.674654752233956e-06, + "loss": 1.4415, + "step": 1434 + }, + { + "epoch": 1.5726027397260274, + "grad_norm": 2.9420814514160156, + "learning_rate": 2.6726238830219335e-06, + "loss": 1.4463, + "step": 1435 + }, + { + "epoch": 1.5736986301369864, + "grad_norm": 2.91435170173645, + "learning_rate": 2.6705930138099107e-06, + "loss": 1.4482, + "step": 1436 + }, + { + "epoch": 1.5747945205479452, + "grad_norm": 2.851418972015381, + "learning_rate": 2.668562144597888e-06, + "loss": 1.4752, + "step": 1437 + }, + { + "epoch": 1.575890410958904, + "grad_norm": 2.7865328788757324, + "learning_rate": 2.6665312753858653e-06, + "loss": 1.3752, + "step": 1438 + }, + { + "epoch": 1.576986301369863, + "grad_norm": 2.8747165203094482, + "learning_rate": 2.6645004061738426e-06, + "loss": 1.4709, + "step": 1439 + }, + { + "epoch": 1.5780821917808219, + "grad_norm": 2.986900568008423, + "learning_rate": 2.66246953696182e-06, + "loss": 1.4524, + "step": 1440 + }, + { + "epoch": 1.5780821917808219, + "eval_loss": 1.7110822200775146, + "eval_runtime": 37.1062, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 1440 + }, + { + "epoch": 1.5791780821917807, + "grad_norm": 2.9051966667175293, + "learning_rate": 2.660438667749797e-06, + "loss": 1.393, + "step": 1441 + }, + { + "epoch": 1.5802739726027397, + "grad_norm": 2.8744468688964844, + "learning_rate": 2.6584077985377744e-06, + "loss": 1.4233, + "step": 1442 + }, + { + "epoch": 1.5813698630136988, + "grad_norm": 2.8655996322631836, + "learning_rate": 2.6563769293257517e-06, + "loss": 1.4501, + "step": 1443 + }, + { + "epoch": 1.5824657534246576, + "grad_norm": 2.8580846786499023, + "learning_rate": 2.654346060113729e-06, + "loss": 1.4612, + "step": 1444 + }, + { + "epoch": 1.5835616438356164, + "grad_norm": 2.824436902999878, + "learning_rate": 2.6523151909017062e-06, + "loss": 1.4263, + "step": 1445 + }, + { + "epoch": 1.5846575342465754, + "grad_norm": 2.8548355102539062, + "learning_rate": 2.6502843216896835e-06, + "loss": 1.4219, + "step": 1446 + }, + { + "epoch": 1.5857534246575342, + "grad_norm": 2.8495287895202637, + "learning_rate": 2.6482534524776603e-06, + "loss": 1.419, + "step": 1447 + }, + { + "epoch": 1.586849315068493, + "grad_norm": 2.7665646076202393, + "learning_rate": 2.646222583265638e-06, + "loss": 1.5077, + "step": 1448 + }, + { + "epoch": 1.587945205479452, + "grad_norm": 2.859886646270752, + "learning_rate": 2.644191714053615e-06, + "loss": 1.4687, + "step": 1449 + }, + { + "epoch": 1.589041095890411, + "grad_norm": 2.8246665000915527, + "learning_rate": 2.6421608448415926e-06, + "loss": 1.4557, + "step": 1450 + }, + { + "epoch": 1.589041095890411, + "eval_loss": 1.7074187994003296, + "eval_runtime": 37.0881, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 1450 + }, + { + "epoch": 1.59013698630137, + "grad_norm": 2.8395724296569824, + "learning_rate": 2.6401299756295694e-06, + "loss": 1.4351, + "step": 1451 + }, + { + "epoch": 1.5912328767123287, + "grad_norm": 2.7422566413879395, + "learning_rate": 2.638099106417547e-06, + "loss": 1.4056, + "step": 1452 + }, + { + "epoch": 1.5923287671232877, + "grad_norm": 2.911670446395874, + "learning_rate": 2.636068237205524e-06, + "loss": 1.4522, + "step": 1453 + }, + { + "epoch": 1.5934246575342466, + "grad_norm": 2.8709356784820557, + "learning_rate": 2.6340373679935017e-06, + "loss": 1.4466, + "step": 1454 + }, + { + "epoch": 1.5945205479452054, + "grad_norm": 2.9503116607666016, + "learning_rate": 2.6320064987814785e-06, + "loss": 1.5055, + "step": 1455 + }, + { + "epoch": 1.5956164383561644, + "grad_norm": 2.752567768096924, + "learning_rate": 2.6299756295694562e-06, + "loss": 1.3711, + "step": 1456 + }, + { + "epoch": 1.5967123287671234, + "grad_norm": 2.907646656036377, + "learning_rate": 2.627944760357433e-06, + "loss": 1.4455, + "step": 1457 + }, + { + "epoch": 1.5978082191780822, + "grad_norm": 2.7967607975006104, + "learning_rate": 2.6259138911454104e-06, + "loss": 1.4972, + "step": 1458 + }, + { + "epoch": 1.598904109589041, + "grad_norm": 2.854980707168579, + "learning_rate": 2.6238830219333877e-06, + "loss": 1.4957, + "step": 1459 + }, + { + "epoch": 1.6, + "grad_norm": 2.8644754886627197, + "learning_rate": 2.621852152721365e-06, + "loss": 1.478, + "step": 1460 + }, + { + "epoch": 1.6, + "eval_loss": 1.706020712852478, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 1460 + }, + { + "epoch": 1.601095890410959, + "grad_norm": 2.7806358337402344, + "learning_rate": 2.619821283509342e-06, + "loss": 1.4027, + "step": 1461 + }, + { + "epoch": 1.6021917808219177, + "grad_norm": 2.829030990600586, + "learning_rate": 2.6177904142973195e-06, + "loss": 1.4583, + "step": 1462 + }, + { + "epoch": 1.6032876712328767, + "grad_norm": 2.8990888595581055, + "learning_rate": 2.6157595450852968e-06, + "loss": 1.4071, + "step": 1463 + }, + { + "epoch": 1.6043835616438358, + "grad_norm": 3.0500295162200928, + "learning_rate": 2.613728675873274e-06, + "loss": 1.4581, + "step": 1464 + }, + { + "epoch": 1.6054794520547944, + "grad_norm": 3.0562543869018555, + "learning_rate": 2.6116978066612513e-06, + "loss": 1.4953, + "step": 1465 + }, + { + "epoch": 1.6065753424657534, + "grad_norm": 2.992666721343994, + "learning_rate": 2.6096669374492286e-06, + "loss": 1.4938, + "step": 1466 + }, + { + "epoch": 1.6076712328767124, + "grad_norm": 2.895270586013794, + "learning_rate": 2.607636068237206e-06, + "loss": 1.4485, + "step": 1467 + }, + { + "epoch": 1.6087671232876712, + "grad_norm": 2.942186117172241, + "learning_rate": 2.605605199025183e-06, + "loss": 1.4303, + "step": 1468 + }, + { + "epoch": 1.60986301369863, + "grad_norm": 2.8370566368103027, + "learning_rate": 2.6035743298131604e-06, + "loss": 1.4619, + "step": 1469 + }, + { + "epoch": 1.610958904109589, + "grad_norm": 2.885728359222412, + "learning_rate": 2.6015434606011377e-06, + "loss": 1.449, + "step": 1470 + }, + { + "epoch": 1.610958904109589, + "eval_loss": 1.708180546760559, + "eval_runtime": 37.0887, + "eval_samples_per_second": 26.962, + "eval_steps_per_second": 1.132, + "step": 1470 + }, + { + "epoch": 1.612054794520548, + "grad_norm": 3.0368149280548096, + "learning_rate": 2.5995125913891145e-06, + "loss": 1.4799, + "step": 1471 + }, + { + "epoch": 1.6131506849315067, + "grad_norm": 3.136568546295166, + "learning_rate": 2.5974817221770922e-06, + "loss": 1.4801, + "step": 1472 + }, + { + "epoch": 1.6142465753424657, + "grad_norm": 2.9858787059783936, + "learning_rate": 2.595450852965069e-06, + "loss": 1.449, + "step": 1473 + }, + { + "epoch": 1.6153424657534248, + "grad_norm": 2.9646689891815186, + "learning_rate": 2.593419983753047e-06, + "loss": 1.4698, + "step": 1474 + }, + { + "epoch": 1.6164383561643836, + "grad_norm": 3.0527894496917725, + "learning_rate": 2.5913891145410236e-06, + "loss": 1.497, + "step": 1475 + }, + { + "epoch": 1.6175342465753424, + "grad_norm": 2.8184142112731934, + "learning_rate": 2.5893582453290013e-06, + "loss": 1.4421, + "step": 1476 + }, + { + "epoch": 1.6186301369863014, + "grad_norm": 2.7851176261901855, + "learning_rate": 2.587327376116978e-06, + "loss": 1.4666, + "step": 1477 + }, + { + "epoch": 1.6197260273972602, + "grad_norm": 2.7636115550994873, + "learning_rate": 2.585296506904956e-06, + "loss": 1.4368, + "step": 1478 + }, + { + "epoch": 1.620821917808219, + "grad_norm": 2.9213268756866455, + "learning_rate": 2.5832656376929327e-06, + "loss": 1.4408, + "step": 1479 + }, + { + "epoch": 1.621917808219178, + "grad_norm": 2.865529775619507, + "learning_rate": 2.5812347684809104e-06, + "loss": 1.4144, + "step": 1480 + }, + { + "epoch": 1.621917808219178, + "eval_loss": 1.7098593711853027, + "eval_runtime": 37.0613, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1480 + }, + { + "epoch": 1.6230136986301371, + "grad_norm": 2.840822458267212, + "learning_rate": 2.5792038992688873e-06, + "loss": 1.3843, + "step": 1481 + }, + { + "epoch": 1.624109589041096, + "grad_norm": 3.051424980163574, + "learning_rate": 2.577173030056865e-06, + "loss": 1.5298, + "step": 1482 + }, + { + "epoch": 1.6252054794520547, + "grad_norm": 2.995697498321533, + "learning_rate": 2.575142160844842e-06, + "loss": 1.3922, + "step": 1483 + }, + { + "epoch": 1.6263013698630138, + "grad_norm": 3.0062193870544434, + "learning_rate": 2.5731112916328187e-06, + "loss": 1.4704, + "step": 1484 + }, + { + "epoch": 1.6273972602739726, + "grad_norm": 2.9391777515411377, + "learning_rate": 2.5710804224207964e-06, + "loss": 1.481, + "step": 1485 + }, + { + "epoch": 1.6284931506849314, + "grad_norm": 2.9193878173828125, + "learning_rate": 2.5690495532087733e-06, + "loss": 1.4038, + "step": 1486 + }, + { + "epoch": 1.6295890410958904, + "grad_norm": 2.8676085472106934, + "learning_rate": 2.567018683996751e-06, + "loss": 1.3752, + "step": 1487 + }, + { + "epoch": 1.6306849315068495, + "grad_norm": 2.9337565898895264, + "learning_rate": 2.564987814784728e-06, + "loss": 1.4469, + "step": 1488 + }, + { + "epoch": 1.6317808219178083, + "grad_norm": 2.9485421180725098, + "learning_rate": 2.5629569455727055e-06, + "loss": 1.4224, + "step": 1489 + }, + { + "epoch": 1.632876712328767, + "grad_norm": 2.9687442779541016, + "learning_rate": 2.5609260763606824e-06, + "loss": 1.4347, + "step": 1490 + }, + { + "epoch": 1.632876712328767, + "eval_loss": 1.7070544958114624, + "eval_runtime": 37.3446, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 1490 + }, + { + "epoch": 1.633972602739726, + "grad_norm": 2.913801908493042, + "learning_rate": 2.55889520714866e-06, + "loss": 1.4478, + "step": 1491 + }, + { + "epoch": 1.635068493150685, + "grad_norm": 3.0276856422424316, + "learning_rate": 2.556864337936637e-06, + "loss": 1.5372, + "step": 1492 + }, + { + "epoch": 1.6361643835616437, + "grad_norm": 2.936163902282715, + "learning_rate": 2.5548334687246146e-06, + "loss": 1.4283, + "step": 1493 + }, + { + "epoch": 1.6372602739726028, + "grad_norm": 2.8936173915863037, + "learning_rate": 2.5528025995125915e-06, + "loss": 1.432, + "step": 1494 + }, + { + "epoch": 1.6383561643835618, + "grad_norm": 3.016876459121704, + "learning_rate": 2.550771730300569e-06, + "loss": 1.4966, + "step": 1495 + }, + { + "epoch": 1.6394520547945204, + "grad_norm": 2.8830440044403076, + "learning_rate": 2.548740861088546e-06, + "loss": 1.4408, + "step": 1496 + }, + { + "epoch": 1.6405479452054794, + "grad_norm": 2.9123940467834473, + "learning_rate": 2.5467099918765233e-06, + "loss": 1.4756, + "step": 1497 + }, + { + "epoch": 1.6416438356164385, + "grad_norm": 2.7645750045776367, + "learning_rate": 2.5446791226645006e-06, + "loss": 1.3705, + "step": 1498 + }, + { + "epoch": 1.6427397260273973, + "grad_norm": 2.8265559673309326, + "learning_rate": 2.542648253452478e-06, + "loss": 1.3738, + "step": 1499 + }, + { + "epoch": 1.643835616438356, + "grad_norm": 3.1146469116210938, + "learning_rate": 2.540617384240455e-06, + "loss": 1.4712, + "step": 1500 + }, + { + "epoch": 1.643835616438356, + "eval_loss": 1.714676022529602, + "eval_runtime": 37.4044, + "eval_samples_per_second": 26.735, + "eval_steps_per_second": 1.123, + "step": 1500 + }, + { + "epoch": 1.644931506849315, + "grad_norm": 3.0422415733337402, + "learning_rate": 2.5385865150284324e-06, + "loss": 1.4514, + "step": 1501 + }, + { + "epoch": 1.6460273972602741, + "grad_norm": 2.9049177169799805, + "learning_rate": 2.5365556458164097e-06, + "loss": 1.408, + "step": 1502 + }, + { + "epoch": 1.6471232876712327, + "grad_norm": 2.9806294441223145, + "learning_rate": 2.534524776604387e-06, + "loss": 1.4423, + "step": 1503 + }, + { + "epoch": 1.6482191780821918, + "grad_norm": 3.1372151374816895, + "learning_rate": 2.5324939073923642e-06, + "loss": 1.4368, + "step": 1504 + }, + { + "epoch": 1.6493150684931508, + "grad_norm": 2.778329849243164, + "learning_rate": 2.5304630381803415e-06, + "loss": 1.3559, + "step": 1505 + }, + { + "epoch": 1.6504109589041096, + "grad_norm": 2.935122013092041, + "learning_rate": 2.5284321689683188e-06, + "loss": 1.4299, + "step": 1506 + }, + { + "epoch": 1.6515068493150684, + "grad_norm": 2.8406238555908203, + "learning_rate": 2.526401299756296e-06, + "loss": 1.4496, + "step": 1507 + }, + { + "epoch": 1.6526027397260274, + "grad_norm": 2.919686794281006, + "learning_rate": 2.524370430544273e-06, + "loss": 1.446, + "step": 1508 + }, + { + "epoch": 1.6536986301369863, + "grad_norm": 2.8712282180786133, + "learning_rate": 2.5223395613322506e-06, + "loss": 1.4192, + "step": 1509 + }, + { + "epoch": 1.654794520547945, + "grad_norm": 2.984957695007324, + "learning_rate": 2.5203086921202274e-06, + "loss": 1.4234, + "step": 1510 + }, + { + "epoch": 1.654794520547945, + "eval_loss": 1.7065510749816895, + "eval_runtime": 36.8721, + "eval_samples_per_second": 27.121, + "eval_steps_per_second": 1.139, + "step": 1510 + }, + { + "epoch": 1.655890410958904, + "grad_norm": 3.0523569583892822, + "learning_rate": 2.518277822908205e-06, + "loss": 1.4488, + "step": 1511 + }, + { + "epoch": 1.6569863013698631, + "grad_norm": 3.1737515926361084, + "learning_rate": 2.516246953696182e-06, + "loss": 1.4311, + "step": 1512 + }, + { + "epoch": 1.658082191780822, + "grad_norm": 2.9754984378814697, + "learning_rate": 2.5142160844841597e-06, + "loss": 1.4363, + "step": 1513 + }, + { + "epoch": 1.6591780821917808, + "grad_norm": 3.0981898307800293, + "learning_rate": 2.5121852152721365e-06, + "loss": 1.4546, + "step": 1514 + }, + { + "epoch": 1.6602739726027398, + "grad_norm": 2.8706960678100586, + "learning_rate": 2.5101543460601142e-06, + "loss": 1.4599, + "step": 1515 + }, + { + "epoch": 1.6613698630136986, + "grad_norm": 2.9565646648406982, + "learning_rate": 2.508123476848091e-06, + "loss": 1.4471, + "step": 1516 + }, + { + "epoch": 1.6624657534246574, + "grad_norm": 2.687807083129883, + "learning_rate": 2.506092607636069e-06, + "loss": 1.3926, + "step": 1517 + }, + { + "epoch": 1.6635616438356164, + "grad_norm": 2.832486629486084, + "learning_rate": 2.5040617384240457e-06, + "loss": 1.4646, + "step": 1518 + }, + { + "epoch": 1.6646575342465755, + "grad_norm": 2.946368455886841, + "learning_rate": 2.5020308692120234e-06, + "loss": 1.4347, + "step": 1519 + }, + { + "epoch": 1.6657534246575343, + "grad_norm": 2.7646501064300537, + "learning_rate": 2.5e-06, + "loss": 1.4214, + "step": 1520 + }, + { + "epoch": 1.6657534246575343, + "eval_loss": 1.7065269947052002, + "eval_runtime": 36.9819, + "eval_samples_per_second": 27.04, + "eval_steps_per_second": 1.136, + "step": 1520 + }, + { + "epoch": 1.666849315068493, + "grad_norm": 2.995950698852539, + "learning_rate": 2.4979691307879775e-06, + "loss": 1.4986, + "step": 1521 + }, + { + "epoch": 1.6679452054794521, + "grad_norm": 2.886826515197754, + "learning_rate": 2.4959382615759548e-06, + "loss": 1.4319, + "step": 1522 + }, + { + "epoch": 1.669041095890411, + "grad_norm": 3.0105748176574707, + "learning_rate": 2.493907392363932e-06, + "loss": 1.4297, + "step": 1523 + }, + { + "epoch": 1.6701369863013698, + "grad_norm": 2.892512083053589, + "learning_rate": 2.4918765231519093e-06, + "loss": 1.4857, + "step": 1524 + }, + { + "epoch": 1.6712328767123288, + "grad_norm": 2.8092124462127686, + "learning_rate": 2.4898456539398866e-06, + "loss": 1.4304, + "step": 1525 + }, + { + "epoch": 1.6723287671232878, + "grad_norm": 2.8145992755889893, + "learning_rate": 2.487814784727864e-06, + "loss": 1.4559, + "step": 1526 + }, + { + "epoch": 1.6734246575342466, + "grad_norm": 2.894888401031494, + "learning_rate": 2.4857839155158407e-06, + "loss": 1.497, + "step": 1527 + }, + { + "epoch": 1.6745205479452054, + "grad_norm": 2.8759849071502686, + "learning_rate": 2.483753046303818e-06, + "loss": 1.4295, + "step": 1528 + }, + { + "epoch": 1.6756164383561645, + "grad_norm": 2.8624649047851562, + "learning_rate": 2.4817221770917953e-06, + "loss": 1.4369, + "step": 1529 + }, + { + "epoch": 1.6767123287671233, + "grad_norm": 2.860114574432373, + "learning_rate": 2.4796913078797725e-06, + "loss": 1.4389, + "step": 1530 + }, + { + "epoch": 1.6767123287671233, + "eval_loss": 1.708011269569397, + "eval_runtime": 37.0257, + "eval_samples_per_second": 27.008, + "eval_steps_per_second": 1.134, + "step": 1530 + }, + { + "epoch": 1.677808219178082, + "grad_norm": 3.062593460083008, + "learning_rate": 2.47766043866775e-06, + "loss": 1.4679, + "step": 1531 + }, + { + "epoch": 1.6789041095890411, + "grad_norm": 2.9316751956939697, + "learning_rate": 2.475629569455727e-06, + "loss": 1.4712, + "step": 1532 + }, + { + "epoch": 1.6800000000000002, + "grad_norm": 2.799980401992798, + "learning_rate": 2.4735987002437044e-06, + "loss": 1.4764, + "step": 1533 + }, + { + "epoch": 1.6810958904109587, + "grad_norm": 2.71172833442688, + "learning_rate": 2.4715678310316816e-06, + "loss": 1.4103, + "step": 1534 + }, + { + "epoch": 1.6821917808219178, + "grad_norm": 2.808029890060425, + "learning_rate": 2.469536961819659e-06, + "loss": 1.4134, + "step": 1535 + }, + { + "epoch": 1.6832876712328768, + "grad_norm": 2.8933329582214355, + "learning_rate": 2.467506092607636e-06, + "loss": 1.4301, + "step": 1536 + }, + { + "epoch": 1.6843835616438356, + "grad_norm": 2.866718053817749, + "learning_rate": 2.4654752233956135e-06, + "loss": 1.4438, + "step": 1537 + }, + { + "epoch": 1.6854794520547944, + "grad_norm": 2.8831021785736084, + "learning_rate": 2.4634443541835907e-06, + "loss": 1.4182, + "step": 1538 + }, + { + "epoch": 1.6865753424657535, + "grad_norm": 3.140631914138794, + "learning_rate": 2.461413484971568e-06, + "loss": 1.5008, + "step": 1539 + }, + { + "epoch": 1.6876712328767123, + "grad_norm": 3.0420989990234375, + "learning_rate": 2.4593826157595453e-06, + "loss": 1.4454, + "step": 1540 + }, + { + "epoch": 1.6876712328767123, + "eval_loss": 1.7109777927398682, + "eval_runtime": 37.0355, + "eval_samples_per_second": 27.001, + "eval_steps_per_second": 1.134, + "step": 1540 + }, + { + "epoch": 1.688767123287671, + "grad_norm": 3.0079169273376465, + "learning_rate": 2.4573517465475226e-06, + "loss": 1.5109, + "step": 1541 + }, + { + "epoch": 1.6898630136986301, + "grad_norm": 2.846114158630371, + "learning_rate": 2.4553208773355e-06, + "loss": 1.4035, + "step": 1542 + }, + { + "epoch": 1.6909589041095892, + "grad_norm": 2.828695297241211, + "learning_rate": 2.453290008123477e-06, + "loss": 1.4166, + "step": 1543 + }, + { + "epoch": 1.692054794520548, + "grad_norm": 2.824759006500244, + "learning_rate": 2.4512591389114544e-06, + "loss": 1.4421, + "step": 1544 + }, + { + "epoch": 1.6931506849315068, + "grad_norm": 2.8216183185577393, + "learning_rate": 2.4492282696994317e-06, + "loss": 1.4654, + "step": 1545 + }, + { + "epoch": 1.6942465753424658, + "grad_norm": 2.879293441772461, + "learning_rate": 2.447197400487409e-06, + "loss": 1.4511, + "step": 1546 + }, + { + "epoch": 1.6953424657534246, + "grad_norm": 2.98406720161438, + "learning_rate": 2.4451665312753862e-06, + "loss": 1.4164, + "step": 1547 + }, + { + "epoch": 1.6964383561643834, + "grad_norm": 3.0683135986328125, + "learning_rate": 2.4431356620633635e-06, + "loss": 1.4563, + "step": 1548 + }, + { + "epoch": 1.6975342465753425, + "grad_norm": 2.986192464828491, + "learning_rate": 2.4411047928513408e-06, + "loss": 1.4643, + "step": 1549 + }, + { + "epoch": 1.6986301369863015, + "grad_norm": 2.923180341720581, + "learning_rate": 2.439073923639318e-06, + "loss": 1.4546, + "step": 1550 + }, + { + "epoch": 1.6986301369863015, + "eval_loss": 1.7101587057113647, + "eval_runtime": 37.0292, + "eval_samples_per_second": 27.006, + "eval_steps_per_second": 1.134, + "step": 1550 + }, + { + "epoch": 1.6997260273972603, + "grad_norm": 2.928861141204834, + "learning_rate": 2.4370430544272953e-06, + "loss": 1.4424, + "step": 1551 + }, + { + "epoch": 1.7008219178082191, + "grad_norm": 2.8597733974456787, + "learning_rate": 2.435012185215272e-06, + "loss": 1.3872, + "step": 1552 + }, + { + "epoch": 1.7019178082191782, + "grad_norm": 3.054025650024414, + "learning_rate": 2.4329813160032495e-06, + "loss": 1.3912, + "step": 1553 + }, + { + "epoch": 1.703013698630137, + "grad_norm": 2.824483871459961, + "learning_rate": 2.4309504467912267e-06, + "loss": 1.4167, + "step": 1554 + }, + { + "epoch": 1.7041095890410958, + "grad_norm": 2.9558396339416504, + "learning_rate": 2.428919577579204e-06, + "loss": 1.4436, + "step": 1555 + }, + { + "epoch": 1.7052054794520548, + "grad_norm": 2.8712751865386963, + "learning_rate": 2.4268887083671813e-06, + "loss": 1.4198, + "step": 1556 + }, + { + "epoch": 1.7063013698630138, + "grad_norm": 2.9776816368103027, + "learning_rate": 2.4248578391551586e-06, + "loss": 1.5133, + "step": 1557 + }, + { + "epoch": 1.7073972602739727, + "grad_norm": 2.930847644805908, + "learning_rate": 2.422826969943136e-06, + "loss": 1.4239, + "step": 1558 + }, + { + "epoch": 1.7084931506849315, + "grad_norm": 2.9059464931488037, + "learning_rate": 2.420796100731113e-06, + "loss": 1.4071, + "step": 1559 + }, + { + "epoch": 1.7095890410958905, + "grad_norm": 2.8655083179473877, + "learning_rate": 2.4187652315190904e-06, + "loss": 1.3911, + "step": 1560 + }, + { + "epoch": 1.7095890410958905, + "eval_loss": 1.705108880996704, + "eval_runtime": 37.0068, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 1560 + }, + { + "epoch": 1.7106849315068493, + "grad_norm": 2.883193016052246, + "learning_rate": 2.4167343623070677e-06, + "loss": 1.444, + "step": 1561 + }, + { + "epoch": 1.7117808219178081, + "grad_norm": 2.8185808658599854, + "learning_rate": 2.414703493095045e-06, + "loss": 1.4444, + "step": 1562 + }, + { + "epoch": 1.7128767123287671, + "grad_norm": 2.8110158443450928, + "learning_rate": 2.4126726238830222e-06, + "loss": 1.4211, + "step": 1563 + }, + { + "epoch": 1.7139726027397262, + "grad_norm": 2.863987445831299, + "learning_rate": 2.410641754670999e-06, + "loss": 1.4319, + "step": 1564 + }, + { + "epoch": 1.7150684931506848, + "grad_norm": 2.977844715118408, + "learning_rate": 2.4086108854589763e-06, + "loss": 1.4112, + "step": 1565 + }, + { + "epoch": 1.7161643835616438, + "grad_norm": 2.8599488735198975, + "learning_rate": 2.4065800162469536e-06, + "loss": 1.4657, + "step": 1566 + }, + { + "epoch": 1.7172602739726028, + "grad_norm": 2.8669981956481934, + "learning_rate": 2.404549147034931e-06, + "loss": 1.4214, + "step": 1567 + }, + { + "epoch": 1.7183561643835616, + "grad_norm": 2.980116367340088, + "learning_rate": 2.402518277822908e-06, + "loss": 1.4189, + "step": 1568 + }, + { + "epoch": 1.7194520547945205, + "grad_norm": 2.936729907989502, + "learning_rate": 2.4004874086108854e-06, + "loss": 1.4689, + "step": 1569 + }, + { + "epoch": 1.7205479452054795, + "grad_norm": 2.8469579219818115, + "learning_rate": 2.3984565393988627e-06, + "loss": 1.4137, + "step": 1570 + }, + { + "epoch": 1.7205479452054795, + "eval_loss": 1.7018489837646484, + "eval_runtime": 37.0142, + "eval_samples_per_second": 27.017, + "eval_steps_per_second": 1.135, + "step": 1570 + }, + { + "epoch": 1.7216438356164385, + "grad_norm": 2.9605817794799805, + "learning_rate": 2.39642567018684e-06, + "loss": 1.4449, + "step": 1571 + }, + { + "epoch": 1.7227397260273971, + "grad_norm": 2.815717935562134, + "learning_rate": 2.3943948009748173e-06, + "loss": 1.4114, + "step": 1572 + }, + { + "epoch": 1.7238356164383561, + "grad_norm": 2.961744785308838, + "learning_rate": 2.3923639317627945e-06, + "loss": 1.4306, + "step": 1573 + }, + { + "epoch": 1.7249315068493152, + "grad_norm": 3.022645950317383, + "learning_rate": 2.390333062550772e-06, + "loss": 1.4619, + "step": 1574 + }, + { + "epoch": 1.726027397260274, + "grad_norm": 2.9433681964874268, + "learning_rate": 2.388302193338749e-06, + "loss": 1.401, + "step": 1575 + }, + { + "epoch": 1.7271232876712328, + "grad_norm": 2.794745683670044, + "learning_rate": 2.3862713241267264e-06, + "loss": 1.3441, + "step": 1576 + }, + { + "epoch": 1.7282191780821918, + "grad_norm": 3.041825294494629, + "learning_rate": 2.3842404549147037e-06, + "loss": 1.4938, + "step": 1577 + }, + { + "epoch": 1.7293150684931506, + "grad_norm": 2.9499473571777344, + "learning_rate": 2.382209585702681e-06, + "loss": 1.4295, + "step": 1578 + }, + { + "epoch": 1.7304109589041095, + "grad_norm": 2.8626675605773926, + "learning_rate": 2.380178716490658e-06, + "loss": 1.3911, + "step": 1579 + }, + { + "epoch": 1.7315068493150685, + "grad_norm": 2.842794179916382, + "learning_rate": 2.3781478472786355e-06, + "loss": 1.4737, + "step": 1580 + }, + { + "epoch": 1.7315068493150685, + "eval_loss": 1.7021467685699463, + "eval_runtime": 37.1037, + "eval_samples_per_second": 26.952, + "eval_steps_per_second": 1.132, + "step": 1580 + }, + { + "epoch": 1.7326027397260275, + "grad_norm": 2.8970463275909424, + "learning_rate": 2.3761169780666128e-06, + "loss": 1.4209, + "step": 1581 + }, + { + "epoch": 1.7336986301369863, + "grad_norm": 2.959874153137207, + "learning_rate": 2.37408610885459e-06, + "loss": 1.4662, + "step": 1582 + }, + { + "epoch": 1.7347945205479451, + "grad_norm": 3.0744926929473877, + "learning_rate": 2.3720552396425673e-06, + "loss": 1.4655, + "step": 1583 + }, + { + "epoch": 1.7358904109589042, + "grad_norm": 2.882037878036499, + "learning_rate": 2.3700243704305446e-06, + "loss": 1.4237, + "step": 1584 + }, + { + "epoch": 1.736986301369863, + "grad_norm": 2.956204652786255, + "learning_rate": 2.367993501218522e-06, + "loss": 1.4908, + "step": 1585 + }, + { + "epoch": 1.7380821917808218, + "grad_norm": 2.9728269577026367, + "learning_rate": 2.365962632006499e-06, + "loss": 1.4557, + "step": 1586 + }, + { + "epoch": 1.7391780821917808, + "grad_norm": 2.9332668781280518, + "learning_rate": 2.3639317627944764e-06, + "loss": 1.4286, + "step": 1587 + }, + { + "epoch": 1.7402739726027399, + "grad_norm": 2.8205621242523193, + "learning_rate": 2.3619008935824537e-06, + "loss": 1.4584, + "step": 1588 + }, + { + "epoch": 1.7413698630136987, + "grad_norm": 2.9190478324890137, + "learning_rate": 2.3598700243704305e-06, + "loss": 1.4657, + "step": 1589 + }, + { + "epoch": 1.7424657534246575, + "grad_norm": 2.8216135501861572, + "learning_rate": 2.357839155158408e-06, + "loss": 1.4655, + "step": 1590 + }, + { + "epoch": 1.7424657534246575, + "eval_loss": 1.7007001638412476, + "eval_runtime": 37.3826, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.124, + "step": 1590 + }, + { + "epoch": 1.7435616438356165, + "grad_norm": 2.827335834503174, + "learning_rate": 2.355808285946385e-06, + "loss": 1.4604, + "step": 1591 + }, + { + "epoch": 1.7446575342465753, + "grad_norm": 2.808551549911499, + "learning_rate": 2.3537774167343624e-06, + "loss": 1.448, + "step": 1592 + }, + { + "epoch": 1.7457534246575341, + "grad_norm": 2.958263635635376, + "learning_rate": 2.3517465475223396e-06, + "loss": 1.4612, + "step": 1593 + }, + { + "epoch": 1.7468493150684932, + "grad_norm": 2.900951623916626, + "learning_rate": 2.349715678310317e-06, + "loss": 1.4572, + "step": 1594 + }, + { + "epoch": 1.7479452054794522, + "grad_norm": 2.933318614959717, + "learning_rate": 2.347684809098294e-06, + "loss": 1.4439, + "step": 1595 + }, + { + "epoch": 1.749041095890411, + "grad_norm": 2.929802417755127, + "learning_rate": 2.3456539398862715e-06, + "loss": 1.4331, + "step": 1596 + }, + { + "epoch": 1.7501369863013698, + "grad_norm": 3.167396068572998, + "learning_rate": 2.3436230706742487e-06, + "loss": 1.4792, + "step": 1597 + }, + { + "epoch": 1.7512328767123289, + "grad_norm": 2.901446580886841, + "learning_rate": 2.341592201462226e-06, + "loss": 1.4143, + "step": 1598 + }, + { + "epoch": 1.7523287671232877, + "grad_norm": 2.976067066192627, + "learning_rate": 2.3395613322502033e-06, + "loss": 1.41, + "step": 1599 + }, + { + "epoch": 1.7534246575342465, + "grad_norm": 2.8032453060150146, + "learning_rate": 2.3375304630381806e-06, + "loss": 1.4304, + "step": 1600 + }, + { + "epoch": 1.7534246575342465, + "eval_loss": 1.700425386428833, + "eval_runtime": 37.4655, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 1600 + }, + { + "epoch": 1.7545205479452055, + "grad_norm": 2.9107625484466553, + "learning_rate": 2.335499593826158e-06, + "loss": 1.3964, + "step": 1601 + }, + { + "epoch": 1.7556164383561645, + "grad_norm": 2.7919020652770996, + "learning_rate": 2.333468724614135e-06, + "loss": 1.4178, + "step": 1602 + }, + { + "epoch": 1.7567123287671231, + "grad_norm": 2.958300828933716, + "learning_rate": 2.3314378554021124e-06, + "loss": 1.4646, + "step": 1603 + }, + { + "epoch": 1.7578082191780822, + "grad_norm": 2.8957359790802, + "learning_rate": 2.3294069861900897e-06, + "loss": 1.4565, + "step": 1604 + }, + { + "epoch": 1.7589041095890412, + "grad_norm": 2.8454794883728027, + "learning_rate": 2.327376116978067e-06, + "loss": 1.3941, + "step": 1605 + }, + { + "epoch": 1.76, + "grad_norm": 2.8161509037017822, + "learning_rate": 2.3253452477660442e-06, + "loss": 1.4012, + "step": 1606 + }, + { + "epoch": 1.7610958904109588, + "grad_norm": 3.020136594772339, + "learning_rate": 2.3233143785540215e-06, + "loss": 1.4629, + "step": 1607 + }, + { + "epoch": 1.7621917808219179, + "grad_norm": 2.901064395904541, + "learning_rate": 2.3212835093419988e-06, + "loss": 1.413, + "step": 1608 + }, + { + "epoch": 1.7632876712328767, + "grad_norm": 3.066730499267578, + "learning_rate": 2.319252640129976e-06, + "loss": 1.4513, + "step": 1609 + }, + { + "epoch": 1.7643835616438355, + "grad_norm": 2.859468460083008, + "learning_rate": 2.3172217709179533e-06, + "loss": 1.4024, + "step": 1610 + }, + { + "epoch": 1.7643835616438355, + "eval_loss": 1.704864501953125, + "eval_runtime": 36.9687, + "eval_samples_per_second": 27.05, + "eval_steps_per_second": 1.136, + "step": 1610 + }, + { + "epoch": 1.7654794520547945, + "grad_norm": 2.990548610687256, + "learning_rate": 2.3151909017059306e-06, + "loss": 1.5025, + "step": 1611 + }, + { + "epoch": 1.7665753424657535, + "grad_norm": 2.9470932483673096, + "learning_rate": 2.313160032493908e-06, + "loss": 1.4658, + "step": 1612 + }, + { + "epoch": 1.7676712328767124, + "grad_norm": 2.8431284427642822, + "learning_rate": 2.311129163281885e-06, + "loss": 1.3847, + "step": 1613 + }, + { + "epoch": 1.7687671232876712, + "grad_norm": 2.864870309829712, + "learning_rate": 2.309098294069862e-06, + "loss": 1.451, + "step": 1614 + }, + { + "epoch": 1.7698630136986302, + "grad_norm": 2.843156099319458, + "learning_rate": 2.3070674248578393e-06, + "loss": 1.4151, + "step": 1615 + }, + { + "epoch": 1.770958904109589, + "grad_norm": 2.9214529991149902, + "learning_rate": 2.3050365556458166e-06, + "loss": 1.3913, + "step": 1616 + }, + { + "epoch": 1.7720547945205478, + "grad_norm": 2.880232572555542, + "learning_rate": 2.303005686433794e-06, + "loss": 1.4301, + "step": 1617 + }, + { + "epoch": 1.7731506849315068, + "grad_norm": 2.8919169902801514, + "learning_rate": 2.300974817221771e-06, + "loss": 1.4661, + "step": 1618 + }, + { + "epoch": 1.7742465753424659, + "grad_norm": 2.950577974319458, + "learning_rate": 2.2989439480097484e-06, + "loss": 1.4316, + "step": 1619 + }, + { + "epoch": 1.7753424657534247, + "grad_norm": 2.8465826511383057, + "learning_rate": 2.2969130787977257e-06, + "loss": 1.3856, + "step": 1620 + }, + { + "epoch": 1.7753424657534247, + "eval_loss": 1.704519510269165, + "eval_runtime": 37.0769, + "eval_samples_per_second": 26.971, + "eval_steps_per_second": 1.133, + "step": 1620 + }, + { + "epoch": 1.7764383561643835, + "grad_norm": 2.91975474357605, + "learning_rate": 2.294882209585703e-06, + "loss": 1.4478, + "step": 1621 + }, + { + "epoch": 1.7775342465753425, + "grad_norm": 3.064791440963745, + "learning_rate": 2.2928513403736802e-06, + "loss": 1.4657, + "step": 1622 + }, + { + "epoch": 1.7786301369863013, + "grad_norm": 2.9099392890930176, + "learning_rate": 2.2908204711616575e-06, + "loss": 1.4162, + "step": 1623 + }, + { + "epoch": 1.7797260273972602, + "grad_norm": 2.9778053760528564, + "learning_rate": 2.2887896019496348e-06, + "loss": 1.4423, + "step": 1624 + }, + { + "epoch": 1.7808219178082192, + "grad_norm": 3.070389747619629, + "learning_rate": 2.286758732737612e-06, + "loss": 1.4845, + "step": 1625 + }, + { + "epoch": 1.7819178082191782, + "grad_norm": 2.860940456390381, + "learning_rate": 2.284727863525589e-06, + "loss": 1.4305, + "step": 1626 + }, + { + "epoch": 1.783013698630137, + "grad_norm": 2.9022462368011475, + "learning_rate": 2.282696994313566e-06, + "loss": 1.4282, + "step": 1627 + }, + { + "epoch": 1.7841095890410958, + "grad_norm": 2.899451971054077, + "learning_rate": 2.2806661251015434e-06, + "loss": 1.4595, + "step": 1628 + }, + { + "epoch": 1.7852054794520549, + "grad_norm": 2.748985528945923, + "learning_rate": 2.2786352558895207e-06, + "loss": 1.4368, + "step": 1629 + }, + { + "epoch": 1.7863013698630137, + "grad_norm": 2.7873692512512207, + "learning_rate": 2.276604386677498e-06, + "loss": 1.4207, + "step": 1630 + }, + { + "epoch": 1.7863013698630137, + "eval_loss": 1.7050743103027344, + "eval_runtime": 37.1442, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 1630 + }, + { + "epoch": 1.7873972602739725, + "grad_norm": 2.894057512283325, + "learning_rate": 2.2745735174654753e-06, + "loss": 1.3917, + "step": 1631 + }, + { + "epoch": 1.7884931506849315, + "grad_norm": 2.931969404220581, + "learning_rate": 2.2725426482534526e-06, + "loss": 1.4056, + "step": 1632 + }, + { + "epoch": 1.7895890410958906, + "grad_norm": 3.027864694595337, + "learning_rate": 2.27051177904143e-06, + "loss": 1.4772, + "step": 1633 + }, + { + "epoch": 1.7906849315068492, + "grad_norm": 2.8892524242401123, + "learning_rate": 2.268480909829407e-06, + "loss": 1.4464, + "step": 1634 + }, + { + "epoch": 1.7917808219178082, + "grad_norm": 3.1151480674743652, + "learning_rate": 2.2664500406173844e-06, + "loss": 1.4244, + "step": 1635 + }, + { + "epoch": 1.7928767123287672, + "grad_norm": 2.991565227508545, + "learning_rate": 2.2644191714053617e-06, + "loss": 1.4498, + "step": 1636 + }, + { + "epoch": 1.793972602739726, + "grad_norm": 3.0030806064605713, + "learning_rate": 2.262388302193339e-06, + "loss": 1.4723, + "step": 1637 + }, + { + "epoch": 1.7950684931506848, + "grad_norm": 2.8650996685028076, + "learning_rate": 2.260357432981316e-06, + "loss": 1.4313, + "step": 1638 + }, + { + "epoch": 1.7961643835616439, + "grad_norm": 2.7710604667663574, + "learning_rate": 2.2583265637692935e-06, + "loss": 1.4179, + "step": 1639 + }, + { + "epoch": 1.7972602739726027, + "grad_norm": 2.8966064453125, + "learning_rate": 2.2562956945572708e-06, + "loss": 1.4394, + "step": 1640 + }, + { + "epoch": 1.7972602739726027, + "eval_loss": 1.6987948417663574, + "eval_runtime": 37.1628, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1640 + }, + { + "epoch": 1.7983561643835615, + "grad_norm": 2.816455841064453, + "learning_rate": 2.254264825345248e-06, + "loss": 1.4477, + "step": 1641 + }, + { + "epoch": 1.7994520547945205, + "grad_norm": 2.8696861267089844, + "learning_rate": 2.2522339561332253e-06, + "loss": 1.4526, + "step": 1642 + }, + { + "epoch": 1.8005479452054796, + "grad_norm": 2.9743974208831787, + "learning_rate": 2.2502030869212026e-06, + "loss": 1.4408, + "step": 1643 + }, + { + "epoch": 1.8016438356164384, + "grad_norm": 2.8802881240844727, + "learning_rate": 2.24817221770918e-06, + "loss": 1.4448, + "step": 1644 + }, + { + "epoch": 1.8027397260273972, + "grad_norm": 3.029087781906128, + "learning_rate": 2.246141348497157e-06, + "loss": 1.4702, + "step": 1645 + }, + { + "epoch": 1.8038356164383562, + "grad_norm": 2.930821180343628, + "learning_rate": 2.2441104792851344e-06, + "loss": 1.4555, + "step": 1646 + }, + { + "epoch": 1.804931506849315, + "grad_norm": 2.888127326965332, + "learning_rate": 2.2420796100731117e-06, + "loss": 1.4297, + "step": 1647 + }, + { + "epoch": 1.8060273972602738, + "grad_norm": 2.8290679454803467, + "learning_rate": 2.240048740861089e-06, + "loss": 1.4313, + "step": 1648 + }, + { + "epoch": 1.8071232876712329, + "grad_norm": 2.8667287826538086, + "learning_rate": 2.2380178716490662e-06, + "loss": 1.4663, + "step": 1649 + }, + { + "epoch": 1.808219178082192, + "grad_norm": 2.8982157707214355, + "learning_rate": 2.2359870024370435e-06, + "loss": 1.3463, + "step": 1650 + }, + { + "epoch": 1.808219178082192, + "eval_loss": 1.698309302330017, + "eval_runtime": 37.1649, + "eval_samples_per_second": 26.907, + "eval_steps_per_second": 1.13, + "step": 1650 + }, + { + "epoch": 1.8093150684931507, + "grad_norm": 2.8749091625213623, + "learning_rate": 2.2339561332250204e-06, + "loss": 1.4489, + "step": 1651 + }, + { + "epoch": 1.8104109589041095, + "grad_norm": 2.9321630001068115, + "learning_rate": 2.2319252640129976e-06, + "loss": 1.4356, + "step": 1652 + }, + { + "epoch": 1.8115068493150686, + "grad_norm": 3.039285898208618, + "learning_rate": 2.229894394800975e-06, + "loss": 1.4332, + "step": 1653 + }, + { + "epoch": 1.8126027397260274, + "grad_norm": 2.9201481342315674, + "learning_rate": 2.227863525588952e-06, + "loss": 1.3828, + "step": 1654 + }, + { + "epoch": 1.8136986301369862, + "grad_norm": 2.8183014392852783, + "learning_rate": 2.2258326563769295e-06, + "loss": 1.3748, + "step": 1655 + }, + { + "epoch": 1.8147945205479452, + "grad_norm": 2.7358760833740234, + "learning_rate": 2.2238017871649067e-06, + "loss": 1.4064, + "step": 1656 + }, + { + "epoch": 1.8158904109589042, + "grad_norm": 2.8340682983398438, + "learning_rate": 2.221770917952884e-06, + "loss": 1.3918, + "step": 1657 + }, + { + "epoch": 1.816986301369863, + "grad_norm": 2.800960063934326, + "learning_rate": 2.2197400487408613e-06, + "loss": 1.4312, + "step": 1658 + }, + { + "epoch": 1.8180821917808219, + "grad_norm": 2.986204147338867, + "learning_rate": 2.2177091795288386e-06, + "loss": 1.457, + "step": 1659 + }, + { + "epoch": 1.819178082191781, + "grad_norm": 2.8145415782928467, + "learning_rate": 2.215678310316816e-06, + "loss": 1.4067, + "step": 1660 + }, + { + "epoch": 1.819178082191781, + "eval_loss": 1.7052077054977417, + "eval_runtime": 37.16, + "eval_samples_per_second": 26.911, + "eval_steps_per_second": 1.13, + "step": 1660 + }, + { + "epoch": 1.8202739726027397, + "grad_norm": 2.936354637145996, + "learning_rate": 2.213647441104793e-06, + "loss": 1.4159, + "step": 1661 + }, + { + "epoch": 1.8213698630136985, + "grad_norm": 2.9661855697631836, + "learning_rate": 2.2116165718927704e-06, + "loss": 1.4532, + "step": 1662 + }, + { + "epoch": 1.8224657534246576, + "grad_norm": 2.943347454071045, + "learning_rate": 2.2095857026807473e-06, + "loss": 1.4179, + "step": 1663 + }, + { + "epoch": 1.8235616438356166, + "grad_norm": 3.0526676177978516, + "learning_rate": 2.2075548334687245e-06, + "loss": 1.4274, + "step": 1664 + }, + { + "epoch": 1.8246575342465754, + "grad_norm": 2.943145751953125, + "learning_rate": 2.205523964256702e-06, + "loss": 1.4012, + "step": 1665 + }, + { + "epoch": 1.8257534246575342, + "grad_norm": 2.9686639308929443, + "learning_rate": 2.203493095044679e-06, + "loss": 1.4283, + "step": 1666 + }, + { + "epoch": 1.8268493150684932, + "grad_norm": 2.951856851577759, + "learning_rate": 2.2014622258326564e-06, + "loss": 1.4493, + "step": 1667 + }, + { + "epoch": 1.827945205479452, + "grad_norm": 2.9208593368530273, + "learning_rate": 2.1994313566206336e-06, + "loss": 1.4452, + "step": 1668 + }, + { + "epoch": 1.8290410958904109, + "grad_norm": 2.9517624378204346, + "learning_rate": 2.197400487408611e-06, + "loss": 1.4529, + "step": 1669 + }, + { + "epoch": 1.83013698630137, + "grad_norm": 2.9841079711914062, + "learning_rate": 2.195369618196588e-06, + "loss": 1.431, + "step": 1670 + }, + { + "epoch": 1.83013698630137, + "eval_loss": 1.6963521242141724, + "eval_runtime": 37.1536, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 1670 + }, + { + "epoch": 1.831232876712329, + "grad_norm": 2.9343044757843018, + "learning_rate": 2.1933387489845655e-06, + "loss": 1.3912, + "step": 1671 + }, + { + "epoch": 1.8323287671232875, + "grad_norm": 2.9684195518493652, + "learning_rate": 2.1913078797725427e-06, + "loss": 1.4222, + "step": 1672 + }, + { + "epoch": 1.8334246575342465, + "grad_norm": 2.929227113723755, + "learning_rate": 2.18927701056052e-06, + "loss": 1.4535, + "step": 1673 + }, + { + "epoch": 1.8345205479452056, + "grad_norm": 3.0814716815948486, + "learning_rate": 2.1872461413484973e-06, + "loss": 1.4984, + "step": 1674 + }, + { + "epoch": 1.8356164383561644, + "grad_norm": 3.0480618476867676, + "learning_rate": 2.1852152721364746e-06, + "loss": 1.4211, + "step": 1675 + }, + { + "epoch": 1.8367123287671232, + "grad_norm": 2.9622223377227783, + "learning_rate": 2.183184402924452e-06, + "loss": 1.4252, + "step": 1676 + }, + { + "epoch": 1.8378082191780822, + "grad_norm": 3.054903745651245, + "learning_rate": 2.181153533712429e-06, + "loss": 1.4253, + "step": 1677 + }, + { + "epoch": 1.838904109589041, + "grad_norm": 2.9990928173065186, + "learning_rate": 2.1791226645004064e-06, + "loss": 1.4515, + "step": 1678 + }, + { + "epoch": 1.8399999999999999, + "grad_norm": 2.914088249206543, + "learning_rate": 2.1770917952883837e-06, + "loss": 1.438, + "step": 1679 + }, + { + "epoch": 1.841095890410959, + "grad_norm": 2.874981641769409, + "learning_rate": 2.175060926076361e-06, + "loss": 1.4247, + "step": 1680 + }, + { + "epoch": 1.841095890410959, + "eval_loss": 1.6927348375320435, + "eval_runtime": 37.1382, + "eval_samples_per_second": 26.926, + "eval_steps_per_second": 1.131, + "step": 1680 + }, + { + "epoch": 1.842191780821918, + "grad_norm": 2.954643726348877, + "learning_rate": 2.1730300568643382e-06, + "loss": 1.4892, + "step": 1681 + }, + { + "epoch": 1.8432876712328767, + "grad_norm": 2.9708850383758545, + "learning_rate": 2.1709991876523155e-06, + "loss": 1.4884, + "step": 1682 + }, + { + "epoch": 1.8443835616438355, + "grad_norm": 2.8535971641540527, + "learning_rate": 2.1689683184402928e-06, + "loss": 1.4231, + "step": 1683 + }, + { + "epoch": 1.8454794520547946, + "grad_norm": 2.9040937423706055, + "learning_rate": 2.16693744922827e-06, + "loss": 1.3577, + "step": 1684 + }, + { + "epoch": 1.8465753424657534, + "grad_norm": 2.9993464946746826, + "learning_rate": 2.1649065800162473e-06, + "loss": 1.4381, + "step": 1685 + }, + { + "epoch": 1.8476712328767122, + "grad_norm": 2.842670202255249, + "learning_rate": 2.1628757108042246e-06, + "loss": 1.4023, + "step": 1686 + }, + { + "epoch": 1.8487671232876712, + "grad_norm": 2.9607810974121094, + "learning_rate": 2.160844841592202e-06, + "loss": 1.4179, + "step": 1687 + }, + { + "epoch": 1.8498630136986303, + "grad_norm": 2.862751007080078, + "learning_rate": 2.1588139723801787e-06, + "loss": 1.4048, + "step": 1688 + }, + { + "epoch": 1.850958904109589, + "grad_norm": 2.910599708557129, + "learning_rate": 2.156783103168156e-06, + "loss": 1.4188, + "step": 1689 + }, + { + "epoch": 1.8520547945205479, + "grad_norm": 2.980849027633667, + "learning_rate": 2.1547522339561333e-06, + "loss": 1.4159, + "step": 1690 + }, + { + "epoch": 1.8520547945205479, + "eval_loss": 1.6937642097473145, + "eval_runtime": 37.3409, + "eval_samples_per_second": 26.78, + "eval_steps_per_second": 1.125, + "step": 1690 + }, + { + "epoch": 1.853150684931507, + "grad_norm": 2.9203343391418457, + "learning_rate": 2.1527213647441106e-06, + "loss": 1.3949, + "step": 1691 + }, + { + "epoch": 1.8542465753424657, + "grad_norm": 2.928570032119751, + "learning_rate": 2.150690495532088e-06, + "loss": 1.4522, + "step": 1692 + }, + { + "epoch": 1.8553424657534245, + "grad_norm": 2.9594554901123047, + "learning_rate": 2.148659626320065e-06, + "loss": 1.4629, + "step": 1693 + }, + { + "epoch": 1.8564383561643836, + "grad_norm": 2.8420305252075195, + "learning_rate": 2.1466287571080424e-06, + "loss": 1.4525, + "step": 1694 + }, + { + "epoch": 1.8575342465753426, + "grad_norm": 2.8952643871307373, + "learning_rate": 2.1445978878960197e-06, + "loss": 1.4606, + "step": 1695 + }, + { + "epoch": 1.8586301369863014, + "grad_norm": 2.8206307888031006, + "learning_rate": 2.142567018683997e-06, + "loss": 1.402, + "step": 1696 + }, + { + "epoch": 1.8597260273972602, + "grad_norm": 2.9090628623962402, + "learning_rate": 2.140536149471974e-06, + "loss": 1.4146, + "step": 1697 + }, + { + "epoch": 1.8608219178082193, + "grad_norm": 3.018481731414795, + "learning_rate": 2.1385052802599515e-06, + "loss": 1.4749, + "step": 1698 + }, + { + "epoch": 1.861917808219178, + "grad_norm": 2.882465362548828, + "learning_rate": 2.1364744110479288e-06, + "loss": 1.4091, + "step": 1699 + }, + { + "epoch": 1.8630136986301369, + "grad_norm": 2.8599698543548584, + "learning_rate": 2.134443541835906e-06, + "loss": 1.4111, + "step": 1700 + }, + { + "epoch": 1.8630136986301369, + "eval_loss": 1.6989127397537231, + "eval_runtime": 37.4015, + "eval_samples_per_second": 26.737, + "eval_steps_per_second": 1.123, + "step": 1700 + }, + { + "epoch": 1.864109589041096, + "grad_norm": 2.9481594562530518, + "learning_rate": 2.132412672623883e-06, + "loss": 1.4633, + "step": 1701 + }, + { + "epoch": 1.865205479452055, + "grad_norm": 3.081756353378296, + "learning_rate": 2.13038180341186e-06, + "loss": 1.4384, + "step": 1702 + }, + { + "epoch": 1.8663013698630135, + "grad_norm": 2.8787295818328857, + "learning_rate": 2.1283509341998374e-06, + "loss": 1.4166, + "step": 1703 + }, + { + "epoch": 1.8673972602739726, + "grad_norm": 3.155799388885498, + "learning_rate": 2.1263200649878147e-06, + "loss": 1.4321, + "step": 1704 + }, + { + "epoch": 1.8684931506849316, + "grad_norm": 3.171860456466675, + "learning_rate": 2.124289195775792e-06, + "loss": 1.4681, + "step": 1705 + }, + { + "epoch": 1.8695890410958904, + "grad_norm": 2.899451971054077, + "learning_rate": 2.1222583265637693e-06, + "loss": 1.4297, + "step": 1706 + }, + { + "epoch": 1.8706849315068492, + "grad_norm": 2.827568292617798, + "learning_rate": 2.1202274573517465e-06, + "loss": 1.4312, + "step": 1707 + }, + { + "epoch": 1.8717808219178083, + "grad_norm": 2.9139182567596436, + "learning_rate": 2.118196588139724e-06, + "loss": 1.3538, + "step": 1708 + }, + { + "epoch": 1.872876712328767, + "grad_norm": 2.7959961891174316, + "learning_rate": 2.116165718927701e-06, + "loss": 1.3951, + "step": 1709 + }, + { + "epoch": 1.8739726027397259, + "grad_norm": 2.953160047531128, + "learning_rate": 2.1141348497156784e-06, + "loss": 1.42, + "step": 1710 + }, + { + "epoch": 1.8739726027397259, + "eval_loss": 1.6972320079803467, + "eval_runtime": 36.8414, + "eval_samples_per_second": 27.143, + "eval_steps_per_second": 1.14, + "step": 1710 + }, + { + "epoch": 1.875068493150685, + "grad_norm": 2.8701608180999756, + "learning_rate": 2.1121039805036556e-06, + "loss": 1.4407, + "step": 1711 + }, + { + "epoch": 1.876164383561644, + "grad_norm": 2.8174476623535156, + "learning_rate": 2.110073111291633e-06, + "loss": 1.3947, + "step": 1712 + }, + { + "epoch": 1.8772602739726028, + "grad_norm": 3.0026297569274902, + "learning_rate": 2.10804224207961e-06, + "loss": 1.4129, + "step": 1713 + }, + { + "epoch": 1.8783561643835616, + "grad_norm": 3.04539155960083, + "learning_rate": 2.1060113728675875e-06, + "loss": 1.4203, + "step": 1714 + }, + { + "epoch": 1.8794520547945206, + "grad_norm": 2.859358072280884, + "learning_rate": 2.1039805036555647e-06, + "loss": 1.4473, + "step": 1715 + }, + { + "epoch": 1.8805479452054794, + "grad_norm": 2.8927764892578125, + "learning_rate": 2.101949634443542e-06, + "loss": 1.4634, + "step": 1716 + }, + { + "epoch": 1.8816438356164382, + "grad_norm": 2.6834025382995605, + "learning_rate": 2.0999187652315193e-06, + "loss": 1.3519, + "step": 1717 + }, + { + "epoch": 1.8827397260273973, + "grad_norm": 2.936937093734741, + "learning_rate": 2.0978878960194966e-06, + "loss": 1.4234, + "step": 1718 + }, + { + "epoch": 1.8838356164383563, + "grad_norm": 2.884028673171997, + "learning_rate": 2.095857026807474e-06, + "loss": 1.4012, + "step": 1719 + }, + { + "epoch": 1.884931506849315, + "grad_norm": 2.9048874378204346, + "learning_rate": 2.093826157595451e-06, + "loss": 1.4283, + "step": 1720 + }, + { + "epoch": 1.884931506849315, + "eval_loss": 1.6949399709701538, + "eval_runtime": 36.9893, + "eval_samples_per_second": 27.035, + "eval_steps_per_second": 1.135, + "step": 1720 + }, + { + "epoch": 1.886027397260274, + "grad_norm": 2.802506923675537, + "learning_rate": 2.0917952883834284e-06, + "loss": 1.4151, + "step": 1721 + }, + { + "epoch": 1.887123287671233, + "grad_norm": 2.9587090015411377, + "learning_rate": 2.0897644191714057e-06, + "loss": 1.4226, + "step": 1722 + }, + { + "epoch": 1.8882191780821918, + "grad_norm": 2.892961263656616, + "learning_rate": 2.087733549959383e-06, + "loss": 1.4124, + "step": 1723 + }, + { + "epoch": 1.8893150684931506, + "grad_norm": 3.1364645957946777, + "learning_rate": 2.0857026807473602e-06, + "loss": 1.4706, + "step": 1724 + }, + { + "epoch": 1.8904109589041096, + "grad_norm": 3.190659284591675, + "learning_rate": 2.0836718115353375e-06, + "loss": 1.3948, + "step": 1725 + }, + { + "epoch": 1.8915068493150686, + "grad_norm": 2.9786376953125, + "learning_rate": 2.0816409423233144e-06, + "loss": 1.3867, + "step": 1726 + }, + { + "epoch": 1.8926027397260274, + "grad_norm": 2.9816296100616455, + "learning_rate": 2.0796100731112916e-06, + "loss": 1.4629, + "step": 1727 + }, + { + "epoch": 1.8936986301369862, + "grad_norm": 2.9736366271972656, + "learning_rate": 2.077579203899269e-06, + "loss": 1.414, + "step": 1728 + }, + { + "epoch": 1.8947945205479453, + "grad_norm": 2.881622552871704, + "learning_rate": 2.075548334687246e-06, + "loss": 1.4327, + "step": 1729 + }, + { + "epoch": 1.895890410958904, + "grad_norm": 2.835374355316162, + "learning_rate": 2.0735174654752235e-06, + "loss": 1.3747, + "step": 1730 + }, + { + "epoch": 1.895890410958904, + "eval_loss": 1.6909430027008057, + "eval_runtime": 36.9718, + "eval_samples_per_second": 27.048, + "eval_steps_per_second": 1.136, + "step": 1730 + }, + { + "epoch": 1.896986301369863, + "grad_norm": 2.883124828338623, + "learning_rate": 2.0714865962632007e-06, + "loss": 1.4439, + "step": 1731 + }, + { + "epoch": 1.898082191780822, + "grad_norm": 2.8171486854553223, + "learning_rate": 2.069455727051178e-06, + "loss": 1.3816, + "step": 1732 + }, + { + "epoch": 1.899178082191781, + "grad_norm": 2.8773303031921387, + "learning_rate": 2.0674248578391553e-06, + "loss": 1.4081, + "step": 1733 + }, + { + "epoch": 1.9002739726027398, + "grad_norm": 2.869682788848877, + "learning_rate": 2.0653939886271326e-06, + "loss": 1.4036, + "step": 1734 + }, + { + "epoch": 1.9013698630136986, + "grad_norm": 3.010930061340332, + "learning_rate": 2.06336311941511e-06, + "loss": 1.3839, + "step": 1735 + }, + { + "epoch": 1.9024657534246576, + "grad_norm": 3.062164306640625, + "learning_rate": 2.061332250203087e-06, + "loss": 1.4705, + "step": 1736 + }, + { + "epoch": 1.9035616438356164, + "grad_norm": 2.8629584312438965, + "learning_rate": 2.0593013809910644e-06, + "loss": 1.4104, + "step": 1737 + }, + { + "epoch": 1.9046575342465752, + "grad_norm": 2.817843198776245, + "learning_rate": 2.0572705117790417e-06, + "loss": 1.3828, + "step": 1738 + }, + { + "epoch": 1.9057534246575343, + "grad_norm": 2.832322597503662, + "learning_rate": 2.055239642567019e-06, + "loss": 1.4327, + "step": 1739 + }, + { + "epoch": 1.9068493150684933, + "grad_norm": 2.8618762493133545, + "learning_rate": 2.0532087733549962e-06, + "loss": 1.4533, + "step": 1740 + }, + { + "epoch": 1.9068493150684933, + "eval_loss": 1.696921467781067, + "eval_runtime": 36.9736, + "eval_samples_per_second": 27.046, + "eval_steps_per_second": 1.136, + "step": 1740 + }, + { + "epoch": 1.907945205479452, + "grad_norm": 2.854097366333008, + "learning_rate": 2.0511779041429735e-06, + "loss": 1.4353, + "step": 1741 + }, + { + "epoch": 1.909041095890411, + "grad_norm": 2.85565185546875, + "learning_rate": 2.0491470349309508e-06, + "loss": 1.4614, + "step": 1742 + }, + { + "epoch": 1.91013698630137, + "grad_norm": 2.7551705837249756, + "learning_rate": 2.047116165718928e-06, + "loss": 1.3815, + "step": 1743 + }, + { + "epoch": 1.9112328767123288, + "grad_norm": 3.0475997924804688, + "learning_rate": 2.0450852965069053e-06, + "loss": 1.4509, + "step": 1744 + }, + { + "epoch": 1.9123287671232876, + "grad_norm": 2.958827018737793, + "learning_rate": 2.0430544272948826e-06, + "loss": 1.4017, + "step": 1745 + }, + { + "epoch": 1.9134246575342466, + "grad_norm": 3.0207631587982178, + "learning_rate": 2.04102355808286e-06, + "loss": 1.4492, + "step": 1746 + }, + { + "epoch": 1.9145205479452054, + "grad_norm": 2.8824949264526367, + "learning_rate": 2.038992688870837e-06, + "loss": 1.4063, + "step": 1747 + }, + { + "epoch": 1.9156164383561642, + "grad_norm": 2.9545154571533203, + "learning_rate": 2.0369618196588144e-06, + "loss": 1.3854, + "step": 1748 + }, + { + "epoch": 1.9167123287671233, + "grad_norm": 2.940086841583252, + "learning_rate": 2.0349309504467917e-06, + "loss": 1.3933, + "step": 1749 + }, + { + "epoch": 1.9178082191780823, + "grad_norm": 2.902759313583374, + "learning_rate": 2.032900081234769e-06, + "loss": 1.4023, + "step": 1750 + }, + { + "epoch": 1.9178082191780823, + "eval_loss": 1.698394775390625, + "eval_runtime": 36.9966, + "eval_samples_per_second": 27.029, + "eval_steps_per_second": 1.135, + "step": 1750 + }, + { + "epoch": 1.9189041095890411, + "grad_norm": 2.911677598953247, + "learning_rate": 2.030869212022746e-06, + "loss": 1.4422, + "step": 1751 + }, + { + "epoch": 1.92, + "grad_norm": 2.9415669441223145, + "learning_rate": 2.028838342810723e-06, + "loss": 1.4459, + "step": 1752 + }, + { + "epoch": 1.921095890410959, + "grad_norm": 2.8038763999938965, + "learning_rate": 2.0268074735987004e-06, + "loss": 1.4024, + "step": 1753 + }, + { + "epoch": 1.9221917808219178, + "grad_norm": 2.9599249362945557, + "learning_rate": 2.0247766043866777e-06, + "loss": 1.4005, + "step": 1754 + }, + { + "epoch": 1.9232876712328766, + "grad_norm": 2.925330638885498, + "learning_rate": 2.022745735174655e-06, + "loss": 1.405, + "step": 1755 + }, + { + "epoch": 1.9243835616438356, + "grad_norm": 3.033116102218628, + "learning_rate": 2.020714865962632e-06, + "loss": 1.439, + "step": 1756 + }, + { + "epoch": 1.9254794520547946, + "grad_norm": 2.928602933883667, + "learning_rate": 2.0186839967506095e-06, + "loss": 1.4246, + "step": 1757 + }, + { + "epoch": 1.9265753424657535, + "grad_norm": 2.9250142574310303, + "learning_rate": 2.0166531275385868e-06, + "loss": 1.4335, + "step": 1758 + }, + { + "epoch": 1.9276712328767123, + "grad_norm": 2.8789937496185303, + "learning_rate": 2.014622258326564e-06, + "loss": 1.4172, + "step": 1759 + }, + { + "epoch": 1.9287671232876713, + "grad_norm": 2.899266481399536, + "learning_rate": 2.0125913891145413e-06, + "loss": 1.3733, + "step": 1760 + }, + { + "epoch": 1.9287671232876713, + "eval_loss": 1.695684790611267, + "eval_runtime": 37.0188, + "eval_samples_per_second": 27.013, + "eval_steps_per_second": 1.135, + "step": 1760 + }, + { + "epoch": 1.9298630136986301, + "grad_norm": 2.962355136871338, + "learning_rate": 2.0105605199025186e-06, + "loss": 1.4441, + "step": 1761 + }, + { + "epoch": 1.930958904109589, + "grad_norm": 2.8060641288757324, + "learning_rate": 2.008529650690496e-06, + "loss": 1.4185, + "step": 1762 + }, + { + "epoch": 1.932054794520548, + "grad_norm": 2.8926310539245605, + "learning_rate": 2.0064987814784727e-06, + "loss": 1.4217, + "step": 1763 + }, + { + "epoch": 1.933150684931507, + "grad_norm": 2.918185234069824, + "learning_rate": 2.00446791226645e-06, + "loss": 1.4566, + "step": 1764 + }, + { + "epoch": 1.9342465753424658, + "grad_norm": 2.9657108783721924, + "learning_rate": 2.0024370430544273e-06, + "loss": 1.3967, + "step": 1765 + }, + { + "epoch": 1.9353424657534246, + "grad_norm": 2.7738616466522217, + "learning_rate": 2.0004061738424045e-06, + "loss": 1.371, + "step": 1766 + }, + { + "epoch": 1.9364383561643836, + "grad_norm": 2.919724941253662, + "learning_rate": 1.998375304630382e-06, + "loss": 1.4405, + "step": 1767 + }, + { + "epoch": 1.9375342465753425, + "grad_norm": 3.1406972408294678, + "learning_rate": 1.996344435418359e-06, + "loss": 1.4645, + "step": 1768 + }, + { + "epoch": 1.9386301369863013, + "grad_norm": 2.8543457984924316, + "learning_rate": 1.9943135662063364e-06, + "loss": 1.4421, + "step": 1769 + }, + { + "epoch": 1.9397260273972603, + "grad_norm": 3.021143674850464, + "learning_rate": 1.9922826969943136e-06, + "loss": 1.4438, + "step": 1770 + }, + { + "epoch": 1.9397260273972603, + "eval_loss": 1.6932497024536133, + "eval_runtime": 37.0311, + "eval_samples_per_second": 27.004, + "eval_steps_per_second": 1.134, + "step": 1770 + }, + { + "epoch": 1.9408219178082193, + "grad_norm": 3.104332685470581, + "learning_rate": 1.990251827782291e-06, + "loss": 1.4577, + "step": 1771 + }, + { + "epoch": 1.941917808219178, + "grad_norm": 2.956894874572754, + "learning_rate": 1.988220958570268e-06, + "loss": 1.4392, + "step": 1772 + }, + { + "epoch": 1.943013698630137, + "grad_norm": 2.866647243499756, + "learning_rate": 1.9861900893582455e-06, + "loss": 1.4492, + "step": 1773 + }, + { + "epoch": 1.944109589041096, + "grad_norm": 2.908168315887451, + "learning_rate": 1.9841592201462227e-06, + "loss": 1.4252, + "step": 1774 + }, + { + "epoch": 1.9452054794520548, + "grad_norm": 2.9659180641174316, + "learning_rate": 1.9821283509342e-06, + "loss": 1.4237, + "step": 1775 + }, + { + "epoch": 1.9463013698630136, + "grad_norm": 2.7656569480895996, + "learning_rate": 1.9800974817221773e-06, + "loss": 1.3896, + "step": 1776 + }, + { + "epoch": 1.9473972602739726, + "grad_norm": 2.931061267852783, + "learning_rate": 1.9780666125101546e-06, + "loss": 1.3917, + "step": 1777 + }, + { + "epoch": 1.9484931506849315, + "grad_norm": 3.035163402557373, + "learning_rate": 1.976035743298132e-06, + "loss": 1.4509, + "step": 1778 + }, + { + "epoch": 1.9495890410958903, + "grad_norm": 2.932955741882324, + "learning_rate": 1.974004874086109e-06, + "loss": 1.4007, + "step": 1779 + }, + { + "epoch": 1.9506849315068493, + "grad_norm": 3.1727302074432373, + "learning_rate": 1.9719740048740864e-06, + "loss": 1.4896, + "step": 1780 + }, + { + "epoch": 1.9506849315068493, + "eval_loss": 1.6910525560379028, + "eval_runtime": 37.2984, + "eval_samples_per_second": 26.811, + "eval_steps_per_second": 1.126, + "step": 1780 + }, + { + "epoch": 1.9517808219178083, + "grad_norm": 2.908858060836792, + "learning_rate": 1.9699431356620637e-06, + "loss": 1.3855, + "step": 1781 + }, + { + "epoch": 1.9528767123287671, + "grad_norm": 2.8864009380340576, + "learning_rate": 1.967912266450041e-06, + "loss": 1.4295, + "step": 1782 + }, + { + "epoch": 1.953972602739726, + "grad_norm": 2.927523136138916, + "learning_rate": 1.9658813972380182e-06, + "loss": 1.4145, + "step": 1783 + }, + { + "epoch": 1.955068493150685, + "grad_norm": 2.8562896251678467, + "learning_rate": 1.9638505280259955e-06, + "loss": 1.3858, + "step": 1784 + }, + { + "epoch": 1.9561643835616438, + "grad_norm": 3.007047653198242, + "learning_rate": 1.9618196588139728e-06, + "loss": 1.4802, + "step": 1785 + }, + { + "epoch": 1.9572602739726026, + "grad_norm": 2.8853635787963867, + "learning_rate": 1.95978878960195e-06, + "loss": 1.3956, + "step": 1786 + }, + { + "epoch": 1.9583561643835616, + "grad_norm": 2.980919361114502, + "learning_rate": 1.9577579203899273e-06, + "loss": 1.4323, + "step": 1787 + }, + { + "epoch": 1.9594520547945207, + "grad_norm": 2.959479570388794, + "learning_rate": 1.955727051177904e-06, + "loss": 1.484, + "step": 1788 + }, + { + "epoch": 1.9605479452054795, + "grad_norm": 2.959350347518921, + "learning_rate": 1.9536961819658815e-06, + "loss": 1.4426, + "step": 1789 + }, + { + "epoch": 1.9616438356164383, + "grad_norm": 3.0871453285217285, + "learning_rate": 1.9516653127538587e-06, + "loss": 1.4136, + "step": 1790 + }, + { + "epoch": 1.9616438356164383, + "eval_loss": 1.69205641746521, + "eval_runtime": 37.4, + "eval_samples_per_second": 26.738, + "eval_steps_per_second": 1.123, + "step": 1790 + }, + { + "epoch": 1.9627397260273973, + "grad_norm": 3.0216007232666016, + "learning_rate": 1.949634443541836e-06, + "loss": 1.402, + "step": 1791 + }, + { + "epoch": 1.9638356164383561, + "grad_norm": 3.036532163619995, + "learning_rate": 1.9476035743298133e-06, + "loss": 1.3962, + "step": 1792 + }, + { + "epoch": 1.964931506849315, + "grad_norm": 2.937406301498413, + "learning_rate": 1.9455727051177906e-06, + "loss": 1.3911, + "step": 1793 + }, + { + "epoch": 1.966027397260274, + "grad_norm": 2.9001076221466064, + "learning_rate": 1.943541835905768e-06, + "loss": 1.4495, + "step": 1794 + }, + { + "epoch": 1.967123287671233, + "grad_norm": 2.950719118118286, + "learning_rate": 1.941510966693745e-06, + "loss": 1.4021, + "step": 1795 + }, + { + "epoch": 1.9682191780821918, + "grad_norm": 2.9256081581115723, + "learning_rate": 1.9394800974817224e-06, + "loss": 1.4526, + "step": 1796 + }, + { + "epoch": 1.9693150684931506, + "grad_norm": 2.793117046356201, + "learning_rate": 1.9374492282696997e-06, + "loss": 1.3817, + "step": 1797 + }, + { + "epoch": 1.9704109589041097, + "grad_norm": 2.9317514896392822, + "learning_rate": 1.935418359057677e-06, + "loss": 1.4452, + "step": 1798 + }, + { + "epoch": 1.9715068493150685, + "grad_norm": 2.887756824493408, + "learning_rate": 1.9333874898456542e-06, + "loss": 1.4738, + "step": 1799 + }, + { + "epoch": 1.9726027397260273, + "grad_norm": 3.0734620094299316, + "learning_rate": 1.931356620633631e-06, + "loss": 1.4422, + "step": 1800 + }, + { + "epoch": 1.9726027397260273, + "eval_loss": 1.6927837133407593, + "eval_runtime": 37.3234, + "eval_samples_per_second": 26.793, + "eval_steps_per_second": 1.125, + "step": 1800 + }, + { + "epoch": 1.9736986301369863, + "grad_norm": 2.944976329803467, + "learning_rate": 1.9293257514216083e-06, + "loss": 1.4388, + "step": 1801 + }, + { + "epoch": 1.9747945205479454, + "grad_norm": 3.0636444091796875, + "learning_rate": 1.9272948822095856e-06, + "loss": 1.3737, + "step": 1802 + }, + { + "epoch": 1.9758904109589042, + "grad_norm": 2.952333450317383, + "learning_rate": 1.925264012997563e-06, + "loss": 1.419, + "step": 1803 + }, + { + "epoch": 1.976986301369863, + "grad_norm": 2.974330186843872, + "learning_rate": 1.92323314378554e-06, + "loss": 1.3994, + "step": 1804 + }, + { + "epoch": 1.978082191780822, + "grad_norm": 2.8695459365844727, + "learning_rate": 1.9212022745735174e-06, + "loss": 1.4307, + "step": 1805 + }, + { + "epoch": 1.9791780821917808, + "grad_norm": 3.035994052886963, + "learning_rate": 1.9191714053614947e-06, + "loss": 1.4398, + "step": 1806 + }, + { + "epoch": 1.9802739726027396, + "grad_norm": 2.935548782348633, + "learning_rate": 1.917140536149472e-06, + "loss": 1.3938, + "step": 1807 + }, + { + "epoch": 1.9813698630136987, + "grad_norm": 2.970191240310669, + "learning_rate": 1.9151096669374493e-06, + "loss": 1.3912, + "step": 1808 + }, + { + "epoch": 1.9824657534246577, + "grad_norm": 2.902297019958496, + "learning_rate": 1.9130787977254266e-06, + "loss": 1.3691, + "step": 1809 + }, + { + "epoch": 1.9835616438356163, + "grad_norm": 3.057100534439087, + "learning_rate": 1.911047928513404e-06, + "loss": 1.4293, + "step": 1810 + }, + { + "epoch": 1.9835616438356163, + "eval_loss": 1.6943376064300537, + "eval_runtime": 36.9253, + "eval_samples_per_second": 27.082, + "eval_steps_per_second": 1.137, + "step": 1810 + }, + { + "epoch": 1.9846575342465753, + "grad_norm": 3.0681138038635254, + "learning_rate": 1.909017059301381e-06, + "loss": 1.421, + "step": 1811 + }, + { + "epoch": 1.9857534246575343, + "grad_norm": 2.911562919616699, + "learning_rate": 1.9069861900893586e-06, + "loss": 1.4041, + "step": 1812 + }, + { + "epoch": 1.9868493150684932, + "grad_norm": 3.0662248134613037, + "learning_rate": 1.9049553208773354e-06, + "loss": 1.4073, + "step": 1813 + }, + { + "epoch": 1.987945205479452, + "grad_norm": 2.8864450454711914, + "learning_rate": 1.9029244516653127e-06, + "loss": 1.4322, + "step": 1814 + }, + { + "epoch": 1.989041095890411, + "grad_norm": 2.834134101867676, + "learning_rate": 1.90089358245329e-06, + "loss": 1.4123, + "step": 1815 + }, + { + "epoch": 1.9901369863013698, + "grad_norm": 2.840860366821289, + "learning_rate": 1.8988627132412673e-06, + "loss": 1.3708, + "step": 1816 + }, + { + "epoch": 1.9912328767123286, + "grad_norm": 2.8203835487365723, + "learning_rate": 1.8968318440292445e-06, + "loss": 1.4014, + "step": 1817 + }, + { + "epoch": 1.9923287671232877, + "grad_norm": 2.8455588817596436, + "learning_rate": 1.8948009748172218e-06, + "loss": 1.4349, + "step": 1818 + }, + { + "epoch": 1.9934246575342467, + "grad_norm": 2.8895022869110107, + "learning_rate": 1.892770105605199e-06, + "loss": 1.3951, + "step": 1819 + }, + { + "epoch": 1.9945205479452055, + "grad_norm": 2.8265910148620605, + "learning_rate": 1.8907392363931764e-06, + "loss": 1.4221, + "step": 1820 + }, + { + "epoch": 1.9945205479452055, + "eval_loss": 1.6890499591827393, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.999, + "eval_steps_per_second": 1.134, + "step": 1820 + }, + { + "epoch": 1.9956164383561643, + "grad_norm": 3.0007691383361816, + "learning_rate": 1.8887083671811536e-06, + "loss": 1.4442, + "step": 1821 + }, + { + "epoch": 1.9967123287671233, + "grad_norm": 3.0173628330230713, + "learning_rate": 1.886677497969131e-06, + "loss": 1.3975, + "step": 1822 + }, + { + "epoch": 1.9978082191780822, + "grad_norm": 2.95487904548645, + "learning_rate": 1.8846466287571082e-06, + "loss": 1.4277, + "step": 1823 + }, + { + "epoch": 1.998904109589041, + "grad_norm": 2.9277896881103516, + "learning_rate": 1.8826157595450855e-06, + "loss": 1.3423, + "step": 1824 + }, + { + "epoch": 2.0, + "grad_norm": 3.0760998725891113, + "learning_rate": 1.8805848903330625e-06, + "loss": 1.4272, + "step": 1825 + }, + { + "epoch": 2.001095890410959, + "grad_norm": 2.9844436645507812, + "learning_rate": 1.8785540211210398e-06, + "loss": 1.3486, + "step": 1826 + }, + { + "epoch": 2.0021917808219176, + "grad_norm": 3.2406389713287354, + "learning_rate": 1.876523151909017e-06, + "loss": 1.4703, + "step": 1827 + }, + { + "epoch": 2.0032876712328767, + "grad_norm": 3.3263025283813477, + "learning_rate": 1.8744922826969944e-06, + "loss": 1.3731, + "step": 1828 + }, + { + "epoch": 2.0043835616438357, + "grad_norm": 3.2702267169952393, + "learning_rate": 1.8724614134849716e-06, + "loss": 1.3585, + "step": 1829 + }, + { + "epoch": 2.0054794520547947, + "grad_norm": 3.20688796043396, + "learning_rate": 1.870430544272949e-06, + "loss": 1.3604, + "step": 1830 + }, + { + "epoch": 2.0054794520547947, + "eval_loss": 1.6996374130249023, + "eval_runtime": 37.1062, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 1830 + }, + { + "epoch": 2.0065753424657533, + "grad_norm": 3.1598732471466064, + "learning_rate": 1.8683996750609262e-06, + "loss": 1.37, + "step": 1831 + }, + { + "epoch": 2.0076712328767123, + "grad_norm": 3.580209970474243, + "learning_rate": 1.8663688058489035e-06, + "loss": 1.4601, + "step": 1832 + }, + { + "epoch": 2.0087671232876714, + "grad_norm": 3.365760564804077, + "learning_rate": 1.8643379366368807e-06, + "loss": 1.3636, + "step": 1833 + }, + { + "epoch": 2.00986301369863, + "grad_norm": 3.3804492950439453, + "learning_rate": 1.862307067424858e-06, + "loss": 1.3684, + "step": 1834 + }, + { + "epoch": 2.010958904109589, + "grad_norm": 3.510930061340332, + "learning_rate": 1.8602761982128353e-06, + "loss": 1.3713, + "step": 1835 + }, + { + "epoch": 2.012054794520548, + "grad_norm": 3.299884557723999, + "learning_rate": 1.8582453290008126e-06, + "loss": 1.3907, + "step": 1836 + }, + { + "epoch": 2.013150684931507, + "grad_norm": 3.3502068519592285, + "learning_rate": 1.8562144597887898e-06, + "loss": 1.4178, + "step": 1837 + }, + { + "epoch": 2.0142465753424657, + "grad_norm": 3.1439013481140137, + "learning_rate": 1.854183590576767e-06, + "loss": 1.3672, + "step": 1838 + }, + { + "epoch": 2.0153424657534247, + "grad_norm": 3.3565664291381836, + "learning_rate": 1.8521527213647442e-06, + "loss": 1.3785, + "step": 1839 + }, + { + "epoch": 2.0164383561643837, + "grad_norm": 3.3047420978546143, + "learning_rate": 1.8501218521527215e-06, + "loss": 1.4195, + "step": 1840 + }, + { + "epoch": 2.0164383561643837, + "eval_loss": 1.696365475654602, + "eval_runtime": 37.1064, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 1840 + }, + { + "epoch": 2.0175342465753423, + "grad_norm": 3.27755069732666, + "learning_rate": 1.8480909829406987e-06, + "loss": 1.3518, + "step": 1841 + }, + { + "epoch": 2.0186301369863013, + "grad_norm": 3.3251404762268066, + "learning_rate": 1.846060113728676e-06, + "loss": 1.3156, + "step": 1842 + }, + { + "epoch": 2.0197260273972604, + "grad_norm": 3.345583915710449, + "learning_rate": 1.8440292445166533e-06, + "loss": 1.35, + "step": 1843 + }, + { + "epoch": 2.020821917808219, + "grad_norm": 3.296478509902954, + "learning_rate": 1.8419983753046306e-06, + "loss": 1.3983, + "step": 1844 + }, + { + "epoch": 2.021917808219178, + "grad_norm": 3.2481424808502197, + "learning_rate": 1.8399675060926078e-06, + "loss": 1.3219, + "step": 1845 + }, + { + "epoch": 2.023013698630137, + "grad_norm": 3.3519129753112793, + "learning_rate": 1.8379366368805851e-06, + "loss": 1.4329, + "step": 1846 + }, + { + "epoch": 2.024109589041096, + "grad_norm": 3.2225515842437744, + "learning_rate": 1.8359057676685624e-06, + "loss": 1.3769, + "step": 1847 + }, + { + "epoch": 2.0252054794520546, + "grad_norm": 3.1705095767974854, + "learning_rate": 1.8338748984565397e-06, + "loss": 1.3731, + "step": 1848 + }, + { + "epoch": 2.0263013698630137, + "grad_norm": 3.2630999088287354, + "learning_rate": 1.831844029244517e-06, + "loss": 1.3581, + "step": 1849 + }, + { + "epoch": 2.0273972602739727, + "grad_norm": 3.141310214996338, + "learning_rate": 1.829813160032494e-06, + "loss": 1.3779, + "step": 1850 + }, + { + "epoch": 2.0273972602739727, + "eval_loss": 1.6947910785675049, + "eval_runtime": 37.091, + "eval_samples_per_second": 26.961, + "eval_steps_per_second": 1.132, + "step": 1850 + }, + { + "epoch": 2.0284931506849313, + "grad_norm": 3.2116565704345703, + "learning_rate": 1.8277822908204713e-06, + "loss": 1.3419, + "step": 1851 + }, + { + "epoch": 2.0295890410958903, + "grad_norm": 3.255674123764038, + "learning_rate": 1.8257514216084486e-06, + "loss": 1.3688, + "step": 1852 + }, + { + "epoch": 2.0306849315068494, + "grad_norm": 3.458174467086792, + "learning_rate": 1.8237205523964258e-06, + "loss": 1.3872, + "step": 1853 + }, + { + "epoch": 2.0317808219178084, + "grad_norm": 3.430821418762207, + "learning_rate": 1.8216896831844031e-06, + "loss": 1.3806, + "step": 1854 + }, + { + "epoch": 2.032876712328767, + "grad_norm": 3.4024930000305176, + "learning_rate": 1.8196588139723804e-06, + "loss": 1.3789, + "step": 1855 + }, + { + "epoch": 2.033972602739726, + "grad_norm": 3.523206949234009, + "learning_rate": 1.8176279447603577e-06, + "loss": 1.4102, + "step": 1856 + }, + { + "epoch": 2.035068493150685, + "grad_norm": 3.2846553325653076, + "learning_rate": 1.815597075548335e-06, + "loss": 1.34, + "step": 1857 + }, + { + "epoch": 2.0361643835616436, + "grad_norm": 3.2146921157836914, + "learning_rate": 1.8135662063363122e-06, + "loss": 1.3324, + "step": 1858 + }, + { + "epoch": 2.0372602739726027, + "grad_norm": 3.3313839435577393, + "learning_rate": 1.8115353371242895e-06, + "loss": 1.382, + "step": 1859 + }, + { + "epoch": 2.0383561643835617, + "grad_norm": 3.3586273193359375, + "learning_rate": 1.8095044679122668e-06, + "loss": 1.4161, + "step": 1860 + }, + { + "epoch": 2.0383561643835617, + "eval_loss": 1.6938049793243408, + "eval_runtime": 37.0874, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 1860 + }, + { + "epoch": 2.0394520547945207, + "grad_norm": 3.1919045448303223, + "learning_rate": 1.807473598700244e-06, + "loss": 1.3089, + "step": 1861 + }, + { + "epoch": 2.0405479452054793, + "grad_norm": 3.212191104888916, + "learning_rate": 1.805442729488221e-06, + "loss": 1.3466, + "step": 1862 + }, + { + "epoch": 2.0416438356164384, + "grad_norm": 3.2063724994659424, + "learning_rate": 1.8034118602761982e-06, + "loss": 1.3421, + "step": 1863 + }, + { + "epoch": 2.0427397260273974, + "grad_norm": 3.2476260662078857, + "learning_rate": 1.8013809910641754e-06, + "loss": 1.3616, + "step": 1864 + }, + { + "epoch": 2.043835616438356, + "grad_norm": 3.244793653488159, + "learning_rate": 1.7993501218521527e-06, + "loss": 1.3615, + "step": 1865 + }, + { + "epoch": 2.044931506849315, + "grad_norm": 3.336434841156006, + "learning_rate": 1.79731925264013e-06, + "loss": 1.3719, + "step": 1866 + }, + { + "epoch": 2.046027397260274, + "grad_norm": 3.3474669456481934, + "learning_rate": 1.7952883834281073e-06, + "loss": 1.3392, + "step": 1867 + }, + { + "epoch": 2.047123287671233, + "grad_norm": 3.3928866386413574, + "learning_rate": 1.7932575142160846e-06, + "loss": 1.3884, + "step": 1868 + }, + { + "epoch": 2.0482191780821917, + "grad_norm": 3.2143406867980957, + "learning_rate": 1.7912266450040618e-06, + "loss": 1.3475, + "step": 1869 + }, + { + "epoch": 2.0493150684931507, + "grad_norm": 3.2816200256347656, + "learning_rate": 1.789195775792039e-06, + "loss": 1.403, + "step": 1870 + }, + { + "epoch": 2.0493150684931507, + "eval_loss": 1.6954723596572876, + "eval_runtime": 37.0803, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 1870 + }, + { + "epoch": 2.0504109589041097, + "grad_norm": 3.3261868953704834, + "learning_rate": 1.7871649065800164e-06, + "loss": 1.397, + "step": 1871 + }, + { + "epoch": 2.0515068493150683, + "grad_norm": 3.216740131378174, + "learning_rate": 1.7851340373679937e-06, + "loss": 1.323, + "step": 1872 + }, + { + "epoch": 2.0526027397260274, + "grad_norm": 3.2935290336608887, + "learning_rate": 1.783103168155971e-06, + "loss": 1.4114, + "step": 1873 + }, + { + "epoch": 2.0536986301369864, + "grad_norm": 3.1856577396392822, + "learning_rate": 1.7810722989439482e-06, + "loss": 1.3441, + "step": 1874 + }, + { + "epoch": 2.0547945205479454, + "grad_norm": 3.1668901443481445, + "learning_rate": 1.7790414297319253e-06, + "loss": 1.3624, + "step": 1875 + }, + { + "epoch": 2.055890410958904, + "grad_norm": 3.2604808807373047, + "learning_rate": 1.7770105605199025e-06, + "loss": 1.3661, + "step": 1876 + }, + { + "epoch": 2.056986301369863, + "grad_norm": 3.3410542011260986, + "learning_rate": 1.7749796913078798e-06, + "loss": 1.3562, + "step": 1877 + }, + { + "epoch": 2.058082191780822, + "grad_norm": 3.2885286808013916, + "learning_rate": 1.772948822095857e-06, + "loss": 1.3472, + "step": 1878 + }, + { + "epoch": 2.0591780821917807, + "grad_norm": 3.4261269569396973, + "learning_rate": 1.7709179528838344e-06, + "loss": 1.3728, + "step": 1879 + }, + { + "epoch": 2.0602739726027397, + "grad_norm": 3.474125862121582, + "learning_rate": 1.7688870836718116e-06, + "loss": 1.4013, + "step": 1880 + }, + { + "epoch": 2.0602739726027397, + "eval_loss": 1.6941890716552734, + "eval_runtime": 37.0664, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 1880 + }, + { + "epoch": 2.0613698630136987, + "grad_norm": 3.468127727508545, + "learning_rate": 1.766856214459789e-06, + "loss": 1.4034, + "step": 1881 + }, + { + "epoch": 2.0624657534246573, + "grad_norm": 3.304748296737671, + "learning_rate": 1.7648253452477662e-06, + "loss": 1.3565, + "step": 1882 + }, + { + "epoch": 2.0635616438356164, + "grad_norm": 3.348489761352539, + "learning_rate": 1.7627944760357435e-06, + "loss": 1.3612, + "step": 1883 + }, + { + "epoch": 2.0646575342465754, + "grad_norm": 3.329779863357544, + "learning_rate": 1.7607636068237208e-06, + "loss": 1.3463, + "step": 1884 + }, + { + "epoch": 2.0657534246575344, + "grad_norm": 3.363576889038086, + "learning_rate": 1.758732737611698e-06, + "loss": 1.4082, + "step": 1885 + }, + { + "epoch": 2.066849315068493, + "grad_norm": 3.2667229175567627, + "learning_rate": 1.7567018683996753e-06, + "loss": 1.3731, + "step": 1886 + }, + { + "epoch": 2.067945205479452, + "grad_norm": 3.250871181488037, + "learning_rate": 1.7546709991876524e-06, + "loss": 1.3729, + "step": 1887 + }, + { + "epoch": 2.069041095890411, + "grad_norm": 3.351614475250244, + "learning_rate": 1.7526401299756296e-06, + "loss": 1.3649, + "step": 1888 + }, + { + "epoch": 2.0701369863013697, + "grad_norm": 3.196298122406006, + "learning_rate": 1.750609260763607e-06, + "loss": 1.3405, + "step": 1889 + }, + { + "epoch": 2.0712328767123287, + "grad_norm": 3.156193494796753, + "learning_rate": 1.7485783915515842e-06, + "loss": 1.3572, + "step": 1890 + }, + { + "epoch": 2.0712328767123287, + "eval_loss": 1.69057297706604, + "eval_runtime": 37.2496, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1890 + }, + { + "epoch": 2.0723287671232877, + "grad_norm": 3.503941059112549, + "learning_rate": 1.7465475223395615e-06, + "loss": 1.3699, + "step": 1891 + }, + { + "epoch": 2.0734246575342468, + "grad_norm": 3.309138298034668, + "learning_rate": 1.7445166531275387e-06, + "loss": 1.359, + "step": 1892 + }, + { + "epoch": 2.0745205479452054, + "grad_norm": 3.210146903991699, + "learning_rate": 1.742485783915516e-06, + "loss": 1.3247, + "step": 1893 + }, + { + "epoch": 2.0756164383561644, + "grad_norm": 3.6310641765594482, + "learning_rate": 1.7404549147034933e-06, + "loss": 1.4335, + "step": 1894 + }, + { + "epoch": 2.0767123287671234, + "grad_norm": 3.5245156288146973, + "learning_rate": 1.7384240454914706e-06, + "loss": 1.4295, + "step": 1895 + }, + { + "epoch": 2.077808219178082, + "grad_norm": 3.3318674564361572, + "learning_rate": 1.7363931762794478e-06, + "loss": 1.3068, + "step": 1896 + }, + { + "epoch": 2.078904109589041, + "grad_norm": 3.3639466762542725, + "learning_rate": 1.7343623070674251e-06, + "loss": 1.37, + "step": 1897 + }, + { + "epoch": 2.08, + "grad_norm": 3.2703022956848145, + "learning_rate": 1.7323314378554024e-06, + "loss": 1.3847, + "step": 1898 + }, + { + "epoch": 2.081095890410959, + "grad_norm": 3.318593740463257, + "learning_rate": 1.7303005686433797e-06, + "loss": 1.3597, + "step": 1899 + }, + { + "epoch": 2.0821917808219177, + "grad_norm": 3.3997914791107178, + "learning_rate": 1.7282696994313567e-06, + "loss": 1.3802, + "step": 1900 + }, + { + "epoch": 2.0821917808219177, + "eval_loss": 1.6921792030334473, + "eval_runtime": 37.4078, + "eval_samples_per_second": 26.732, + "eval_steps_per_second": 1.123, + "step": 1900 + }, + { + "epoch": 2.0832876712328767, + "grad_norm": 3.3228092193603516, + "learning_rate": 1.726238830219334e-06, + "loss": 1.3674, + "step": 1901 + }, + { + "epoch": 2.0843835616438358, + "grad_norm": 3.487579345703125, + "learning_rate": 1.7242079610073113e-06, + "loss": 1.3681, + "step": 1902 + }, + { + "epoch": 2.0854794520547943, + "grad_norm": 3.306591749191284, + "learning_rate": 1.7221770917952886e-06, + "loss": 1.371, + "step": 1903 + }, + { + "epoch": 2.0865753424657534, + "grad_norm": 3.483325242996216, + "learning_rate": 1.7201462225832658e-06, + "loss": 1.4151, + "step": 1904 + }, + { + "epoch": 2.0876712328767124, + "grad_norm": 3.4242193698883057, + "learning_rate": 1.7181153533712431e-06, + "loss": 1.3765, + "step": 1905 + }, + { + "epoch": 2.0887671232876714, + "grad_norm": 3.3638198375701904, + "learning_rate": 1.7160844841592204e-06, + "loss": 1.3977, + "step": 1906 + }, + { + "epoch": 2.08986301369863, + "grad_norm": 3.2706453800201416, + "learning_rate": 1.7140536149471977e-06, + "loss": 1.3004, + "step": 1907 + }, + { + "epoch": 2.090958904109589, + "grad_norm": 3.1796586513519287, + "learning_rate": 1.712022745735175e-06, + "loss": 1.3106, + "step": 1908 + }, + { + "epoch": 2.092054794520548, + "grad_norm": 3.211973190307617, + "learning_rate": 1.7099918765231522e-06, + "loss": 1.3555, + "step": 1909 + }, + { + "epoch": 2.0931506849315067, + "grad_norm": 3.2266228199005127, + "learning_rate": 1.7079610073111295e-06, + "loss": 1.3576, + "step": 1910 + }, + { + "epoch": 2.0931506849315067, + "eval_loss": 1.6927424669265747, + "eval_runtime": 36.8805, + "eval_samples_per_second": 27.115, + "eval_steps_per_second": 1.139, + "step": 1910 + }, + { + "epoch": 2.0942465753424657, + "grad_norm": 3.406233787536621, + "learning_rate": 1.7059301380991068e-06, + "loss": 1.3588, + "step": 1911 + }, + { + "epoch": 2.0953424657534248, + "grad_norm": 3.185476303100586, + "learning_rate": 1.7038992688870836e-06, + "loss": 1.3233, + "step": 1912 + }, + { + "epoch": 2.0964383561643833, + "grad_norm": 3.183128595352173, + "learning_rate": 1.701868399675061e-06, + "loss": 1.362, + "step": 1913 + }, + { + "epoch": 2.0975342465753424, + "grad_norm": 3.351947546005249, + "learning_rate": 1.6998375304630382e-06, + "loss": 1.375, + "step": 1914 + }, + { + "epoch": 2.0986301369863014, + "grad_norm": 3.497518539428711, + "learning_rate": 1.6978066612510155e-06, + "loss": 1.3955, + "step": 1915 + }, + { + "epoch": 2.0997260273972604, + "grad_norm": 3.47510027885437, + "learning_rate": 1.6957757920389927e-06, + "loss": 1.4383, + "step": 1916 + }, + { + "epoch": 2.100821917808219, + "grad_norm": 3.456883668899536, + "learning_rate": 1.69374492282697e-06, + "loss": 1.3675, + "step": 1917 + }, + { + "epoch": 2.101917808219178, + "grad_norm": 3.441713809967041, + "learning_rate": 1.6917140536149473e-06, + "loss": 1.354, + "step": 1918 + }, + { + "epoch": 2.103013698630137, + "grad_norm": 3.209798574447632, + "learning_rate": 1.6896831844029246e-06, + "loss": 1.3516, + "step": 1919 + }, + { + "epoch": 2.1041095890410957, + "grad_norm": 3.3628475666046143, + "learning_rate": 1.6876523151909018e-06, + "loss": 1.4311, + "step": 1920 + }, + { + "epoch": 2.1041095890410957, + "eval_loss": 1.6944255828857422, + "eval_runtime": 37.0053, + "eval_samples_per_second": 27.023, + "eval_steps_per_second": 1.135, + "step": 1920 + }, + { + "epoch": 2.1052054794520547, + "grad_norm": 3.2812838554382324, + "learning_rate": 1.6856214459788791e-06, + "loss": 1.3503, + "step": 1921 + }, + { + "epoch": 2.1063013698630138, + "grad_norm": 3.396134853363037, + "learning_rate": 1.6835905767668564e-06, + "loss": 1.376, + "step": 1922 + }, + { + "epoch": 2.107397260273973, + "grad_norm": 3.4108328819274902, + "learning_rate": 1.6815597075548337e-06, + "loss": 1.3429, + "step": 1923 + }, + { + "epoch": 2.1084931506849314, + "grad_norm": 3.3519976139068604, + "learning_rate": 1.679528838342811e-06, + "loss": 1.3475, + "step": 1924 + }, + { + "epoch": 2.1095890410958904, + "grad_norm": 3.337141275405884, + "learning_rate": 1.677497969130788e-06, + "loss": 1.4016, + "step": 1925 + }, + { + "epoch": 2.1106849315068494, + "grad_norm": 3.353137493133545, + "learning_rate": 1.6754670999187653e-06, + "loss": 1.3634, + "step": 1926 + }, + { + "epoch": 2.111780821917808, + "grad_norm": 3.2123942375183105, + "learning_rate": 1.6734362307067426e-06, + "loss": 1.362, + "step": 1927 + }, + { + "epoch": 2.112876712328767, + "grad_norm": 3.293612003326416, + "learning_rate": 1.6714053614947198e-06, + "loss": 1.3702, + "step": 1928 + }, + { + "epoch": 2.113972602739726, + "grad_norm": 3.1277008056640625, + "learning_rate": 1.669374492282697e-06, + "loss": 1.3277, + "step": 1929 + }, + { + "epoch": 2.115068493150685, + "grad_norm": 3.202958345413208, + "learning_rate": 1.6673436230706744e-06, + "loss": 1.3018, + "step": 1930 + }, + { + "epoch": 2.115068493150685, + "eval_loss": 1.699993371963501, + "eval_runtime": 36.9732, + "eval_samples_per_second": 27.047, + "eval_steps_per_second": 1.136, + "step": 1930 + }, + { + "epoch": 2.1161643835616437, + "grad_norm": 3.420098066329956, + "learning_rate": 1.6653127538586517e-06, + "loss": 1.3973, + "step": 1931 + }, + { + "epoch": 2.1172602739726027, + "grad_norm": 3.5087451934814453, + "learning_rate": 1.663281884646629e-06, + "loss": 1.3805, + "step": 1932 + }, + { + "epoch": 2.118356164383562, + "grad_norm": 3.40878963470459, + "learning_rate": 1.6612510154346062e-06, + "loss": 1.3733, + "step": 1933 + }, + { + "epoch": 2.1194520547945204, + "grad_norm": 3.2129666805267334, + "learning_rate": 1.6592201462225835e-06, + "loss": 1.3613, + "step": 1934 + }, + { + "epoch": 2.1205479452054794, + "grad_norm": 3.4247281551361084, + "learning_rate": 1.6571892770105608e-06, + "loss": 1.4109, + "step": 1935 + }, + { + "epoch": 2.1216438356164384, + "grad_norm": 3.5532023906707764, + "learning_rate": 1.655158407798538e-06, + "loss": 1.3876, + "step": 1936 + }, + { + "epoch": 2.1227397260273975, + "grad_norm": 3.413281202316284, + "learning_rate": 1.653127538586515e-06, + "loss": 1.4162, + "step": 1937 + }, + { + "epoch": 2.123835616438356, + "grad_norm": 3.4193389415740967, + "learning_rate": 1.6510966693744924e-06, + "loss": 1.3642, + "step": 1938 + }, + { + "epoch": 2.124931506849315, + "grad_norm": 3.2126522064208984, + "learning_rate": 1.6490658001624696e-06, + "loss": 1.3859, + "step": 1939 + }, + { + "epoch": 2.126027397260274, + "grad_norm": 3.340851068496704, + "learning_rate": 1.647034930950447e-06, + "loss": 1.403, + "step": 1940 + }, + { + "epoch": 2.126027397260274, + "eval_loss": 1.697942852973938, + "eval_runtime": 37.0059, + "eval_samples_per_second": 27.023, + "eval_steps_per_second": 1.135, + "step": 1940 + }, + { + "epoch": 2.1271232876712327, + "grad_norm": 3.2976479530334473, + "learning_rate": 1.6450040617384242e-06, + "loss": 1.3875, + "step": 1941 + }, + { + "epoch": 2.1282191780821917, + "grad_norm": 3.1940505504608154, + "learning_rate": 1.6429731925264015e-06, + "loss": 1.3503, + "step": 1942 + }, + { + "epoch": 2.1293150684931508, + "grad_norm": 3.3898298740386963, + "learning_rate": 1.6409423233143788e-06, + "loss": 1.3772, + "step": 1943 + }, + { + "epoch": 2.1304109589041094, + "grad_norm": 3.5407347679138184, + "learning_rate": 1.638911454102356e-06, + "loss": 1.4022, + "step": 1944 + }, + { + "epoch": 2.1315068493150684, + "grad_norm": 3.2861413955688477, + "learning_rate": 1.6368805848903333e-06, + "loss": 1.3525, + "step": 1945 + }, + { + "epoch": 2.1326027397260274, + "grad_norm": 3.1556038856506348, + "learning_rate": 1.6348497156783106e-06, + "loss": 1.3583, + "step": 1946 + }, + { + "epoch": 2.1336986301369865, + "grad_norm": 3.4463050365448, + "learning_rate": 1.6328188464662879e-06, + "loss": 1.4073, + "step": 1947 + }, + { + "epoch": 2.134794520547945, + "grad_norm": 3.38688588142395, + "learning_rate": 1.6307879772542651e-06, + "loss": 1.3507, + "step": 1948 + }, + { + "epoch": 2.135890410958904, + "grad_norm": 3.318810224533081, + "learning_rate": 1.628757108042242e-06, + "loss": 1.3578, + "step": 1949 + }, + { + "epoch": 2.136986301369863, + "grad_norm": 3.3318817615509033, + "learning_rate": 1.6267262388302193e-06, + "loss": 1.3733, + "step": 1950 + }, + { + "epoch": 2.136986301369863, + "eval_loss": 1.6968495845794678, + "eval_runtime": 37.0309, + "eval_samples_per_second": 27.004, + "eval_steps_per_second": 1.134, + "step": 1950 + }, + { + "epoch": 2.138082191780822, + "grad_norm": 3.376007556915283, + "learning_rate": 1.6246953696181965e-06, + "loss": 1.4084, + "step": 1951 + }, + { + "epoch": 2.1391780821917807, + "grad_norm": 3.332242012023926, + "learning_rate": 1.6226645004061738e-06, + "loss": 1.3515, + "step": 1952 + }, + { + "epoch": 2.1402739726027398, + "grad_norm": 3.4526102542877197, + "learning_rate": 1.620633631194151e-06, + "loss": 1.3685, + "step": 1953 + }, + { + "epoch": 2.141369863013699, + "grad_norm": 3.5033974647521973, + "learning_rate": 1.6186027619821284e-06, + "loss": 1.4413, + "step": 1954 + }, + { + "epoch": 2.1424657534246574, + "grad_norm": 3.4636034965515137, + "learning_rate": 1.6165718927701056e-06, + "loss": 1.4018, + "step": 1955 + }, + { + "epoch": 2.1435616438356164, + "grad_norm": 3.3892483711242676, + "learning_rate": 1.614541023558083e-06, + "loss": 1.3677, + "step": 1956 + }, + { + "epoch": 2.1446575342465755, + "grad_norm": 3.3363494873046875, + "learning_rate": 1.6125101543460602e-06, + "loss": 1.3942, + "step": 1957 + }, + { + "epoch": 2.145753424657534, + "grad_norm": 3.241360664367676, + "learning_rate": 1.6104792851340375e-06, + "loss": 1.3674, + "step": 1958 + }, + { + "epoch": 2.146849315068493, + "grad_norm": 3.3540799617767334, + "learning_rate": 1.6084484159220147e-06, + "loss": 1.371, + "step": 1959 + }, + { + "epoch": 2.147945205479452, + "grad_norm": 3.406891345977783, + "learning_rate": 1.606417546709992e-06, + "loss": 1.3928, + "step": 1960 + }, + { + "epoch": 2.147945205479452, + "eval_loss": 1.6973413228988647, + "eval_runtime": 37.0315, + "eval_samples_per_second": 27.004, + "eval_steps_per_second": 1.134, + "step": 1960 + }, + { + "epoch": 2.149041095890411, + "grad_norm": 3.281581401824951, + "learning_rate": 1.6043866774979693e-06, + "loss": 1.3536, + "step": 1961 + }, + { + "epoch": 2.1501369863013697, + "grad_norm": 3.3481616973876953, + "learning_rate": 1.6023558082859464e-06, + "loss": 1.3597, + "step": 1962 + }, + { + "epoch": 2.1512328767123288, + "grad_norm": 3.2706754207611084, + "learning_rate": 1.6003249390739236e-06, + "loss": 1.3677, + "step": 1963 + }, + { + "epoch": 2.152328767123288, + "grad_norm": 3.117994785308838, + "learning_rate": 1.598294069861901e-06, + "loss": 1.3336, + "step": 1964 + }, + { + "epoch": 2.1534246575342464, + "grad_norm": 3.3349826335906982, + "learning_rate": 1.5962632006498782e-06, + "loss": 1.3793, + "step": 1965 + }, + { + "epoch": 2.1545205479452054, + "grad_norm": 3.3232476711273193, + "learning_rate": 1.5942323314378555e-06, + "loss": 1.3424, + "step": 1966 + }, + { + "epoch": 2.1556164383561645, + "grad_norm": 3.4362082481384277, + "learning_rate": 1.5922014622258327e-06, + "loss": 1.4164, + "step": 1967 + }, + { + "epoch": 2.1567123287671235, + "grad_norm": 3.319347381591797, + "learning_rate": 1.59017059301381e-06, + "loss": 1.3398, + "step": 1968 + }, + { + "epoch": 2.157808219178082, + "grad_norm": 3.3089616298675537, + "learning_rate": 1.5881397238017873e-06, + "loss": 1.3326, + "step": 1969 + }, + { + "epoch": 2.158904109589041, + "grad_norm": 3.4224839210510254, + "learning_rate": 1.5861088545897646e-06, + "loss": 1.3824, + "step": 1970 + }, + { + "epoch": 2.158904109589041, + "eval_loss": 1.6987097263336182, + "eval_runtime": 37.0446, + "eval_samples_per_second": 26.994, + "eval_steps_per_second": 1.134, + "step": 1970 + }, + { + "epoch": 2.16, + "grad_norm": 3.4151504039764404, + "learning_rate": 1.5840779853777418e-06, + "loss": 1.3707, + "step": 1971 + }, + { + "epoch": 2.1610958904109587, + "grad_norm": 3.340728998184204, + "learning_rate": 1.5820471161657191e-06, + "loss": 1.3405, + "step": 1972 + }, + { + "epoch": 2.1621917808219178, + "grad_norm": 3.312610626220703, + "learning_rate": 1.5800162469536964e-06, + "loss": 1.3733, + "step": 1973 + }, + { + "epoch": 2.163287671232877, + "grad_norm": 3.305577278137207, + "learning_rate": 1.5779853777416735e-06, + "loss": 1.355, + "step": 1974 + }, + { + "epoch": 2.1643835616438354, + "grad_norm": 3.5228960514068604, + "learning_rate": 1.5759545085296507e-06, + "loss": 1.3909, + "step": 1975 + }, + { + "epoch": 2.1654794520547944, + "grad_norm": 3.2608871459960938, + "learning_rate": 1.573923639317628e-06, + "loss": 1.3841, + "step": 1976 + }, + { + "epoch": 2.1665753424657535, + "grad_norm": 3.5178403854370117, + "learning_rate": 1.5718927701056053e-06, + "loss": 1.3707, + "step": 1977 + }, + { + "epoch": 2.1676712328767125, + "grad_norm": 3.2331478595733643, + "learning_rate": 1.5698619008935826e-06, + "loss": 1.373, + "step": 1978 + }, + { + "epoch": 2.168767123287671, + "grad_norm": 3.5636820793151855, + "learning_rate": 1.5678310316815598e-06, + "loss": 1.3982, + "step": 1979 + }, + { + "epoch": 2.16986301369863, + "grad_norm": 3.31272292137146, + "learning_rate": 1.5658001624695371e-06, + "loss": 1.3949, + "step": 1980 + }, + { + "epoch": 2.16986301369863, + "eval_loss": 1.694874882698059, + "eval_runtime": 37.0818, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 1980 + }, + { + "epoch": 2.170958904109589, + "grad_norm": 3.300340414047241, + "learning_rate": 1.5637692932575144e-06, + "loss": 1.345, + "step": 1981 + }, + { + "epoch": 2.172054794520548, + "grad_norm": 3.410402297973633, + "learning_rate": 1.5617384240454917e-06, + "loss": 1.389, + "step": 1982 + }, + { + "epoch": 2.1731506849315068, + "grad_norm": 3.3093440532684326, + "learning_rate": 1.559707554833469e-06, + "loss": 1.3646, + "step": 1983 + }, + { + "epoch": 2.174246575342466, + "grad_norm": 3.336097478866577, + "learning_rate": 1.5576766856214462e-06, + "loss": 1.385, + "step": 1984 + }, + { + "epoch": 2.175342465753425, + "grad_norm": 3.290992498397827, + "learning_rate": 1.5556458164094235e-06, + "loss": 1.4105, + "step": 1985 + }, + { + "epoch": 2.1764383561643834, + "grad_norm": 3.1935672760009766, + "learning_rate": 1.5536149471974008e-06, + "loss": 1.3132, + "step": 1986 + }, + { + "epoch": 2.1775342465753424, + "grad_norm": 3.167464256286621, + "learning_rate": 1.5515840779853778e-06, + "loss": 1.3518, + "step": 1987 + }, + { + "epoch": 2.1786301369863015, + "grad_norm": 3.312204122543335, + "learning_rate": 1.549553208773355e-06, + "loss": 1.3251, + "step": 1988 + }, + { + "epoch": 2.17972602739726, + "grad_norm": 3.2165799140930176, + "learning_rate": 1.5475223395613324e-06, + "loss": 1.3574, + "step": 1989 + }, + { + "epoch": 2.180821917808219, + "grad_norm": 3.4713053703308105, + "learning_rate": 1.5454914703493097e-06, + "loss": 1.3526, + "step": 1990 + }, + { + "epoch": 2.180821917808219, + "eval_loss": 1.6958508491516113, + "eval_runtime": 37.342, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 1990 + }, + { + "epoch": 2.181917808219178, + "grad_norm": 3.424330949783325, + "learning_rate": 1.543460601137287e-06, + "loss": 1.3465, + "step": 1991 + }, + { + "epoch": 2.183013698630137, + "grad_norm": 3.4916937351226807, + "learning_rate": 1.5414297319252642e-06, + "loss": 1.3988, + "step": 1992 + }, + { + "epoch": 2.1841095890410958, + "grad_norm": 3.4839375019073486, + "learning_rate": 1.5393988627132415e-06, + "loss": 1.3595, + "step": 1993 + }, + { + "epoch": 2.185205479452055, + "grad_norm": 3.3232152462005615, + "learning_rate": 1.5373679935012188e-06, + "loss": 1.3744, + "step": 1994 + }, + { + "epoch": 2.186301369863014, + "grad_norm": 3.499152421951294, + "learning_rate": 1.535337124289196e-06, + "loss": 1.3694, + "step": 1995 + }, + { + "epoch": 2.1873972602739724, + "grad_norm": 3.46787166595459, + "learning_rate": 1.5333062550771733e-06, + "loss": 1.3125, + "step": 1996 + }, + { + "epoch": 2.1884931506849314, + "grad_norm": 3.3333165645599365, + "learning_rate": 1.5312753858651506e-06, + "loss": 1.3464, + "step": 1997 + }, + { + "epoch": 2.1895890410958905, + "grad_norm": 3.5153214931488037, + "learning_rate": 1.5292445166531279e-06, + "loss": 1.435, + "step": 1998 + }, + { + "epoch": 2.1906849315068495, + "grad_norm": 3.342599630355835, + "learning_rate": 1.5272136474411047e-06, + "loss": 1.3935, + "step": 1999 + }, + { + "epoch": 2.191780821917808, + "grad_norm": 3.4318430423736572, + "learning_rate": 1.525182778229082e-06, + "loss": 1.3702, + "step": 2000 + }, + { + "epoch": 2.191780821917808, + "eval_loss": 1.6950546503067017, + "eval_runtime": 37.4816, + "eval_samples_per_second": 26.68, + "eval_steps_per_second": 1.121, + "step": 2000 + }, + { + "epoch": 2.192876712328767, + "grad_norm": 3.4610133171081543, + "learning_rate": 1.5231519090170593e-06, + "loss": 1.4065, + "step": 2001 + }, + { + "epoch": 2.193972602739726, + "grad_norm": 3.3207743167877197, + "learning_rate": 1.5211210398050365e-06, + "loss": 1.393, + "step": 2002 + }, + { + "epoch": 2.1950684931506848, + "grad_norm": 3.3207743167877197, + "learning_rate": 1.5211210398050365e-06, + "loss": 1.368, + "step": 2003 + }, + { + "epoch": 2.196164383561644, + "grad_norm": 3.2459397315979004, + "learning_rate": 1.5190901705930138e-06, + "loss": 1.341, + "step": 2004 + }, + { + "epoch": 2.197260273972603, + "grad_norm": 3.2874672412872314, + "learning_rate": 1.517059301380991e-06, + "loss": 1.386, + "step": 2005 + }, + { + "epoch": 2.198356164383562, + "grad_norm": 3.413085460662842, + "learning_rate": 1.5150284321689684e-06, + "loss": 1.3573, + "step": 2006 + }, + { + "epoch": 2.1994520547945204, + "grad_norm": 3.332045793533325, + "learning_rate": 1.5129975629569456e-06, + "loss": 1.3646, + "step": 2007 + }, + { + "epoch": 2.2005479452054795, + "grad_norm": 3.4042816162109375, + "learning_rate": 1.510966693744923e-06, + "loss": 1.3449, + "step": 2008 + }, + { + "epoch": 2.2016438356164385, + "grad_norm": 3.5643274784088135, + "learning_rate": 1.5089358245329002e-06, + "loss": 1.3718, + "step": 2009 + }, + { + "epoch": 2.202739726027397, + "grad_norm": 3.2719311714172363, + "learning_rate": 1.5069049553208775e-06, + "loss": 1.3372, + "step": 2010 + }, + { + "epoch": 2.202739726027397, + "eval_loss": 1.6964045763015747, + "eval_runtime": 36.9263, + "eval_samples_per_second": 27.081, + "eval_steps_per_second": 1.137, + "step": 2010 + }, + { + "epoch": 2.203835616438356, + "grad_norm": 3.3720433712005615, + "learning_rate": 1.5048740861088547e-06, + "loss": 1.3691, + "step": 2011 + }, + { + "epoch": 2.204931506849315, + "grad_norm": 3.3895585536956787, + "learning_rate": 1.502843216896832e-06, + "loss": 1.3857, + "step": 2012 + }, + { + "epoch": 2.206027397260274, + "grad_norm": 3.218393087387085, + "learning_rate": 1.500812347684809e-06, + "loss": 1.3155, + "step": 2013 + }, + { + "epoch": 2.207123287671233, + "grad_norm": 3.3249216079711914, + "learning_rate": 1.4987814784727864e-06, + "loss": 1.3736, + "step": 2014 + }, + { + "epoch": 2.208219178082192, + "grad_norm": 3.30904483795166, + "learning_rate": 1.4967506092607636e-06, + "loss": 1.354, + "step": 2015 + }, + { + "epoch": 2.209315068493151, + "grad_norm": 3.1706366539001465, + "learning_rate": 1.494719740048741e-06, + "loss": 1.345, + "step": 2016 + }, + { + "epoch": 2.2104109589041094, + "grad_norm": 3.4368274211883545, + "learning_rate": 1.4926888708367182e-06, + "loss": 1.3699, + "step": 2017 + }, + { + "epoch": 2.2115068493150685, + "grad_norm": 3.364492893218994, + "learning_rate": 1.4906580016246955e-06, + "loss": 1.3911, + "step": 2018 + }, + { + "epoch": 2.2126027397260275, + "grad_norm": 3.397193193435669, + "learning_rate": 1.4886271324126727e-06, + "loss": 1.3927, + "step": 2019 + }, + { + "epoch": 2.213698630136986, + "grad_norm": 3.432245969772339, + "learning_rate": 1.48659626320065e-06, + "loss": 1.4013, + "step": 2020 + }, + { + "epoch": 2.213698630136986, + "eval_loss": 1.6935715675354004, + "eval_runtime": 37.0955, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 2020 + }, + { + "epoch": 2.214794520547945, + "grad_norm": 3.2825889587402344, + "learning_rate": 1.4845653939886273e-06, + "loss": 1.2974, + "step": 2021 + }, + { + "epoch": 2.215890410958904, + "grad_norm": 3.296154737472534, + "learning_rate": 1.4825345247766046e-06, + "loss": 1.3127, + "step": 2022 + }, + { + "epoch": 2.216986301369863, + "grad_norm": 3.4842185974121094, + "learning_rate": 1.4805036555645818e-06, + "loss": 1.3201, + "step": 2023 + }, + { + "epoch": 2.2180821917808218, + "grad_norm": 3.314819097518921, + "learning_rate": 1.4784727863525591e-06, + "loss": 1.3519, + "step": 2024 + }, + { + "epoch": 2.219178082191781, + "grad_norm": 3.207789421081543, + "learning_rate": 1.4764419171405362e-06, + "loss": 1.324, + "step": 2025 + }, + { + "epoch": 2.22027397260274, + "grad_norm": 3.516042470932007, + "learning_rate": 1.4744110479285135e-06, + "loss": 1.3828, + "step": 2026 + }, + { + "epoch": 2.2213698630136984, + "grad_norm": 3.556605339050293, + "learning_rate": 1.4723801787164907e-06, + "loss": 1.3954, + "step": 2027 + }, + { + "epoch": 2.2224657534246575, + "grad_norm": 3.4396398067474365, + "learning_rate": 1.470349309504468e-06, + "loss": 1.387, + "step": 2028 + }, + { + "epoch": 2.2235616438356165, + "grad_norm": 3.391446352005005, + "learning_rate": 1.4683184402924453e-06, + "loss": 1.3658, + "step": 2029 + }, + { + "epoch": 2.2246575342465755, + "grad_norm": 3.4033215045928955, + "learning_rate": 1.4662875710804226e-06, + "loss": 1.3995, + "step": 2030 + }, + { + "epoch": 2.2246575342465755, + "eval_loss": 1.6942157745361328, + "eval_runtime": 37.0973, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 2030 + }, + { + "epoch": 2.225753424657534, + "grad_norm": 3.4440598487854004, + "learning_rate": 1.4642567018683998e-06, + "loss": 1.4188, + "step": 2031 + }, + { + "epoch": 2.226849315068493, + "grad_norm": 3.4236795902252197, + "learning_rate": 1.4622258326563771e-06, + "loss": 1.3592, + "step": 2032 + }, + { + "epoch": 2.227945205479452, + "grad_norm": 3.369434118270874, + "learning_rate": 1.4601949634443544e-06, + "loss": 1.3527, + "step": 2033 + }, + { + "epoch": 2.2290410958904108, + "grad_norm": 3.3993430137634277, + "learning_rate": 1.4581640942323317e-06, + "loss": 1.3502, + "step": 2034 + }, + { + "epoch": 2.23013698630137, + "grad_norm": 3.4025020599365234, + "learning_rate": 1.456133225020309e-06, + "loss": 1.4365, + "step": 2035 + }, + { + "epoch": 2.231232876712329, + "grad_norm": 3.418386459350586, + "learning_rate": 1.4541023558082862e-06, + "loss": 1.3722, + "step": 2036 + }, + { + "epoch": 2.232328767123288, + "grad_norm": 3.357462167739868, + "learning_rate": 1.4520714865962635e-06, + "loss": 1.3568, + "step": 2037 + }, + { + "epoch": 2.2334246575342465, + "grad_norm": 3.590677261352539, + "learning_rate": 1.4500406173842406e-06, + "loss": 1.3913, + "step": 2038 + }, + { + "epoch": 2.2345205479452055, + "grad_norm": 3.531836986541748, + "learning_rate": 1.4480097481722178e-06, + "loss": 1.3635, + "step": 2039 + }, + { + "epoch": 2.2356164383561645, + "grad_norm": 3.505734920501709, + "learning_rate": 1.4459788789601951e-06, + "loss": 1.3793, + "step": 2040 + }, + { + "epoch": 2.2356164383561645, + "eval_loss": 1.6933643817901611, + "eval_runtime": 37.0972, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 2040 + }, + { + "epoch": 2.236712328767123, + "grad_norm": 3.336024045944214, + "learning_rate": 1.4439480097481724e-06, + "loss": 1.3456, + "step": 2041 + }, + { + "epoch": 2.237808219178082, + "grad_norm": 3.302778720855713, + "learning_rate": 1.4419171405361497e-06, + "loss": 1.3352, + "step": 2042 + }, + { + "epoch": 2.238904109589041, + "grad_norm": 3.446685791015625, + "learning_rate": 1.439886271324127e-06, + "loss": 1.3763, + "step": 2043 + }, + { + "epoch": 2.24, + "grad_norm": 3.3545477390289307, + "learning_rate": 1.4378554021121042e-06, + "loss": 1.3755, + "step": 2044 + }, + { + "epoch": 2.241095890410959, + "grad_norm": 3.3778200149536133, + "learning_rate": 1.4358245329000815e-06, + "loss": 1.3583, + "step": 2045 + }, + { + "epoch": 2.242191780821918, + "grad_norm": 3.293827533721924, + "learning_rate": 1.4337936636880588e-06, + "loss": 1.3729, + "step": 2046 + }, + { + "epoch": 2.243287671232877, + "grad_norm": 3.2763779163360596, + "learning_rate": 1.431762794476036e-06, + "loss": 1.3968, + "step": 2047 + }, + { + "epoch": 2.2443835616438355, + "grad_norm": 3.3150572776794434, + "learning_rate": 1.4297319252640133e-06, + "loss": 1.3293, + "step": 2048 + }, + { + "epoch": 2.2454794520547945, + "grad_norm": 3.4018642902374268, + "learning_rate": 1.4277010560519906e-06, + "loss": 1.3563, + "step": 2049 + }, + { + "epoch": 2.2465753424657535, + "grad_norm": 3.3171231746673584, + "learning_rate": 1.4256701868399674e-06, + "loss": 1.3486, + "step": 2050 + }, + { + "epoch": 2.2465753424657535, + "eval_loss": 1.6919909715652466, + "eval_runtime": 37.1277, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 2050 + }, + { + "epoch": 2.247671232876712, + "grad_norm": 3.395368814468384, + "learning_rate": 1.4236393176279447e-06, + "loss": 1.3946, + "step": 2051 + }, + { + "epoch": 2.248767123287671, + "grad_norm": 3.312668800354004, + "learning_rate": 1.421608448415922e-06, + "loss": 1.3682, + "step": 2052 + }, + { + "epoch": 2.24986301369863, + "grad_norm": 3.474766254425049, + "learning_rate": 1.4195775792038993e-06, + "loss": 1.4361, + "step": 2053 + }, + { + "epoch": 2.250958904109589, + "grad_norm": 3.337005138397217, + "learning_rate": 1.4175467099918765e-06, + "loss": 1.3526, + "step": 2054 + }, + { + "epoch": 2.252054794520548, + "grad_norm": 3.3669068813323975, + "learning_rate": 1.4155158407798538e-06, + "loss": 1.3606, + "step": 2055 + }, + { + "epoch": 2.253150684931507, + "grad_norm": 3.4782967567443848, + "learning_rate": 1.413484971567831e-06, + "loss": 1.3918, + "step": 2056 + }, + { + "epoch": 2.254246575342466, + "grad_norm": 3.288242816925049, + "learning_rate": 1.4114541023558084e-06, + "loss": 1.3111, + "step": 2057 + }, + { + "epoch": 2.255342465753425, + "grad_norm": 3.452920436859131, + "learning_rate": 1.4094232331437856e-06, + "loss": 1.3544, + "step": 2058 + }, + { + "epoch": 2.2564383561643835, + "grad_norm": 3.3995440006256104, + "learning_rate": 1.407392363931763e-06, + "loss": 1.3885, + "step": 2059 + }, + { + "epoch": 2.2575342465753425, + "grad_norm": 3.3539721965789795, + "learning_rate": 1.4053614947197402e-06, + "loss": 1.4397, + "step": 2060 + }, + { + "epoch": 2.2575342465753425, + "eval_loss": 1.6925660371780396, + "eval_runtime": 37.0945, + "eval_samples_per_second": 26.958, + "eval_steps_per_second": 1.132, + "step": 2060 + }, + { + "epoch": 2.2586301369863016, + "grad_norm": 3.4854390621185303, + "learning_rate": 1.4033306255077175e-06, + "loss": 1.356, + "step": 2061 + }, + { + "epoch": 2.25972602739726, + "grad_norm": 3.224963426589966, + "learning_rate": 1.4012997562956945e-06, + "loss": 1.3185, + "step": 2062 + }, + { + "epoch": 2.260821917808219, + "grad_norm": 3.423572063446045, + "learning_rate": 1.3992688870836718e-06, + "loss": 1.4013, + "step": 2063 + }, + { + "epoch": 2.261917808219178, + "grad_norm": 3.409107208251953, + "learning_rate": 1.397238017871649e-06, + "loss": 1.3921, + "step": 2064 + }, + { + "epoch": 2.263013698630137, + "grad_norm": 3.430488348007202, + "learning_rate": 1.3952071486596264e-06, + "loss": 1.3821, + "step": 2065 + }, + { + "epoch": 2.264109589041096, + "grad_norm": 3.3903696537017822, + "learning_rate": 1.3931762794476036e-06, + "loss": 1.3435, + "step": 2066 + }, + { + "epoch": 2.265205479452055, + "grad_norm": 3.451477289199829, + "learning_rate": 1.391145410235581e-06, + "loss": 1.4008, + "step": 2067 + }, + { + "epoch": 2.266301369863014, + "grad_norm": 3.3731205463409424, + "learning_rate": 1.3891145410235582e-06, + "loss": 1.3587, + "step": 2068 + }, + { + "epoch": 2.2673972602739725, + "grad_norm": 3.323897123336792, + "learning_rate": 1.3870836718115355e-06, + "loss": 1.2697, + "step": 2069 + }, + { + "epoch": 2.2684931506849315, + "grad_norm": 3.4095637798309326, + "learning_rate": 1.3850528025995127e-06, + "loss": 1.3705, + "step": 2070 + }, + { + "epoch": 2.2684931506849315, + "eval_loss": 1.6916862726211548, + "eval_runtime": 37.0917, + "eval_samples_per_second": 26.96, + "eval_steps_per_second": 1.132, + "step": 2070 + }, + { + "epoch": 2.2695890410958905, + "grad_norm": 3.4273080825805664, + "learning_rate": 1.38302193338749e-06, + "loss": 1.3624, + "step": 2071 + }, + { + "epoch": 2.270684931506849, + "grad_norm": 3.337531566619873, + "learning_rate": 1.3809910641754673e-06, + "loss": 1.2882, + "step": 2072 + }, + { + "epoch": 2.271780821917808, + "grad_norm": 3.47322678565979, + "learning_rate": 1.3789601949634446e-06, + "loss": 1.3818, + "step": 2073 + }, + { + "epoch": 2.272876712328767, + "grad_norm": 3.518423557281494, + "learning_rate": 1.3769293257514218e-06, + "loss": 1.3371, + "step": 2074 + }, + { + "epoch": 2.2739726027397262, + "grad_norm": 3.423983573913574, + "learning_rate": 1.374898456539399e-06, + "loss": 1.3667, + "step": 2075 + }, + { + "epoch": 2.275068493150685, + "grad_norm": 3.240217924118042, + "learning_rate": 1.3728675873273762e-06, + "loss": 1.328, + "step": 2076 + }, + { + "epoch": 2.276164383561644, + "grad_norm": 3.45780348777771, + "learning_rate": 1.3708367181153535e-06, + "loss": 1.3949, + "step": 2077 + }, + { + "epoch": 2.277260273972603, + "grad_norm": 3.4683196544647217, + "learning_rate": 1.3688058489033307e-06, + "loss": 1.3581, + "step": 2078 + }, + { + "epoch": 2.2783561643835615, + "grad_norm": 3.5016560554504395, + "learning_rate": 1.366774979691308e-06, + "loss": 1.373, + "step": 2079 + }, + { + "epoch": 2.2794520547945205, + "grad_norm": 3.48295521736145, + "learning_rate": 1.3647441104792853e-06, + "loss": 1.3943, + "step": 2080 + }, + { + "epoch": 2.2794520547945205, + "eval_loss": 1.6909493207931519, + "eval_runtime": 37.0602, + "eval_samples_per_second": 26.983, + "eval_steps_per_second": 1.133, + "step": 2080 + }, + { + "epoch": 2.2805479452054795, + "grad_norm": 3.4415335655212402, + "learning_rate": 1.3627132412672626e-06, + "loss": 1.3657, + "step": 2081 + }, + { + "epoch": 2.281643835616438, + "grad_norm": 3.352210760116577, + "learning_rate": 1.3606823720552398e-06, + "loss": 1.3334, + "step": 2082 + }, + { + "epoch": 2.282739726027397, + "grad_norm": 3.296957492828369, + "learning_rate": 1.3586515028432171e-06, + "loss": 1.3634, + "step": 2083 + }, + { + "epoch": 2.283835616438356, + "grad_norm": 3.3552114963531494, + "learning_rate": 1.3566206336311944e-06, + "loss": 1.3171, + "step": 2084 + }, + { + "epoch": 2.2849315068493152, + "grad_norm": 3.3183212280273438, + "learning_rate": 1.3545897644191717e-06, + "loss": 1.3627, + "step": 2085 + }, + { + "epoch": 2.286027397260274, + "grad_norm": 3.275786876678467, + "learning_rate": 1.352558895207149e-06, + "loss": 1.3566, + "step": 2086 + }, + { + "epoch": 2.287123287671233, + "grad_norm": 3.384044885635376, + "learning_rate": 1.3505280259951258e-06, + "loss": 1.3934, + "step": 2087 + }, + { + "epoch": 2.288219178082192, + "grad_norm": 3.5280847549438477, + "learning_rate": 1.348497156783103e-06, + "loss": 1.3666, + "step": 2088 + }, + { + "epoch": 2.289315068493151, + "grad_norm": 3.416024923324585, + "learning_rate": 1.3464662875710804e-06, + "loss": 1.389, + "step": 2089 + }, + { + "epoch": 2.2904109589041095, + "grad_norm": 3.4078176021575928, + "learning_rate": 1.3444354183590576e-06, + "loss": 1.3456, + "step": 2090 + }, + { + "epoch": 2.2904109589041095, + "eval_loss": 1.6899075508117676, + "eval_runtime": 37.147, + "eval_samples_per_second": 26.92, + "eval_steps_per_second": 1.131, + "step": 2090 + }, + { + "epoch": 2.2915068493150685, + "grad_norm": 3.306574821472168, + "learning_rate": 1.342404549147035e-06, + "loss": 1.3912, + "step": 2091 + }, + { + "epoch": 2.2926027397260276, + "grad_norm": 3.313765048980713, + "learning_rate": 1.3403736799350122e-06, + "loss": 1.3607, + "step": 2092 + }, + { + "epoch": 2.293698630136986, + "grad_norm": 3.3027026653289795, + "learning_rate": 1.3383428107229895e-06, + "loss": 1.3798, + "step": 2093 + }, + { + "epoch": 2.294794520547945, + "grad_norm": 3.4547550678253174, + "learning_rate": 1.3363119415109667e-06, + "loss": 1.3973, + "step": 2094 + }, + { + "epoch": 2.2958904109589042, + "grad_norm": 3.4237265586853027, + "learning_rate": 1.334281072298944e-06, + "loss": 1.3909, + "step": 2095 + }, + { + "epoch": 2.296986301369863, + "grad_norm": 3.4097959995269775, + "learning_rate": 1.3322502030869213e-06, + "loss": 1.3769, + "step": 2096 + }, + { + "epoch": 2.298082191780822, + "grad_norm": 3.3365914821624756, + "learning_rate": 1.3302193338748986e-06, + "loss": 1.3755, + "step": 2097 + }, + { + "epoch": 2.299178082191781, + "grad_norm": 3.420936107635498, + "learning_rate": 1.3281884646628758e-06, + "loss": 1.3912, + "step": 2098 + }, + { + "epoch": 2.30027397260274, + "grad_norm": 3.293266773223877, + "learning_rate": 1.3261575954508531e-06, + "loss": 1.384, + "step": 2099 + }, + { + "epoch": 2.3013698630136985, + "grad_norm": 3.3154945373535156, + "learning_rate": 1.3241267262388302e-06, + "loss": 1.3624, + "step": 2100 + }, + { + "epoch": 2.3013698630136985, + "eval_loss": 1.6889803409576416, + "eval_runtime": 37.33, + "eval_samples_per_second": 26.788, + "eval_steps_per_second": 1.125, + "step": 2100 + }, + { + "epoch": 2.3024657534246575, + "grad_norm": 3.341747760772705, + "learning_rate": 1.3220958570268074e-06, + "loss": 1.3658, + "step": 2101 + }, + { + "epoch": 2.3035616438356166, + "grad_norm": 3.302058458328247, + "learning_rate": 1.3200649878147847e-06, + "loss": 1.3085, + "step": 2102 + }, + { + "epoch": 2.304657534246575, + "grad_norm": 3.363069772720337, + "learning_rate": 1.318034118602762e-06, + "loss": 1.3466, + "step": 2103 + }, + { + "epoch": 2.305753424657534, + "grad_norm": 3.4709537029266357, + "learning_rate": 1.3160032493907393e-06, + "loss": 1.392, + "step": 2104 + }, + { + "epoch": 2.3068493150684932, + "grad_norm": 3.3935465812683105, + "learning_rate": 1.3139723801787166e-06, + "loss": 1.3852, + "step": 2105 + }, + { + "epoch": 2.3079452054794523, + "grad_norm": 3.4387412071228027, + "learning_rate": 1.3119415109666938e-06, + "loss": 1.3891, + "step": 2106 + }, + { + "epoch": 2.309041095890411, + "grad_norm": 3.3640964031219482, + "learning_rate": 1.309910641754671e-06, + "loss": 1.367, + "step": 2107 + }, + { + "epoch": 2.31013698630137, + "grad_norm": 3.3887405395507812, + "learning_rate": 1.3078797725426484e-06, + "loss": 1.3265, + "step": 2108 + }, + { + "epoch": 2.311232876712329, + "grad_norm": 3.381732702255249, + "learning_rate": 1.3058489033306257e-06, + "loss": 1.34, + "step": 2109 + }, + { + "epoch": 2.3123287671232875, + "grad_norm": 3.42246413230896, + "learning_rate": 1.303818034118603e-06, + "loss": 1.3508, + "step": 2110 + }, + { + "epoch": 2.3123287671232875, + "eval_loss": 1.693956732749939, + "eval_runtime": 36.8725, + "eval_samples_per_second": 27.12, + "eval_steps_per_second": 1.139, + "step": 2110 + }, + { + "epoch": 2.3134246575342465, + "grad_norm": 3.467451572418213, + "learning_rate": 1.3017871649065802e-06, + "loss": 1.3586, + "step": 2111 + }, + { + "epoch": 2.3145205479452056, + "grad_norm": 3.387568712234497, + "learning_rate": 1.2997562956945573e-06, + "loss": 1.3654, + "step": 2112 + }, + { + "epoch": 2.315616438356164, + "grad_norm": 3.500000238418579, + "learning_rate": 1.2977254264825345e-06, + "loss": 1.3396, + "step": 2113 + }, + { + "epoch": 2.316712328767123, + "grad_norm": 3.5994179248809814, + "learning_rate": 1.2956945572705118e-06, + "loss": 1.3409, + "step": 2114 + }, + { + "epoch": 2.317808219178082, + "grad_norm": 3.469545841217041, + "learning_rate": 1.293663688058489e-06, + "loss": 1.3481, + "step": 2115 + }, + { + "epoch": 2.3189041095890413, + "grad_norm": 3.3346712589263916, + "learning_rate": 1.2916328188464664e-06, + "loss": 1.3381, + "step": 2116 + }, + { + "epoch": 2.32, + "grad_norm": 3.4600489139556885, + "learning_rate": 1.2896019496344436e-06, + "loss": 1.3076, + "step": 2117 + }, + { + "epoch": 2.321095890410959, + "grad_norm": 3.407301425933838, + "learning_rate": 1.287571080422421e-06, + "loss": 1.3375, + "step": 2118 + }, + { + "epoch": 2.322191780821918, + "grad_norm": 3.3791961669921875, + "learning_rate": 1.2855402112103982e-06, + "loss": 1.3827, + "step": 2119 + }, + { + "epoch": 2.323287671232877, + "grad_norm": 3.2421553134918213, + "learning_rate": 1.2835093419983755e-06, + "loss": 1.3389, + "step": 2120 + }, + { + "epoch": 2.323287671232877, + "eval_loss": 1.6887240409851074, + "eval_runtime": 36.9945, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 2120 + }, + { + "epoch": 2.3243835616438355, + "grad_norm": 3.451120376586914, + "learning_rate": 1.2814784727863528e-06, + "loss": 1.3244, + "step": 2121 + }, + { + "epoch": 2.3254794520547946, + "grad_norm": 3.2467851638793945, + "learning_rate": 1.27944760357433e-06, + "loss": 1.3448, + "step": 2122 + }, + { + "epoch": 2.3265753424657536, + "grad_norm": 3.387474536895752, + "learning_rate": 1.2774167343623073e-06, + "loss": 1.3957, + "step": 2123 + }, + { + "epoch": 2.327671232876712, + "grad_norm": 3.455059289932251, + "learning_rate": 1.2753858651502846e-06, + "loss": 1.3984, + "step": 2124 + }, + { + "epoch": 2.328767123287671, + "grad_norm": 3.5118813514709473, + "learning_rate": 1.2733549959382616e-06, + "loss": 1.3268, + "step": 2125 + }, + { + "epoch": 2.3298630136986302, + "grad_norm": 3.427954912185669, + "learning_rate": 1.271324126726239e-06, + "loss": 1.336, + "step": 2126 + }, + { + "epoch": 2.330958904109589, + "grad_norm": 3.4469428062438965, + "learning_rate": 1.2692932575142162e-06, + "loss": 1.3739, + "step": 2127 + }, + { + "epoch": 2.332054794520548, + "grad_norm": 3.354433536529541, + "learning_rate": 1.2672623883021935e-06, + "loss": 1.312, + "step": 2128 + }, + { + "epoch": 2.333150684931507, + "grad_norm": 3.2023873329162598, + "learning_rate": 1.2652315190901707e-06, + "loss": 1.3154, + "step": 2129 + }, + { + "epoch": 2.334246575342466, + "grad_norm": 3.4237990379333496, + "learning_rate": 1.263200649878148e-06, + "loss": 1.3817, + "step": 2130 + }, + { + "epoch": 2.334246575342466, + "eval_loss": 1.6899415254592896, + "eval_runtime": 37.0259, + "eval_samples_per_second": 27.008, + "eval_steps_per_second": 1.134, + "step": 2130 + }, + { + "epoch": 2.3353424657534245, + "grad_norm": 3.384200096130371, + "learning_rate": 1.2611697806661253e-06, + "loss": 1.3746, + "step": 2131 + }, + { + "epoch": 2.3364383561643836, + "grad_norm": 3.3518500328063965, + "learning_rate": 1.2591389114541026e-06, + "loss": 1.3851, + "step": 2132 + }, + { + "epoch": 2.3375342465753426, + "grad_norm": 3.3043456077575684, + "learning_rate": 1.2571080422420798e-06, + "loss": 1.3663, + "step": 2133 + }, + { + "epoch": 2.338630136986301, + "grad_norm": 3.387157917022705, + "learning_rate": 1.2550771730300571e-06, + "loss": 1.3314, + "step": 2134 + }, + { + "epoch": 2.33972602739726, + "grad_norm": 3.3655288219451904, + "learning_rate": 1.2530463038180344e-06, + "loss": 1.4132, + "step": 2135 + }, + { + "epoch": 2.3408219178082192, + "grad_norm": 3.201500654220581, + "learning_rate": 1.2510154346060117e-06, + "loss": 1.3695, + "step": 2136 + }, + { + "epoch": 2.3419178082191783, + "grad_norm": 3.255248546600342, + "learning_rate": 1.2489845653939887e-06, + "loss": 1.3616, + "step": 2137 + }, + { + "epoch": 2.343013698630137, + "grad_norm": 3.4628982543945312, + "learning_rate": 1.246953696181966e-06, + "loss": 1.3921, + "step": 2138 + }, + { + "epoch": 2.344109589041096, + "grad_norm": 3.3804187774658203, + "learning_rate": 1.2449228269699433e-06, + "loss": 1.4242, + "step": 2139 + }, + { + "epoch": 2.345205479452055, + "grad_norm": 3.367130756378174, + "learning_rate": 1.2428919577579204e-06, + "loss": 1.3596, + "step": 2140 + }, + { + "epoch": 2.345205479452055, + "eval_loss": 1.6879291534423828, + "eval_runtime": 37.0536, + "eval_samples_per_second": 26.988, + "eval_steps_per_second": 1.133, + "step": 2140 + }, + { + "epoch": 2.3463013698630135, + "grad_norm": 3.3789920806884766, + "learning_rate": 1.2408610885458976e-06, + "loss": 1.3362, + "step": 2141 + }, + { + "epoch": 2.3473972602739726, + "grad_norm": 3.2852022647857666, + "learning_rate": 1.238830219333875e-06, + "loss": 1.3659, + "step": 2142 + }, + { + "epoch": 2.3484931506849316, + "grad_norm": 3.377454996109009, + "learning_rate": 1.2367993501218522e-06, + "loss": 1.3347, + "step": 2143 + }, + { + "epoch": 2.34958904109589, + "grad_norm": 3.284188985824585, + "learning_rate": 1.2347684809098295e-06, + "loss": 1.4019, + "step": 2144 + }, + { + "epoch": 2.350684931506849, + "grad_norm": 3.359066963195801, + "learning_rate": 1.2327376116978067e-06, + "loss": 1.3911, + "step": 2145 + }, + { + "epoch": 2.3517808219178082, + "grad_norm": 3.249520778656006, + "learning_rate": 1.230706742485784e-06, + "loss": 1.322, + "step": 2146 + }, + { + "epoch": 2.3528767123287673, + "grad_norm": 3.3531105518341064, + "learning_rate": 1.2286758732737613e-06, + "loss": 1.3391, + "step": 2147 + }, + { + "epoch": 2.353972602739726, + "grad_norm": 3.493309736251831, + "learning_rate": 1.2266450040617386e-06, + "loss": 1.4254, + "step": 2148 + }, + { + "epoch": 2.355068493150685, + "grad_norm": 3.554919481277466, + "learning_rate": 1.2246141348497158e-06, + "loss": 1.4103, + "step": 2149 + }, + { + "epoch": 2.356164383561644, + "grad_norm": 3.397555112838745, + "learning_rate": 1.2225832656376931e-06, + "loss": 1.3533, + "step": 2150 + }, + { + "epoch": 2.356164383561644, + "eval_loss": 1.6935802698135376, + "eval_runtime": 37.0677, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 2150 + }, + { + "epoch": 2.357260273972603, + "grad_norm": 3.40470552444458, + "learning_rate": 1.2205523964256704e-06, + "loss": 1.3726, + "step": 2151 + }, + { + "epoch": 2.3583561643835615, + "grad_norm": 3.4772186279296875, + "learning_rate": 1.2185215272136477e-06, + "loss": 1.3449, + "step": 2152 + }, + { + "epoch": 2.3594520547945206, + "grad_norm": 3.624591827392578, + "learning_rate": 1.2164906580016247e-06, + "loss": 1.375, + "step": 2153 + }, + { + "epoch": 2.3605479452054796, + "grad_norm": 3.4178595542907715, + "learning_rate": 1.214459788789602e-06, + "loss": 1.3905, + "step": 2154 + }, + { + "epoch": 2.361643835616438, + "grad_norm": 3.309950351715088, + "learning_rate": 1.2124289195775793e-06, + "loss": 1.2774, + "step": 2155 + }, + { + "epoch": 2.3627397260273972, + "grad_norm": 3.2514491081237793, + "learning_rate": 1.2103980503655566e-06, + "loss": 1.3356, + "step": 2156 + }, + { + "epoch": 2.3638356164383563, + "grad_norm": 3.3560686111450195, + "learning_rate": 1.2083671811535338e-06, + "loss": 1.3413, + "step": 2157 + }, + { + "epoch": 2.364931506849315, + "grad_norm": 3.3756141662597656, + "learning_rate": 1.2063363119415111e-06, + "loss": 1.3561, + "step": 2158 + }, + { + "epoch": 2.366027397260274, + "grad_norm": 3.288309335708618, + "learning_rate": 1.2043054427294882e-06, + "loss": 1.3476, + "step": 2159 + }, + { + "epoch": 2.367123287671233, + "grad_norm": 3.2800111770629883, + "learning_rate": 1.2022745735174654e-06, + "loss": 1.3542, + "step": 2160 + }, + { + "epoch": 2.367123287671233, + "eval_loss": 1.6914314031600952, + "eval_runtime": 37.0986, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 2160 + }, + { + "epoch": 2.368219178082192, + "grad_norm": 3.355591058731079, + "learning_rate": 1.2002437043054427e-06, + "loss": 1.3782, + "step": 2161 + }, + { + "epoch": 2.3693150684931505, + "grad_norm": 3.3569586277008057, + "learning_rate": 1.19821283509342e-06, + "loss": 1.4118, + "step": 2162 + }, + { + "epoch": 2.3704109589041096, + "grad_norm": 3.3605072498321533, + "learning_rate": 1.1961819658813973e-06, + "loss": 1.3849, + "step": 2163 + }, + { + "epoch": 2.3715068493150686, + "grad_norm": 3.401731252670288, + "learning_rate": 1.1941510966693746e-06, + "loss": 1.3629, + "step": 2164 + }, + { + "epoch": 2.3726027397260276, + "grad_norm": 3.327308177947998, + "learning_rate": 1.1921202274573518e-06, + "loss": 1.343, + "step": 2165 + }, + { + "epoch": 2.3736986301369862, + "grad_norm": 3.3242149353027344, + "learning_rate": 1.190089358245329e-06, + "loss": 1.3659, + "step": 2166 + }, + { + "epoch": 2.3747945205479453, + "grad_norm": 3.5616390705108643, + "learning_rate": 1.1880584890333064e-06, + "loss": 1.3982, + "step": 2167 + }, + { + "epoch": 2.3758904109589043, + "grad_norm": 3.3133833408355713, + "learning_rate": 1.1860276198212837e-06, + "loss": 1.3446, + "step": 2168 + }, + { + "epoch": 2.376986301369863, + "grad_norm": 3.5213828086853027, + "learning_rate": 1.183996750609261e-06, + "loss": 1.3817, + "step": 2169 + }, + { + "epoch": 2.378082191780822, + "grad_norm": 3.475719451904297, + "learning_rate": 1.1819658813972382e-06, + "loss": 1.3592, + "step": 2170 + }, + { + "epoch": 2.378082191780822, + "eval_loss": 1.6902751922607422, + "eval_runtime": 37.0792, + "eval_samples_per_second": 26.969, + "eval_steps_per_second": 1.133, + "step": 2170 + }, + { + "epoch": 2.379178082191781, + "grad_norm": 3.320626735687256, + "learning_rate": 1.1799350121852153e-06, + "loss": 1.3582, + "step": 2171 + }, + { + "epoch": 2.3802739726027395, + "grad_norm": 3.554243803024292, + "learning_rate": 1.1779041429731925e-06, + "loss": 1.3993, + "step": 2172 + }, + { + "epoch": 2.3813698630136986, + "grad_norm": 3.524912118911743, + "learning_rate": 1.1758732737611698e-06, + "loss": 1.3963, + "step": 2173 + }, + { + "epoch": 2.3824657534246576, + "grad_norm": 3.501112222671509, + "learning_rate": 1.173842404549147e-06, + "loss": 1.3207, + "step": 2174 + }, + { + "epoch": 2.383561643835616, + "grad_norm": 3.4857540130615234, + "learning_rate": 1.1718115353371244e-06, + "loss": 1.3942, + "step": 2175 + }, + { + "epoch": 2.3846575342465752, + "grad_norm": 3.3781418800354004, + "learning_rate": 1.1697806661251016e-06, + "loss": 1.3444, + "step": 2176 + }, + { + "epoch": 2.3857534246575343, + "grad_norm": 3.4888405799865723, + "learning_rate": 1.167749796913079e-06, + "loss": 1.4008, + "step": 2177 + }, + { + "epoch": 2.3868493150684933, + "grad_norm": 3.424044370651245, + "learning_rate": 1.1657189277010562e-06, + "loss": 1.3692, + "step": 2178 + }, + { + "epoch": 2.387945205479452, + "grad_norm": 3.4120898246765137, + "learning_rate": 1.1636880584890335e-06, + "loss": 1.3655, + "step": 2179 + }, + { + "epoch": 2.389041095890411, + "grad_norm": 3.52172589302063, + "learning_rate": 1.1616571892770108e-06, + "loss": 1.3662, + "step": 2180 + }, + { + "epoch": 2.389041095890411, + "eval_loss": 1.6922929286956787, + "eval_runtime": 37.0883, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 2180 + }, + { + "epoch": 2.39013698630137, + "grad_norm": 3.4904558658599854, + "learning_rate": 1.159626320064988e-06, + "loss": 1.368, + "step": 2181 + }, + { + "epoch": 2.391232876712329, + "grad_norm": 3.4123220443725586, + "learning_rate": 1.1575954508529653e-06, + "loss": 1.3785, + "step": 2182 + }, + { + "epoch": 2.3923287671232876, + "grad_norm": 3.534339427947998, + "learning_rate": 1.1555645816409426e-06, + "loss": 1.3857, + "step": 2183 + }, + { + "epoch": 2.3934246575342466, + "grad_norm": 3.3286800384521484, + "learning_rate": 1.1535337124289196e-06, + "loss": 1.3888, + "step": 2184 + }, + { + "epoch": 2.3945205479452056, + "grad_norm": 3.281928777694702, + "learning_rate": 1.151502843216897e-06, + "loss": 1.3329, + "step": 2185 + }, + { + "epoch": 2.3956164383561642, + "grad_norm": 3.3129794597625732, + "learning_rate": 1.1494719740048742e-06, + "loss": 1.3077, + "step": 2186 + }, + { + "epoch": 2.3967123287671233, + "grad_norm": 3.417710542678833, + "learning_rate": 1.1474411047928515e-06, + "loss": 1.3858, + "step": 2187 + }, + { + "epoch": 2.3978082191780823, + "grad_norm": 3.2835235595703125, + "learning_rate": 1.1454102355808287e-06, + "loss": 1.3701, + "step": 2188 + }, + { + "epoch": 2.398904109589041, + "grad_norm": 3.2926671504974365, + "learning_rate": 1.143379366368806e-06, + "loss": 1.3301, + "step": 2189 + }, + { + "epoch": 2.4, + "grad_norm": 3.393537759780884, + "learning_rate": 1.141348497156783e-06, + "loss": 1.3342, + "step": 2190 + }, + { + "epoch": 2.4, + "eval_loss": 1.6916675567626953, + "eval_runtime": 37.3782, + "eval_samples_per_second": 26.754, + "eval_steps_per_second": 1.124, + "step": 2190 + }, + { + "epoch": 2.401095890410959, + "grad_norm": 3.668956995010376, + "learning_rate": 1.1393176279447604e-06, + "loss": 1.4463, + "step": 2191 + }, + { + "epoch": 2.402191780821918, + "grad_norm": 3.4752628803253174, + "learning_rate": 1.1372867587327376e-06, + "loss": 1.4111, + "step": 2192 + }, + { + "epoch": 2.4032876712328766, + "grad_norm": 3.210810899734497, + "learning_rate": 1.135255889520715e-06, + "loss": 1.2916, + "step": 2193 + }, + { + "epoch": 2.4043835616438356, + "grad_norm": 3.5378470420837402, + "learning_rate": 1.1332250203086922e-06, + "loss": 1.3718, + "step": 2194 + }, + { + "epoch": 2.4054794520547946, + "grad_norm": 3.362588882446289, + "learning_rate": 1.1311941510966695e-06, + "loss": 1.3707, + "step": 2195 + }, + { + "epoch": 2.4065753424657537, + "grad_norm": 3.411147356033325, + "learning_rate": 1.1291632818846467e-06, + "loss": 1.3597, + "step": 2196 + }, + { + "epoch": 2.4076712328767123, + "grad_norm": 3.404151439666748, + "learning_rate": 1.127132412672624e-06, + "loss": 1.3306, + "step": 2197 + }, + { + "epoch": 2.4087671232876713, + "grad_norm": 3.306877374649048, + "learning_rate": 1.1251015434606013e-06, + "loss": 1.3509, + "step": 2198 + }, + { + "epoch": 2.4098630136986303, + "grad_norm": 3.4185047149658203, + "learning_rate": 1.1230706742485786e-06, + "loss": 1.3816, + "step": 2199 + }, + { + "epoch": 2.410958904109589, + "grad_norm": 3.573544979095459, + "learning_rate": 1.1210398050365558e-06, + "loss": 1.394, + "step": 2200 + }, + { + "epoch": 2.410958904109589, + "eval_loss": 1.6875046491622925, + "eval_runtime": 37.4533, + "eval_samples_per_second": 26.7, + "eval_steps_per_second": 1.121, + "step": 2200 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 6.858080700596224e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-2200/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/latest new file mode 100644 index 0000000000000000000000000000000000000000..e5bdf58d4f29d34e909da25905fad376f73e7c29 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/latest @@ -0,0 +1 @@ +global_step400 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..f8bd94c5541a0de92995bdd2eeeff2a85a886193 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/trainer_state.json @@ -0,0 +1,3153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.4383561643835616, + "eval_steps": 10, + "global_step": 400, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.246923763744768e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-400/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/latest new file mode 100644 index 0000000000000000000000000000000000000000..12cae1adf3af8546b4141c6f62261c8e99839a54 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/latest @@ -0,0 +1 @@ +global_step600 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..e7f6d40e08e365b40fe6641f3b8e058a3286cfca --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/trainer_state.json @@ -0,0 +1,4713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.6575342465753424, + "eval_steps": 10, + "global_step": 600, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6624906063079834, + "learning_rate": 4.77051177904143e-06, + "loss": 1.704, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 3.0075268745422363, + "learning_rate": 4.768480909829407e-06, + "loss": 1.6322, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.6070828437805176, + "learning_rate": 4.766450040617385e-06, + "loss": 1.6871, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 3.044222831726074, + "learning_rate": 4.764419171405362e-06, + "loss": 1.6652, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.773167610168457, + "learning_rate": 4.762388302193339e-06, + "loss": 1.7063, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.880176305770874, + "learning_rate": 4.760357432981316e-06, + "loss": 1.6775, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.6402764320373535, + "learning_rate": 4.758326563769294e-06, + "loss": 1.6694, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.9245846271514893, + "learning_rate": 4.756295694557271e-06, + "loss": 1.6688, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.623044967651367, + "learning_rate": 4.754264825345248e-06, + "loss": 1.6819, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.739896059036255, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.6401, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.8707164525985718, + "eval_runtime": 36.9513, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.603599786758423, + "learning_rate": 4.750203086921203e-06, + "loss": 1.6412, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.5571675300598145, + "learning_rate": 4.74817221770918e-06, + "loss": 1.5779, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.87091326713562, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.6878, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.8140673637390137, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7023, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.6550583839416504, + "learning_rate": 4.742079610073111e-06, + "loss": 1.5995, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.6259925365448, + "learning_rate": 4.740048740861089e-06, + "loss": 1.6194, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.971848726272583, + "learning_rate": 4.738017871649066e-06, + "loss": 1.6634, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6344778537750244, + "learning_rate": 4.735987002437044e-06, + "loss": 1.6502, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 2.9362456798553467, + "learning_rate": 4.73395613322502e-06, + "loss": 1.6948, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.8098835945129395, + "learning_rate": 4.731925264012998e-06, + "loss": 1.7161, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.8724393844604492, + "eval_runtime": 37.0876, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.933009386062622, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7014, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.8216991424560547, + "learning_rate": 4.727863525588953e-06, + "loss": 1.6192, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.7028493881225586, + "learning_rate": 4.725832656376929e-06, + "loss": 1.607, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.816818952560425, + "learning_rate": 4.723801787164907e-06, + "loss": 1.6183, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.7707865238189697, + "learning_rate": 4.721770917952885e-06, + "loss": 1.6795, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.779562473297119, + "learning_rate": 4.719740048740861e-06, + "loss": 1.63, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.907338857650757, + "learning_rate": 4.717709179528839e-06, + "loss": 1.644, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 3.073558807373047, + "learning_rate": 4.715678310316816e-06, + "loss": 1.6821, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.7556493282318115, + "learning_rate": 4.713647441104794e-06, + "loss": 1.6501, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6718344688415527, + "learning_rate": 4.71161657189277e-06, + "loss": 1.6225, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.854687213897705, + "eval_runtime": 36.9773, + "eval_samples_per_second": 27.044, + "eval_steps_per_second": 1.136, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.763335943222046, + "learning_rate": 4.709585702680748e-06, + "loss": 1.6311, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.7669479846954346, + "learning_rate": 4.707554833468725e-06, + "loss": 1.6809, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.711562395095825, + "learning_rate": 4.705523964256703e-06, + "loss": 1.741, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.799954652786255, + "learning_rate": 4.703493095044679e-06, + "loss": 1.6409, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.7278690338134766, + "learning_rate": 4.701462225832657e-06, + "loss": 1.6325, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.6802427768707275, + "learning_rate": 4.699431356620634e-06, + "loss": 1.5988, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.8826160430908203, + "learning_rate": 4.697400487408612e-06, + "loss": 1.7119, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.784343719482422, + "learning_rate": 4.695369618196588e-06, + "loss": 1.6855, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.683337688446045, + "learning_rate": 4.693338748984566e-06, + "loss": 1.6588, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.6646857261657715, + "learning_rate": 4.691307879772543e-06, + "loss": 1.623, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.8511513471603394, + "eval_runtime": 36.9852, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.8181183338165283, + "learning_rate": 4.68927701056052e-06, + "loss": 1.6171, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.8270857334136963, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.678, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.7685446739196777, + "learning_rate": 4.685215272136475e-06, + "loss": 1.6575, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.7003183364868164, + "learning_rate": 4.683184402924452e-06, + "loss": 1.6405, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.668750286102295, + "learning_rate": 4.681153533712429e-06, + "loss": 1.6541, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 3.065800905227661, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7288, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.615601062774658, + "learning_rate": 4.677091795288384e-06, + "loss": 1.6306, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.5948891639709473, + "learning_rate": 4.675060926076361e-06, + "loss": 1.6018, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.7550079822540283, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7009, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.868687868118286, + "learning_rate": 4.670999187652316e-06, + "loss": 1.6034, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.8473671674728394, + "eval_runtime": 37.0015, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.705962657928467, + "learning_rate": 4.668968318440293e-06, + "loss": 1.6084, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.7901813983917236, + "learning_rate": 4.66693744922827e-06, + "loss": 1.6171, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.688774824142456, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.6774, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.734617233276367, + "learning_rate": 4.662875710804225e-06, + "loss": 1.5796, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.742017984390259, + "learning_rate": 4.660844841592202e-06, + "loss": 1.6671, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.6438794136047363, + "learning_rate": 4.658813972380179e-06, + "loss": 1.5835, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6441051959991455, + "learning_rate": 4.656783103168157e-06, + "loss": 1.6909, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.824204206466675, + "learning_rate": 4.654752233956134e-06, + "loss": 1.5758, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.556638479232788, + "learning_rate": 4.652721364744111e-06, + "loss": 1.6504, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.897812843322754, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.6406, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.8537043333053589, + "eval_runtime": 37.257, + "eval_samples_per_second": 26.841, + "eval_steps_per_second": 1.127, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.6594088077545166, + "learning_rate": 4.648659626320066e-06, + "loss": 1.6966, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.79524302482605, + "learning_rate": 4.646628757108043e-06, + "loss": 1.6645, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.80291748046875, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.6281, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.7723848819732666, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.5959, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 3.0625650882720947, + "learning_rate": 4.640536149471974e-06, + "loss": 1.6313, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.775726079940796, + "learning_rate": 4.638505280259952e-06, + "loss": 1.6563, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.837758779525757, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.6727, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 3.041355609893799, + "learning_rate": 4.634443541835907e-06, + "loss": 1.6282, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.5775389671325684, + "learning_rate": 4.632412672623883e-06, + "loss": 1.651, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 3.026287794113159, + "learning_rate": 4.630381803411861e-06, + "loss": 1.6458, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.8461565971374512, + "eval_runtime": 37.3881, + "eval_samples_per_second": 26.747, + "eval_steps_per_second": 1.123, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.6781764030456543, + "learning_rate": 4.628350934199838e-06, + "loss": 1.5232, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.763350486755371, + "learning_rate": 4.626320064987816e-06, + "loss": 1.6186, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.7063372135162354, + "learning_rate": 4.624289195775792e-06, + "loss": 1.6784, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.824563980102539, + "learning_rate": 4.62225832656377e-06, + "loss": 1.6219, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.5335569381713867, + "learning_rate": 4.620227457351747e-06, + "loss": 1.6152, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.773644208908081, + "learning_rate": 4.618196588139724e-06, + "loss": 1.6179, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.9841721057891846, + "learning_rate": 4.616165718927701e-06, + "loss": 1.6258, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.684171676635742, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.6066, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.8817965984344482, + "learning_rate": 4.612103980503656e-06, + "loss": 1.6697, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.9846866130828857, + "learning_rate": 4.610073111291633e-06, + "loss": 1.6541, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.8485002517700195, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.904170513153076, + "learning_rate": 4.60804224207961e-06, + "loss": 1.6362, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.684476375579834, + "learning_rate": 4.606011372867588e-06, + "loss": 1.6303, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6692216396331787, + "learning_rate": 4.603980503655565e-06, + "loss": 1.638, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.73555588722229, + "learning_rate": 4.601949634443542e-06, + "loss": 1.6125, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.729525089263916, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.565, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.5972018241882324, + "learning_rate": 4.597887896019497e-06, + "loss": 1.5856, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7859137058258057, + "learning_rate": 4.595857026807474e-06, + "loss": 1.6325, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.662587881088257, + "learning_rate": 4.593826157595451e-06, + "loss": 1.6172, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.6380577087402344, + "learning_rate": 4.591795288383429e-06, + "loss": 1.6005, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8375349044799805, + "learning_rate": 4.589764419171406e-06, + "loss": 1.6507, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.833074688911438, + "eval_runtime": 37.0695, + "eval_samples_per_second": 26.976, + "eval_steps_per_second": 1.133, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.786713123321533, + "learning_rate": 4.587733549959383e-06, + "loss": 1.6283, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.6959030628204346, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.5994, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.6100692749023438, + "learning_rate": 4.583671811535338e-06, + "loss": 1.5884, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.7281863689422607, + "learning_rate": 4.581640942323315e-06, + "loss": 1.5737, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.7519376277923584, + "learning_rate": 4.579610073111292e-06, + "loss": 1.6531, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.9314866065979004, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.6755, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.9250576496124268, + "learning_rate": 4.575548334687247e-06, + "loss": 1.6604, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.6382815837860107, + "learning_rate": 4.573517465475224e-06, + "loss": 1.6773, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.6891751289367676, + "learning_rate": 4.571486596263201e-06, + "loss": 1.5849, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.7018136978149414, + "learning_rate": 4.569455727051178e-06, + "loss": 1.6358, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.838019847869873, + "eval_runtime": 37.0496, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.671952247619629, + "learning_rate": 4.567424857839156e-06, + "loss": 1.6207, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.882539749145508, + "learning_rate": 4.565393988627132e-06, + "loss": 1.6618, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.684070587158203, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.6502, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.5295751094818115, + "learning_rate": 4.561332250203087e-06, + "loss": 1.6078, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.0268194675445557, + "learning_rate": 4.559301380991065e-06, + "loss": 1.6312, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6895201206207275, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.5848, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.869950771331787, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.5951, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.909505605697632, + "learning_rate": 4.553208773354996e-06, + "loss": 1.6589, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.7604334354400635, + "learning_rate": 4.551177904142974e-06, + "loss": 1.5767, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.8792080879211426, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.6864, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.836847186088562, + "eval_runtime": 36.9142, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.607037305831909, + "learning_rate": 4.547116165718929e-06, + "loss": 1.5925, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.9338927268981934, + "learning_rate": 4.545085296506905e-06, + "loss": 1.6433, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.7723233699798584, + "learning_rate": 4.543054427294882e-06, + "loss": 1.5963, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.4122188091278076, + "learning_rate": 4.54102355808286e-06, + "loss": 1.5978, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.7582709789276123, + "learning_rate": 4.538992688870837e-06, + "loss": 1.581, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.6983039379119873, + "learning_rate": 4.536961819658814e-06, + "loss": 1.6253, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.6036365032196045, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.5957, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.759645700454712, + "learning_rate": 4.532900081234769e-06, + "loss": 1.6199, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.797744035720825, + "learning_rate": 4.530869212022746e-06, + "loss": 1.5854, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.752868890762329, + "learning_rate": 4.528838342810723e-06, + "loss": 1.619, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.8305871486663818, + "eval_runtime": 37.0307, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.7421889305114746, + "learning_rate": 4.526807473598701e-06, + "loss": 1.6182, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.6852707862854004, + "learning_rate": 4.524776604386678e-06, + "loss": 1.6135, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.861720323562622, + "learning_rate": 4.522745735174655e-06, + "loss": 1.6099, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.706851005554199, + "learning_rate": 4.520714865962632e-06, + "loss": 1.6597, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7992191314697266, + "learning_rate": 4.51868399675061e-06, + "loss": 1.6077, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.6248183250427246, + "learning_rate": 4.516653127538587e-06, + "loss": 1.6485, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.696040391921997, + "learning_rate": 4.514622258326564e-06, + "loss": 1.5853, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.5958261489868164, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.571, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.9715874195098877, + "learning_rate": 4.510560519902519e-06, + "loss": 1.6044, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.664834499359131, + "learning_rate": 4.508529650690496e-06, + "loss": 1.6568, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.8277279138565063, + "eval_runtime": 37.0489, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.869558811187744, + "learning_rate": 4.506498781478473e-06, + "loss": 1.5654, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.5814828872680664, + "learning_rate": 4.504467912266451e-06, + "loss": 1.5484, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.6158947944641113, + "learning_rate": 4.502437043054428e-06, + "loss": 1.6202, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.7424654960632324, + "learning_rate": 4.500406173842405e-06, + "loss": 1.6755, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.8670949935913086, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.6144, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.7914652824401855, + "learning_rate": 4.49634443541836e-06, + "loss": 1.6164, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.866380214691162, + "learning_rate": 4.494313566206336e-06, + "loss": 1.5961, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.966474771499634, + "learning_rate": 4.492282696994314e-06, + "loss": 1.6681, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.751291513442993, + "learning_rate": 4.490251827782291e-06, + "loss": 1.6338, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.58840012550354, + "learning_rate": 4.488220958570269e-06, + "loss": 1.5619, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.824889898300171, + "eval_runtime": 37.0821, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.786945104598999, + "learning_rate": 4.486190089358245e-06, + "loss": 1.5712, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4536845684051514, + "learning_rate": 4.484159220146223e-06, + "loss": 1.5487, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.780278205871582, + "learning_rate": 4.4821283509342e-06, + "loss": 1.6269, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.5427017211914062, + "learning_rate": 4.480097481722178e-06, + "loss": 1.6254, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.7540087699890137, + "learning_rate": 4.478066612510154e-06, + "loss": 1.6112, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.616352081298828, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.5637, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.686035633087158, + "learning_rate": 4.474004874086109e-06, + "loss": 1.5974, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.982365846633911, + "learning_rate": 4.471974004874087e-06, + "loss": 1.7034, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.6106085777282715, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.5482, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.7798519134521484, + "learning_rate": 4.467912266450041e-06, + "loss": 1.6475, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.820489764213562, + "eval_runtime": 37.0672, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.8634185791015625, + "learning_rate": 4.465881397238018e-06, + "loss": 1.6258, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.633626699447632, + "learning_rate": 4.463850528025995e-06, + "loss": 1.5806, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.719074249267578, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.6404, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.7545487880706787, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6404, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.6559669971466064, + "learning_rate": 4.457757920389927e-06, + "loss": 1.6104, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.9004898071289062, + "learning_rate": 4.455727051177904e-06, + "loss": 1.5513, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.6640021800994873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.5859, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.842026472091675, + "learning_rate": 4.451665312753859e-06, + "loss": 1.6241, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.784647226333618, + "learning_rate": 4.449634443541836e-06, + "loss": 1.627, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.7067618370056152, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.6276, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.8115602731704712, + "eval_runtime": 37.078, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.6582882404327393, + "learning_rate": 4.445572705117791e-06, + "loss": 1.5746, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7881102561950684, + "learning_rate": 4.443541835905768e-06, + "loss": 1.5749, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.9323439598083496, + "learning_rate": 4.441510966693745e-06, + "loss": 1.6421, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.63948655128479, + "learning_rate": 4.439480097481723e-06, + "loss": 1.5853, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.7746710777282715, + "learning_rate": 4.4374492282697e-06, + "loss": 1.6455, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.093895435333252, + "learning_rate": 4.435418359057677e-06, + "loss": 1.5311, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.526385545730591, + "learning_rate": 4.433387489845654e-06, + "loss": 1.6082, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.6604297161102295, + "learning_rate": 4.431356620633632e-06, + "loss": 1.61, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.6636247634887695, + "learning_rate": 4.429325751421609e-06, + "loss": 1.5815, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.5567514896392822, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6086, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.8170957565307617, + "eval_runtime": 37.4095, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.695563793182373, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.5836, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6009771823883057, + "learning_rate": 4.423233143785541e-06, + "loss": 1.5264, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.0490736961364746, + "learning_rate": 4.421202274573518e-06, + "loss": 1.5982, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.8609886169433594, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6265, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4415369033813477, + "learning_rate": 4.417140536149473e-06, + "loss": 1.4987, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.771657943725586, + "learning_rate": 4.415109666937449e-06, + "loss": 1.5494, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.739617347717285, + "learning_rate": 4.413078797725427e-06, + "loss": 1.6274, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.805098295211792, + "learning_rate": 4.411047928513404e-06, + "loss": 1.5686, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.7248735427856445, + "learning_rate": 4.409017059301382e-06, + "loss": 1.6027, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6653318405151367, + "learning_rate": 4.406986190089358e-06, + "loss": 1.5909, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.8076283931732178, + "eval_runtime": 37.5073, + "eval_samples_per_second": 26.661, + "eval_steps_per_second": 1.12, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.626783847808838, + "learning_rate": 4.404955320877336e-06, + "loss": 1.6205, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.6929268836975098, + "learning_rate": 4.402924451665313e-06, + "loss": 1.6146, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.6806800365448, + "learning_rate": 4.400893582453291e-06, + "loss": 1.5702, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.5407259464263916, + "learning_rate": 4.398862713241267e-06, + "loss": 1.5525, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.639122247695923, + "learning_rate": 4.396831844029245e-06, + "loss": 1.5826, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.669396162033081, + "learning_rate": 4.394800974817222e-06, + "loss": 1.5982, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.6377906799316406, + "learning_rate": 4.392770105605199e-06, + "loss": 1.5797, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6769638061523438, + "learning_rate": 4.390739236393176e-06, + "loss": 1.607, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.6347222328186035, + "learning_rate": 4.388708367181154e-06, + "loss": 1.6196, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.670680284500122, + "learning_rate": 4.386677497969131e-06, + "loss": 1.619, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.8124384880065918, + "eval_runtime": 37.4751, + "eval_samples_per_second": 26.684, + "eval_steps_per_second": 1.121, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.5886521339416504, + "learning_rate": 4.384646628757108e-06, + "loss": 1.6258, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.644566297531128, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6447, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.6432394981384277, + "learning_rate": 4.380584890333063e-06, + "loss": 1.5888, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.5602784156799316, + "learning_rate": 4.37855402112104e-06, + "loss": 1.5544, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.4874439239501953, + "learning_rate": 4.376523151909017e-06, + "loss": 1.5075, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.694748640060425, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.6498, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.582211494445801, + "learning_rate": 4.372461413484972e-06, + "loss": 1.5877, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.675713539123535, + "learning_rate": 4.370430544272949e-06, + "loss": 1.6045, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.7157161235809326, + "learning_rate": 4.368399675060926e-06, + "loss": 1.623, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.795076370239258, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6141, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.8062744140625, + "eval_runtime": 37.2435, + "eval_samples_per_second": 26.85, + "eval_steps_per_second": 1.128, + "step": 600 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.870385645617152e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-600/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/generation_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/latest b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/latest new file mode 100644 index 0000000000000000000000000000000000000000..a0f3e526aa9af2ac647b278f006bb9616843c5d6 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/latest @@ -0,0 +1 @@ +global_step700 \ No newline at end of file diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/model.safetensors.index.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/special_tokens_map.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/tokenizer_config.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/trainer_state.json b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..8108ed9905b56a8059b07cf28a886bf8ed4424e9 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/trainer_state.json @@ -0,0 +1,5493 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.7671232876712328, + "eval_steps": 10, + "global_step": 700, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7904, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7703, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7401, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7709, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7344, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8913, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.789, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8088, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7921, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8587, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.8344645500183105, + "eval_runtime": 37.837, + "eval_samples_per_second": 26.429, + "eval_steps_per_second": 1.11, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8144, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7208, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8085, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7815, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 3.906928539276123, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.8328, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 3.8761353492736816, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.7683, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 3.9978103637695312, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.7677, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 3.8899712562561035, + "learning_rate": 7.299270072992701e-08, + "loss": 2.8026, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.052681922912598, + "learning_rate": 9.124087591240876e-08, + "loss": 2.7572, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 3.9182305335998535, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.851, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.833639621734619, + "eval_runtime": 37.65, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.02630090713501, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.8169, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 3.8186051845550537, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.7255, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 3.8384382724761963, + "learning_rate": 1.642335766423358e-07, + "loss": 2.7719, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 3.8526113033294678, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.8174, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 3.648515224456787, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.8106, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 3.716984748840332, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.801, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.7279837131500244, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.822, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.647592782974243, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.7703, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.4229180812835693, + "learning_rate": 2.737226277372263e-07, + "loss": 2.7515, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.4730520248413086, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.7106, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.7988297939300537, + "eval_runtime": 37.3293, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.0499517917633057, + "learning_rate": 3.102189781021898e-07, + "loss": 2.747, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 2.9903063774108887, + "learning_rate": 3.284671532846716e-07, + "loss": 2.7487, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 2.990891695022583, + "learning_rate": 3.467153284671533e-07, + "loss": 2.7175, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 2.9300644397735596, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.7057, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 2.770928382873535, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.7374, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8250246047973633, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.7655, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.4764456748962402, + "learning_rate": 4.197080291970803e-07, + "loss": 2.6921, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.4779810905456543, + "learning_rate": 4.379562043795621e-07, + "loss": 2.6515, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.517014980316162, + "learning_rate": 4.562043795620438e-07, + "loss": 2.7621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.4234426021575928, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.6568, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.733020544052124, + "eval_runtime": 37.2647, + "eval_samples_per_second": 26.835, + "eval_steps_per_second": 1.127, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.507478952407837, + "learning_rate": 4.927007299270073e-07, + "loss": 2.6906, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.2636210918426514, + "learning_rate": 5.109489051094891e-07, + "loss": 2.7169, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.2580671310424805, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7015, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.1623055934906006, + "learning_rate": 5.474452554744526e-07, + "loss": 2.7024, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2087788581848145, + "learning_rate": 5.656934306569343e-07, + "loss": 2.6969, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.1327500343322754, + "learning_rate": 5.839416058394161e-07, + "loss": 2.6324, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.0596821308135986, + "learning_rate": 6.021897810218979e-07, + "loss": 2.6123, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 1.952254295349121, + "learning_rate": 6.204379562043796e-07, + "loss": 2.648, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.8542354106903076, + "learning_rate": 6.386861313868614e-07, + "loss": 2.569, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 1.9736655950546265, + "learning_rate": 6.569343065693432e-07, + "loss": 2.715, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.6547672748565674, + "eval_runtime": 37.0691, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 1.9505298137664795, + "learning_rate": 6.751824817518248e-07, + "loss": 2.5774, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.8306165933609009, + "learning_rate": 6.934306569343066e-07, + "loss": 2.6153, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.7291252613067627, + "learning_rate": 7.116788321167883e-07, + "loss": 2.5454, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.5850962400436401, + "learning_rate": 7.299270072992701e-07, + "loss": 2.5566, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.592590570449829, + "learning_rate": 7.48175182481752e-07, + "loss": 2.581, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.5905886888504028, + "learning_rate": 7.664233576642337e-07, + "loss": 2.5866, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.6853892803192139, + "learning_rate": 7.846715328467154e-07, + "loss": 2.594, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.6294301748275757, + "learning_rate": 8.029197080291971e-07, + "loss": 2.5695, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.5434138774871826, + "learning_rate": 8.211678832116789e-07, + "loss": 2.5119, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.684287190437317, + "learning_rate": 8.394160583941606e-07, + "loss": 2.4893, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.596491813659668, + "eval_runtime": 37.0709, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.6595470905303955, + "learning_rate": 8.576642335766425e-07, + "loss": 2.5247, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6877329349517822, + "learning_rate": 8.759124087591242e-07, + "loss": 2.5046, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.5429190397262573, + "learning_rate": 8.94160583941606e-07, + "loss": 2.5182, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7417540550231934, + "learning_rate": 9.124087591240876e-07, + "loss": 2.5345, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.5623447895050049, + "learning_rate": 9.306569343065694e-07, + "loss": 2.4965, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5643991231918335, + "learning_rate": 9.489051094890511e-07, + "loss": 2.52, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.5427721738815308, + "learning_rate": 9.67153284671533e-07, + "loss": 2.5009, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.5804131031036377, + "learning_rate": 9.854014598540146e-07, + "loss": 2.5457, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.679776906967163, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.4679, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.729390025138855, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.4399, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.5290703773498535, + "eval_runtime": 37.0583, + "eval_samples_per_second": 26.984, + "eval_steps_per_second": 1.133, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.779736042022705, + "learning_rate": 1.04014598540146e-06, + "loss": 2.5181, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 1.8900535106658936, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.3951, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.861053228378296, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.3902, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.114215850830078, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.4072, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.2751121520996094, + "learning_rate": 1.113138686131387e-06, + "loss": 2.4419, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.2021124362945557, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.4544, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.837768077850342, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.359, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.189955472946167, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.358, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.4810078144073486, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.3493, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.2023067474365234, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.3517, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.4466662406921387, + "eval_runtime": 37.0566, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 1.133, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.083674430847168, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.3534, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4299442768096924, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.3865, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.254706859588623, + "learning_rate": 1.259124087591241e-06, + "loss": 2.3431, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.115211248397827, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.3561, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.3584327697753906, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.343, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.604055404663086, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.2822, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.262540578842163, + "learning_rate": 1.332116788321168e-06, + "loss": 2.3197, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4614903926849365, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.2344, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.2465147972106934, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.2755, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.6906347274780273, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.1843, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.3721561431884766, + "eval_runtime": 37.0663, + "eval_samples_per_second": 26.979, + "eval_steps_per_second": 1.133, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.481743335723877, + "learning_rate": 1.405109489051095e-06, + "loss": 2.3047, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.399268388748169, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.2581, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.306957960128784, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.2854, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.532820701599121, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.2529, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.8887627124786377, + "learning_rate": 1.478102189781022e-06, + "loss": 2.21, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.950792074203491, + "learning_rate": 1.496350364963504e-06, + "loss": 2.2764, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.380949020385742, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.2461, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 3.1742029190063477, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.2669, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.432572603225708, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2133, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.6373605728149414, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.1722, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.304755449295044, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.328529119491577, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.1542, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.6276895999908447, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.2331, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.719228506088257, + "learning_rate": 1.624087591240876e-06, + "loss": 2.2646, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.596773147583008, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.1856, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 3.006861686706543, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.1299, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.8645336627960205, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.1402, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.6831343173980713, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.1878, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 3.4171595573425293, + "learning_rate": 1.715328467153285e-06, + "loss": 2.0976, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.6250598430633545, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.1577, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 3.9928224086761475, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.1488, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.2569072246551514, + "eval_runtime": 36.9567, + "eval_samples_per_second": 27.059, + "eval_steps_per_second": 1.136, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.4283833503723145, + "learning_rate": 1.77007299270073e-06, + "loss": 2.1893, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 4.3146443367004395, + "learning_rate": 1.788321167883212e-06, + "loss": 2.1801, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.8649468421936035, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2152, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 3.4944796562194824, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.1519, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.8709475994110107, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.0923, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.6932499408721924, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.1153, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.2397680282592773, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.114, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5339198112487793, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.1179, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 3.527005672454834, + "learning_rate": 1.916058394160584e-06, + "loss": 2.1237, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.6510908603668213, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1208, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.218452215194702, + "eval_runtime": 37.0391, + "eval_samples_per_second": 26.998, + "eval_steps_per_second": 1.134, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 4.179861068725586, + "learning_rate": 1.952554744525548e-06, + "loss": 2.0254, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.8919570446014404, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.0874, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.425593137741089, + "learning_rate": 1.989051094890511e-06, + "loss": 2.0624, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 3.2468361854553223, + "learning_rate": 2.007299270072993e-06, + "loss": 2.0829, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.953590154647827, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1266, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.1816306114196777, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.0613, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.504004716873169, + "learning_rate": 2.062043795620438e-06, + "loss": 2.0817, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8403778076171875, + "learning_rate": 2.08029197080292e-06, + "loss": 2.0611, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 3.306710720062256, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1163, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.7831554412841797, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.0161, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.1820547580718994, + "eval_runtime": 37.3884, + "eval_samples_per_second": 26.746, + "eval_steps_per_second": 1.123, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.8647801876068115, + "learning_rate": 2.135036496350365e-06, + "loss": 2.0798, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.5721442699432373, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.0978, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1265504360198975, + "learning_rate": 2.171532846715329e-06, + "loss": 2.0134, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6115546226501465, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.0275, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.8961997032165527, + "learning_rate": 2.208029197080292e-06, + "loss": 2.0637, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.823732852935791, + "learning_rate": 2.226277372262774e-06, + "loss": 2.0348, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.9123544692993164, + "learning_rate": 2.244525547445256e-06, + "loss": 1.9968, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.9987287521362305, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.0687, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.746525764465332, + "learning_rate": 2.281021897810219e-06, + "loss": 2.0599, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.031419038772583, + "learning_rate": 2.299270072992701e-06, + "loss": 2.0189, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.1600100994110107, + "eval_runtime": 37.4178, + "eval_samples_per_second": 26.725, + "eval_steps_per_second": 1.122, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.091212272644043, + "learning_rate": 2.3175182481751825e-06, + "loss": 1.9926, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 3.141331672668457, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.0149, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.9116017818450928, + "learning_rate": 2.354014598540146e-06, + "loss": 1.9782, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.629312515258789, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.0087, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.1184496879577637, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0268, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.708953857421875, + "learning_rate": 2.4087591240875918e-06, + "loss": 1.9599, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.6366217136383057, + "learning_rate": 2.427007299270073e-06, + "loss": 1.9673, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.875124931335449, + "learning_rate": 2.445255474452555e-06, + "loss": 1.9743, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.911785364151001, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0438, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 3.3519198894500732, + "learning_rate": 2.4817518248175183e-06, + "loss": 1.9333, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.1266536712646484, + "eval_runtime": 37.3106, + "eval_samples_per_second": 26.802, + "eval_steps_per_second": 1.126, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.7893099784851074, + "learning_rate": 2.5e-06, + "loss": 1.91, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.813505172729492, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0315, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.356557607650757, + "learning_rate": 2.5364963503649635e-06, + "loss": 1.9322, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.960689067840576, + "learning_rate": 2.5547445255474458e-06, + "loss": 1.9249, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 3.455681800842285, + "learning_rate": 2.5729927007299272e-06, + "loss": 1.9884, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 4.040747165679932, + "learning_rate": 2.591240875912409e-06, + "loss": 1.9698, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 3.0668084621429443, + "learning_rate": 2.6094890510948905e-06, + "loss": 1.9599, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.0219950675964355, + "learning_rate": 2.627737226277373e-06, + "loss": 1.9767, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.1736350059509277, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0069, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 3.402104616165161, + "learning_rate": 2.664233576642336e-06, + "loss": 1.9578, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.101997137069702, + "eval_runtime": 37.1135, + "eval_samples_per_second": 26.944, + "eval_steps_per_second": 1.132, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.9179415702819824, + "learning_rate": 2.6824817518248175e-06, + "loss": 1.9162, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.1186177730560303, + "learning_rate": 2.7007299270072994e-06, + "loss": 1.9813, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.507108211517334, + "learning_rate": 2.718978102189781e-06, + "loss": 1.9344, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.4272000789642334, + "learning_rate": 2.737226277372263e-06, + "loss": 2.0066, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 3.2958714962005615, + "learning_rate": 2.755474452554745e-06, + "loss": 1.9738, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.321594476699829, + "learning_rate": 2.7737226277372264e-06, + "loss": 1.9192, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.727707624435425, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.0256, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.2205474376678467, + "learning_rate": 2.81021897810219e-06, + "loss": 1.9658, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.2150039672851562, + "learning_rate": 2.828467153284672e-06, + "loss": 1.9139, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.8227202892303467, + "learning_rate": 2.8467153284671534e-06, + "loss": 1.9585, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.0871636867523193, + "eval_runtime": 37.0122, + "eval_samples_per_second": 27.018, + "eval_steps_per_second": 1.135, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.9982995986938477, + "learning_rate": 2.8649635036496353e-06, + "loss": 1.8721, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.1639492511749268, + "learning_rate": 2.8832116788321167e-06, + "loss": 1.9075, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.095726490020752, + "learning_rate": 2.901459854014599e-06, + "loss": 1.8645, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 3.3911561965942383, + "learning_rate": 2.9197080291970804e-06, + "loss": 1.9751, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.625232696533203, + "learning_rate": 2.9379562043795623e-06, + "loss": 1.9396, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.8197591304779053, + "learning_rate": 2.956204379562044e-06, + "loss": 1.9394, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.754249334335327, + "learning_rate": 2.9744525547445256e-06, + "loss": 1.9734, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.9164586067199707, + "learning_rate": 2.992700729927008e-06, + "loss": 1.9298, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.039619207382202, + "learning_rate": 3.0109489051094893e-06, + "loss": 1.8975, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.8215222358703613, + "learning_rate": 3.029197080291971e-06, + "loss": 1.8612, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.0672969818115234, + "eval_runtime": 36.9948, + "eval_samples_per_second": 27.031, + "eval_steps_per_second": 1.135, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.7031683921813965, + "learning_rate": 3.0474452554744526e-06, + "loss": 1.9536, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.088528633117676, + "learning_rate": 3.065693430656935e-06, + "loss": 1.9231, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.798271894454956, + "learning_rate": 3.0839416058394163e-06, + "loss": 1.9184, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.8862130641937256, + "learning_rate": 3.102189781021898e-06, + "loss": 1.9463, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.24735951423645, + "learning_rate": 3.1204379562043796e-06, + "loss": 1.9247, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6648035049438477, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.8647, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.0373692512512207, + "learning_rate": 3.156934306569343e-06, + "loss": 1.8434, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.090059518814087, + "learning_rate": 3.175182481751825e-06, + "loss": 1.9731, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5814902782440186, + "learning_rate": 3.193430656934307e-06, + "loss": 1.9192, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 3.121877670288086, + "learning_rate": 3.2116788321167884e-06, + "loss": 1.8721, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.0536279678344727, + "eval_runtime": 37.0296, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 4.011061668395996, + "learning_rate": 3.2299270072992707e-06, + "loss": 1.9053, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 3.306645393371582, + "learning_rate": 3.248175182481752e-06, + "loss": 1.867, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 3.401339054107666, + "learning_rate": 3.266423357664234e-06, + "loss": 1.8213, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 3.637998580932617, + "learning_rate": 3.2846715328467155e-06, + "loss": 1.898, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 3.9283359050750732, + "learning_rate": 3.3029197080291973e-06, + "loss": 1.887, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.79770565032959, + "learning_rate": 3.3211678832116788e-06, + "loss": 1.9022, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.8775925636291504, + "learning_rate": 3.339416058394161e-06, + "loss": 1.8842, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 3.357034206390381, + "learning_rate": 3.3576642335766425e-06, + "loss": 1.8852, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7827579975128174, + "learning_rate": 3.3759124087591243e-06, + "loss": 1.8946, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 3.1603176593780518, + "learning_rate": 3.3941605839416058e-06, + "loss": 1.8716, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.029608964920044, + "eval_runtime": 37.0169, + "eval_samples_per_second": 27.015, + "eval_steps_per_second": 1.135, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.900463104248047, + "learning_rate": 3.412408759124088e-06, + "loss": 1.8697, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 3.0645594596862793, + "learning_rate": 3.43065693430657e-06, + "loss": 1.8887, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.9790565967559814, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.8073, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.4916350841522217, + "learning_rate": 3.467153284671533e-06, + "loss": 1.8798, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.9313464164733887, + "learning_rate": 3.4854014598540146e-06, + "loss": 1.8282, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.1672043800354004, + "learning_rate": 3.503649635036497e-06, + "loss": 1.8646, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.895242929458618, + "learning_rate": 3.5218978102189783e-06, + "loss": 1.8645, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 3.017155885696411, + "learning_rate": 3.54014598540146e-06, + "loss": 1.8739, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.085167169570923, + "learning_rate": 3.5583941605839416e-06, + "loss": 1.9117, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 3.167325735092163, + "learning_rate": 3.576642335766424e-06, + "loss": 1.9411, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.019193649291992, + "eval_runtime": 36.834, + "eval_samples_per_second": 27.149, + "eval_steps_per_second": 1.14, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 3.294370412826538, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9387, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.7739784717559814, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.7767, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0439939498901367, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.7783, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.9774787425994873, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.829, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6825435161590576, + "learning_rate": 3.667883211678833e-06, + "loss": 1.8322, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.43871808052063, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.7991, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 3.5290913581848145, + "learning_rate": 3.704379562043796e-06, + "loss": 1.8046, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9394545555114746, + "learning_rate": 3.7226277372262775e-06, + "loss": 1.9057, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 3.8342912197113037, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.8561, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 3.1885757446289062, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.8051, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.0043280124664307, + "eval_runtime": 36.9629, + "eval_samples_per_second": 27.054, + "eval_steps_per_second": 1.136, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 3.0562901496887207, + "learning_rate": 3.777372262773723e-06, + "loss": 1.8531, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.450380802154541, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.8337, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.8352060317993164, + "learning_rate": 3.8138686131386864e-06, + "loss": 1.8236, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 3.6039183139801025, + "learning_rate": 3.832116788321168e-06, + "loss": 1.8504, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 3.1691510677337646, + "learning_rate": 3.85036496350365e-06, + "loss": 1.7954, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.2054355144500732, + "learning_rate": 3.868613138686132e-06, + "loss": 1.8012, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.130464553833008, + "learning_rate": 3.886861313868613e-06, + "loss": 1.8142, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 3.454741954803467, + "learning_rate": 3.905109489051096e-06, + "loss": 1.8398, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.9741413593292236, + "learning_rate": 3.923357664233577e-06, + "loss": 1.868, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.5700478553771973, + "learning_rate": 3.9416058394160585e-06, + "loss": 1.7983, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 1.9883906841278076, + "eval_runtime": 37.1553, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 3.4521028995513916, + "learning_rate": 3.95985401459854e-06, + "loss": 1.8849, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 3.026761293411255, + "learning_rate": 3.978102189781022e-06, + "loss": 1.8921, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.2922768592834473, + "learning_rate": 3.996350364963504e-06, + "loss": 1.7558, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.8713529109954834, + "learning_rate": 4.014598540145986e-06, + "loss": 1.8317, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 3.0722813606262207, + "learning_rate": 4.032846715328467e-06, + "loss": 1.7507, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 3.09576153755188, + "learning_rate": 4.05109489051095e-06, + "loss": 1.8053, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.1753714084625244, + "learning_rate": 4.069343065693431e-06, + "loss": 1.8052, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.2278690338134766, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.7789, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.7040414810180664, + "learning_rate": 4.105839416058395e-06, + "loss": 1.7958, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.751190423965454, + "learning_rate": 4.124087591240876e-06, + "loss": 1.861, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 1.9721232652664185, + "eval_runtime": 37.3594, + "eval_samples_per_second": 26.767, + "eval_steps_per_second": 1.124, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.715794324874878, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.8018, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.787363052368164, + "learning_rate": 4.16058394160584e-06, + "loss": 1.7951, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.165919303894043, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.8646, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.409665107727051, + "learning_rate": 4.197080291970803e-06, + "loss": 1.7347, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 3.0749592781066895, + "learning_rate": 4.215328467153285e-06, + "loss": 1.8005, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.069650173187256, + "learning_rate": 4.233576642335767e-06, + "loss": 1.7816, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 3.019862174987793, + "learning_rate": 4.251824817518249e-06, + "loss": 1.7842, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 3.3378067016601562, + "learning_rate": 4.27007299270073e-06, + "loss": 1.7868, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 3.116816997528076, + "learning_rate": 4.288321167883212e-06, + "loss": 1.758, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 3.2071259021759033, + "learning_rate": 4.306569343065693e-06, + "loss": 1.7725, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 1.9668830633163452, + "eval_runtime": 37.4267, + "eval_samples_per_second": 26.719, + "eval_steps_per_second": 1.122, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.9114110469818115, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.7451, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.9991559982299805, + "learning_rate": 4.343065693430658e-06, + "loss": 1.7989, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 3.249702215194702, + "learning_rate": 4.361313868613139e-06, + "loss": 1.7893, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.1095340251922607, + "learning_rate": 4.379562043795621e-06, + "loss": 1.7724, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.332470655441284, + "learning_rate": 4.397810218978102e-06, + "loss": 1.825, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 3.201326847076416, + "learning_rate": 4.416058394160584e-06, + "loss": 1.7881, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 3.0723628997802734, + "learning_rate": 4.434306569343066e-06, + "loss": 1.7713, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2011122703552246, + "learning_rate": 4.452554744525548e-06, + "loss": 1.7585, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 3.465801239013672, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.8046, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 3.0417261123657227, + "learning_rate": 4.489051094890512e-06, + "loss": 1.8116, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 1.9586586952209473, + "eval_runtime": 37.2828, + "eval_samples_per_second": 26.822, + "eval_steps_per_second": 1.127, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.2577242851257324, + "learning_rate": 4.507299270072993e-06, + "loss": 1.8334, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.2681455612182617, + "learning_rate": 4.525547445255475e-06, + "loss": 1.8204, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.9197771549224854, + "learning_rate": 4.543795620437956e-06, + "loss": 1.775, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 3.401921272277832, + "learning_rate": 4.562043795620438e-06, + "loss": 1.738, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 3.1170408725738525, + "learning_rate": 4.580291970802921e-06, + "loss": 1.8292, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 3.016622543334961, + "learning_rate": 4.598540145985402e-06, + "loss": 1.839, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.2432861328125, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.7959, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.786032199859619, + "learning_rate": 4.635036496350365e-06, + "loss": 1.7646, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.188246011734009, + "learning_rate": 4.653284671532847e-06, + "loss": 1.7713, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.80027437210083, + "learning_rate": 4.671532846715329e-06, + "loss": 1.8031, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 1.9488393068313599, + "eval_runtime": 37.1054, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.0002360343933105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.7914, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.1637566089630127, + "learning_rate": 4.708029197080292e-06, + "loss": 1.7691, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 3.0462076663970947, + "learning_rate": 4.726277372262774e-06, + "loss": 1.8392, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 3.0676465034484863, + "learning_rate": 4.744525547445255e-06, + "loss": 1.803, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.927471399307251, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.736, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 3.0370371341705322, + "learning_rate": 4.78102189781022e-06, + "loss": 1.7877, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.5490198135375977, + "learning_rate": 4.799270072992701e-06, + "loss": 1.7992, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.85689377784729, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.7398, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.1516075134277344, + "learning_rate": 4.835766423357665e-06, + "loss": 1.7822, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.8532397747039795, + "learning_rate": 4.854014598540146e-06, + "loss": 1.784, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 1.9438674449920654, + "eval_runtime": 37.0689, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 3.1778180599212646, + "learning_rate": 4.872262773722628e-06, + "loss": 1.7127, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.0473344326019287, + "learning_rate": 4.89051094890511e-06, + "loss": 1.7461, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.7586324214935303, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.7255, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.051339626312256, + "learning_rate": 4.927007299270074e-06, + "loss": 1.7664, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 3.054767370223999, + "learning_rate": 4.945255474452555e-06, + "loss": 1.7803, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.9614083766937256, + "learning_rate": 4.963503649635037e-06, + "loss": 1.7412, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 3.252504348754883, + "learning_rate": 4.981751824817518e-06, + "loss": 1.8355, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.999206066131592, + "learning_rate": 5e-06, + "loss": 1.7883, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 3.381014347076416, + "learning_rate": 4.997969130787978e-06, + "loss": 1.7654, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 3.015758514404297, + "learning_rate": 4.995938261575955e-06, + "loss": 1.7821, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 1.9337852001190186, + "eval_runtime": 37.086, + "eval_samples_per_second": 26.964, + "eval_steps_per_second": 1.133, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 3.174635410308838, + "learning_rate": 4.993907392363932e-06, + "loss": 1.6993, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.990565776824951, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.7788, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 3.3665783405303955, + "learning_rate": 4.989845653939887e-06, + "loss": 1.7209, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.5099875926971436, + "learning_rate": 4.987814784727864e-06, + "loss": 1.7286, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.2758665084838867, + "learning_rate": 4.985783915515841e-06, + "loss": 1.7532, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.6986751556396484, + "learning_rate": 4.983753046303819e-06, + "loss": 1.7313, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 3.0444724559783936, + "learning_rate": 4.981722177091796e-06, + "loss": 1.7427, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 3.363081693649292, + "learning_rate": 4.979691307879773e-06, + "loss": 1.7149, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 3.037461519241333, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.7486, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.7800803184509277, + "learning_rate": 4.975629569455728e-06, + "loss": 1.7527, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 1.9319745302200317, + "eval_runtime": 37.0845, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 3.201310634613037, + "learning_rate": 4.973598700243705e-06, + "loss": 1.778, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.966487169265747, + "learning_rate": 4.971567831031681e-06, + "loss": 1.7111, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.0362396240234375, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7324, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.7503223419189453, + "learning_rate": 4.967506092607636e-06, + "loss": 1.6178, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.9468796253204346, + "learning_rate": 4.965475223395614e-06, + "loss": 1.6879, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 3.3060693740844727, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.7197, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.619682550430298, + "learning_rate": 4.961413484971569e-06, + "loss": 1.6925, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.908228635787964, + "learning_rate": 4.959382615759545e-06, + "loss": 1.7539, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 3.0089046955108643, + "learning_rate": 4.957351746547523e-06, + "loss": 1.7144, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.9313178062438965, + "learning_rate": 4.9553208773355e-06, + "loss": 1.7092, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 1.928604006767273, + "eval_runtime": 36.9545, + "eval_samples_per_second": 27.06, + "eval_steps_per_second": 1.137, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.070209264755249, + "learning_rate": 4.953290008123478e-06, + "loss": 1.8563, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.901120901107788, + "learning_rate": 4.951259138911454e-06, + "loss": 1.7364, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.7859580516815186, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.6974, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 3.193673610687256, + "learning_rate": 4.947197400487409e-06, + "loss": 1.7964, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.866114854812622, + "learning_rate": 4.945166531275386e-06, + "loss": 1.7347, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 3.227334976196289, + "learning_rate": 4.943135662063363e-06, + "loss": 1.6867, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.911961317062378, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.7282, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.375117778778076, + "learning_rate": 4.939073923639318e-06, + "loss": 1.7792, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 3.039350986480713, + "learning_rate": 4.937043054427295e-06, + "loss": 1.7625, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 3.1554577350616455, + "learning_rate": 4.935012185215272e-06, + "loss": 1.6978, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 1.9135974645614624, + "eval_runtime": 37.0781, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.874204635620117, + "learning_rate": 4.93298131600325e-06, + "loss": 1.6907, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 3.1355791091918945, + "learning_rate": 4.930950446791227e-06, + "loss": 1.7464, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.776468515396118, + "learning_rate": 4.928919577579204e-06, + "loss": 1.7373, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 3.0690815448760986, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.7471, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.8855857849121094, + "learning_rate": 4.924857839155159e-06, + "loss": 1.7143, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.9788177013397217, + "learning_rate": 4.922826969943136e-06, + "loss": 1.6989, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.8904759883880615, + "learning_rate": 4.920796100731113e-06, + "loss": 1.7292, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 3.154738187789917, + "learning_rate": 4.918765231519091e-06, + "loss": 1.6609, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 3.0258750915527344, + "learning_rate": 4.916734362307068e-06, + "loss": 1.7221, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 3.0016801357269287, + "learning_rate": 4.914703493095045e-06, + "loss": 1.672, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 1.9114375114440918, + "eval_runtime": 37.1092, + "eval_samples_per_second": 26.948, + "eval_steps_per_second": 1.132, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 3.2514264583587646, + "learning_rate": 4.912672623883022e-06, + "loss": 1.6893, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.674236536026001, + "learning_rate": 4.910641754671e-06, + "loss": 1.675, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 3.3388164043426514, + "learning_rate": 4.908610885458977e-06, + "loss": 1.708, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.903599262237549, + "learning_rate": 4.906580016246954e-06, + "loss": 1.7039, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1790928840637207, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.7484, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.906397581100464, + "learning_rate": 4.902518277822909e-06, + "loss": 1.7325, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.698976993560791, + "learning_rate": 4.900487408610886e-06, + "loss": 1.7158, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 2.8924341201782227, + "learning_rate": 4.898456539398863e-06, + "loss": 1.7285, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.028808355331421, + "learning_rate": 4.89642567018684e-06, + "loss": 1.7382, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.8244469165802, + "learning_rate": 4.894394800974818e-06, + "loss": 1.794, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 1.8975961208343506, + "eval_runtime": 37.124, + "eval_samples_per_second": 26.937, + "eval_steps_per_second": 1.131, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.959707260131836, + "learning_rate": 4.892363931762794e-06, + "loss": 1.6751, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.943209409713745, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.7485, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.833904981613159, + "learning_rate": 4.888302193338749e-06, + "loss": 1.674, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.0296709537506104, + "learning_rate": 4.886271324126727e-06, + "loss": 1.6641, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.88393235206604, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.7682, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.7120718955993652, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.7405, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.8697848320007324, + "learning_rate": 4.880178716490658e-06, + "loss": 1.7007, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.895385980606079, + "learning_rate": 4.878147847278636e-06, + "loss": 1.6712, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.704183578491211, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.675, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.811582088470459, + "learning_rate": 4.874086108854591e-06, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 1.8998559713363647, + "eval_runtime": 37.4341, + "eval_samples_per_second": 26.714, + "eval_steps_per_second": 1.122, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.605085849761963, + "learning_rate": 4.872055239642567e-06, + "loss": 1.6651, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.637470006942749, + "learning_rate": 4.870024370430544e-06, + "loss": 1.6316, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8741681575775146, + "learning_rate": 4.867993501218522e-06, + "loss": 1.6983, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.639359474182129, + "learning_rate": 4.865962632006499e-06, + "loss": 1.7107, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.962362766265869, + "learning_rate": 4.863931762794476e-06, + "loss": 1.6438, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 3.2190587520599365, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.7117, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.857741355895996, + "learning_rate": 4.859870024370431e-06, + "loss": 1.7359, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.867485761642456, + "learning_rate": 4.857839155158408e-06, + "loss": 1.7365, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.706620454788208, + "learning_rate": 4.855808285946385e-06, + "loss": 1.7538, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.7733891010284424, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.693, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 1.8900014162063599, + "eval_runtime": 37.5422, + "eval_samples_per_second": 26.637, + "eval_steps_per_second": 1.119, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.787938356399536, + "learning_rate": 4.85174654752234e-06, + "loss": 1.6923, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6490252017974854, + "learning_rate": 4.849715678310317e-06, + "loss": 1.6984, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.7721261978149414, + "learning_rate": 4.847684809098294e-06, + "loss": 1.7087, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.6365602016448975, + "learning_rate": 4.845653939886272e-06, + "loss": 1.6771, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6294384002685547, + "learning_rate": 4.843623070674249e-06, + "loss": 1.7036, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.542022228240967, + "learning_rate": 4.841592201462226e-06, + "loss": 1.6908, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.605208396911621, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.7268, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.5954573154449463, + "learning_rate": 4.837530463038181e-06, + "loss": 1.6543, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.8794212341308594, + "learning_rate": 4.835499593826158e-06, + "loss": 1.6596, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.641418933868408, + "learning_rate": 4.833468724614135e-06, + "loss": 1.717, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 1.883540153503418, + "eval_runtime": 37.4659, + "eval_samples_per_second": 26.691, + "eval_steps_per_second": 1.121, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.6206278800964355, + "learning_rate": 4.831437855402113e-06, + "loss": 1.6887, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.8248493671417236, + "learning_rate": 4.82940698619009e-06, + "loss": 1.5921, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.5635788440704346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.6714, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.8677446842193604, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.7045, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.895571231842041, + "learning_rate": 4.823314378554022e-06, + "loss": 1.7113, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.887521743774414, + "learning_rate": 4.821283509341998e-06, + "loss": 1.7053, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.807668447494507, + "learning_rate": 4.819252640129976e-06, + "loss": 1.6764, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.8177578449249268, + "learning_rate": 4.817221770917953e-06, + "loss": 1.6827, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.8412656784057617, + "learning_rate": 4.815190901705931e-06, + "loss": 1.6466, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.763336420059204, + "learning_rate": 4.813160032493907e-06, + "loss": 1.6826, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.8777310848236084, + "eval_runtime": 37.2238, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.7151660919189453, + "learning_rate": 4.811129163281885e-06, + "loss": 1.6469, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 3.066633939743042, + "learning_rate": 4.809098294069862e-06, + "loss": 1.7217, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.7772305011749268, + "learning_rate": 4.80706742485784e-06, + "loss": 1.6729, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.938558578491211, + "learning_rate": 4.805036555645816e-06, + "loss": 1.7224, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6988492012023926, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.6555, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.9132323265075684, + "learning_rate": 4.800974817221771e-06, + "loss": 1.7356, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.679778814315796, + "learning_rate": 4.798943948009749e-06, + "loss": 1.6628, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.9366323947906494, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.6952, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.9841928482055664, + "learning_rate": 4.794882209585703e-06, + "loss": 1.6883, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 3.130336284637451, + "learning_rate": 4.79285134037368e-06, + "loss": 1.7179, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.875152587890625, + "eval_runtime": 37.2033, + "eval_samples_per_second": 26.879, + "eval_steps_per_second": 1.129, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.8412537574768066, + "learning_rate": 4.790820471161657e-06, + "loss": 1.6651, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.871779680252075, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.659, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 3.1089956760406494, + "learning_rate": 4.786758732737612e-06, + "loss": 1.6545, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7381818294525146, + "learning_rate": 4.784727863525589e-06, + "loss": 1.6981, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 3.0174624919891357, + "learning_rate": 4.782696994313566e-06, + "loss": 1.6195, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 3.0664613246917725, + "learning_rate": 4.780666125101544e-06, + "loss": 1.6767, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 3.014626979827881, + "learning_rate": 4.778635255889521e-06, + "loss": 1.6128, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.7011501789093018, + "learning_rate": 4.776604386677498e-06, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.7285525798797607, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.6722, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.8392159938812256, + "learning_rate": 4.772542648253453e-06, + "loss": 1.7141, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.872631311416626, + "eval_runtime": 37.1715, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6624906063079834, + "learning_rate": 4.77051177904143e-06, + "loss": 1.704, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 3.0075268745422363, + "learning_rate": 4.768480909829407e-06, + "loss": 1.6322, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.6070828437805176, + "learning_rate": 4.766450040617385e-06, + "loss": 1.6871, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 3.044222831726074, + "learning_rate": 4.764419171405362e-06, + "loss": 1.6652, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.773167610168457, + "learning_rate": 4.762388302193339e-06, + "loss": 1.7063, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.880176305770874, + "learning_rate": 4.760357432981316e-06, + "loss": 1.6775, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.6402764320373535, + "learning_rate": 4.758326563769294e-06, + "loss": 1.6694, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.9245846271514893, + "learning_rate": 4.756295694557271e-06, + "loss": 1.6688, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.623044967651367, + "learning_rate": 4.754264825345248e-06, + "loss": 1.6819, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.739896059036255, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.6401, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.8707164525985718, + "eval_runtime": 36.9513, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.603599786758423, + "learning_rate": 4.750203086921203e-06, + "loss": 1.6412, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.5571675300598145, + "learning_rate": 4.74817221770918e-06, + "loss": 1.5779, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.87091326713562, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.6878, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.8140673637390137, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7023, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.6550583839416504, + "learning_rate": 4.742079610073111e-06, + "loss": 1.5995, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.6259925365448, + "learning_rate": 4.740048740861089e-06, + "loss": 1.6194, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.971848726272583, + "learning_rate": 4.738017871649066e-06, + "loss": 1.6634, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6344778537750244, + "learning_rate": 4.735987002437044e-06, + "loss": 1.6502, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 2.9362456798553467, + "learning_rate": 4.73395613322502e-06, + "loss": 1.6948, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.8098835945129395, + "learning_rate": 4.731925264012998e-06, + "loss": 1.7161, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.8724393844604492, + "eval_runtime": 37.0876, + "eval_samples_per_second": 26.963, + "eval_steps_per_second": 1.132, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.933009386062622, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7014, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.8216991424560547, + "learning_rate": 4.727863525588953e-06, + "loss": 1.6192, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.7028493881225586, + "learning_rate": 4.725832656376929e-06, + "loss": 1.607, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.816818952560425, + "learning_rate": 4.723801787164907e-06, + "loss": 1.6183, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.7707865238189697, + "learning_rate": 4.721770917952885e-06, + "loss": 1.6795, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.779562473297119, + "learning_rate": 4.719740048740861e-06, + "loss": 1.63, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.907338857650757, + "learning_rate": 4.717709179528839e-06, + "loss": 1.644, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 3.073558807373047, + "learning_rate": 4.715678310316816e-06, + "loss": 1.6821, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.7556493282318115, + "learning_rate": 4.713647441104794e-06, + "loss": 1.6501, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6718344688415527, + "learning_rate": 4.71161657189277e-06, + "loss": 1.6225, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.854687213897705, + "eval_runtime": 36.9773, + "eval_samples_per_second": 27.044, + "eval_steps_per_second": 1.136, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.763335943222046, + "learning_rate": 4.709585702680748e-06, + "loss": 1.6311, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.7669479846954346, + "learning_rate": 4.707554833468725e-06, + "loss": 1.6809, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.711562395095825, + "learning_rate": 4.705523964256703e-06, + "loss": 1.741, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.799954652786255, + "learning_rate": 4.703493095044679e-06, + "loss": 1.6409, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.7278690338134766, + "learning_rate": 4.701462225832657e-06, + "loss": 1.6325, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.6802427768707275, + "learning_rate": 4.699431356620634e-06, + "loss": 1.5988, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.8826160430908203, + "learning_rate": 4.697400487408612e-06, + "loss": 1.7119, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.784343719482422, + "learning_rate": 4.695369618196588e-06, + "loss": 1.6855, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.683337688446045, + "learning_rate": 4.693338748984566e-06, + "loss": 1.6588, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.6646857261657715, + "learning_rate": 4.691307879772543e-06, + "loss": 1.623, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.8511513471603394, + "eval_runtime": 36.9852, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.8181183338165283, + "learning_rate": 4.68927701056052e-06, + "loss": 1.6171, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.8270857334136963, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.678, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.7685446739196777, + "learning_rate": 4.685215272136475e-06, + "loss": 1.6575, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.7003183364868164, + "learning_rate": 4.683184402924452e-06, + "loss": 1.6405, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.668750286102295, + "learning_rate": 4.681153533712429e-06, + "loss": 1.6541, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 3.065800905227661, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7288, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.615601062774658, + "learning_rate": 4.677091795288384e-06, + "loss": 1.6306, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.5948891639709473, + "learning_rate": 4.675060926076361e-06, + "loss": 1.6018, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.7550079822540283, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7009, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.868687868118286, + "learning_rate": 4.670999187652316e-06, + "loss": 1.6034, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.8473671674728394, + "eval_runtime": 37.0015, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.705962657928467, + "learning_rate": 4.668968318440293e-06, + "loss": 1.6084, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.7901813983917236, + "learning_rate": 4.66693744922827e-06, + "loss": 1.6171, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.688774824142456, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.6774, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.734617233276367, + "learning_rate": 4.662875710804225e-06, + "loss": 1.5796, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.742017984390259, + "learning_rate": 4.660844841592202e-06, + "loss": 1.6671, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.6438794136047363, + "learning_rate": 4.658813972380179e-06, + "loss": 1.5835, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6441051959991455, + "learning_rate": 4.656783103168157e-06, + "loss": 1.6909, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.824204206466675, + "learning_rate": 4.654752233956134e-06, + "loss": 1.5758, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.556638479232788, + "learning_rate": 4.652721364744111e-06, + "loss": 1.6504, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.897812843322754, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.6406, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.8537043333053589, + "eval_runtime": 37.257, + "eval_samples_per_second": 26.841, + "eval_steps_per_second": 1.127, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.6594088077545166, + "learning_rate": 4.648659626320066e-06, + "loss": 1.6966, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.79524302482605, + "learning_rate": 4.646628757108043e-06, + "loss": 1.6645, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.80291748046875, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.6281, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.7723848819732666, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.5959, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 3.0625650882720947, + "learning_rate": 4.640536149471974e-06, + "loss": 1.6313, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.775726079940796, + "learning_rate": 4.638505280259952e-06, + "loss": 1.6563, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.837758779525757, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.6727, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 3.041355609893799, + "learning_rate": 4.634443541835907e-06, + "loss": 1.6282, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.5775389671325684, + "learning_rate": 4.632412672623883e-06, + "loss": 1.651, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 3.026287794113159, + "learning_rate": 4.630381803411861e-06, + "loss": 1.6458, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.8461565971374512, + "eval_runtime": 37.3881, + "eval_samples_per_second": 26.747, + "eval_steps_per_second": 1.123, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.6781764030456543, + "learning_rate": 4.628350934199838e-06, + "loss": 1.5232, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.763350486755371, + "learning_rate": 4.626320064987816e-06, + "loss": 1.6186, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.7063372135162354, + "learning_rate": 4.624289195775792e-06, + "loss": 1.6784, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.824563980102539, + "learning_rate": 4.62225832656377e-06, + "loss": 1.6219, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.5335569381713867, + "learning_rate": 4.620227457351747e-06, + "loss": 1.6152, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.773644208908081, + "learning_rate": 4.618196588139724e-06, + "loss": 1.6179, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.9841721057891846, + "learning_rate": 4.616165718927701e-06, + "loss": 1.6258, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.684171676635742, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.6066, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.8817965984344482, + "learning_rate": 4.612103980503656e-06, + "loss": 1.6697, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.9846866130828857, + "learning_rate": 4.610073111291633e-06, + "loss": 1.6541, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.8485002517700195, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.779, + "eval_steps_per_second": 1.125, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.904170513153076, + "learning_rate": 4.60804224207961e-06, + "loss": 1.6362, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.684476375579834, + "learning_rate": 4.606011372867588e-06, + "loss": 1.6303, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6692216396331787, + "learning_rate": 4.603980503655565e-06, + "loss": 1.638, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.73555588722229, + "learning_rate": 4.601949634443542e-06, + "loss": 1.6125, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.729525089263916, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.565, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.5972018241882324, + "learning_rate": 4.597887896019497e-06, + "loss": 1.5856, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7859137058258057, + "learning_rate": 4.595857026807474e-06, + "loss": 1.6325, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.662587881088257, + "learning_rate": 4.593826157595451e-06, + "loss": 1.6172, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.6380577087402344, + "learning_rate": 4.591795288383429e-06, + "loss": 1.6005, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8375349044799805, + "learning_rate": 4.589764419171406e-06, + "loss": 1.6507, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.833074688911438, + "eval_runtime": 37.0695, + "eval_samples_per_second": 26.976, + "eval_steps_per_second": 1.133, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.786713123321533, + "learning_rate": 4.587733549959383e-06, + "loss": 1.6283, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.6959030628204346, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.5994, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.6100692749023438, + "learning_rate": 4.583671811535338e-06, + "loss": 1.5884, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.7281863689422607, + "learning_rate": 4.581640942323315e-06, + "loss": 1.5737, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.7519376277923584, + "learning_rate": 4.579610073111292e-06, + "loss": 1.6531, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.9314866065979004, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.6755, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.9250576496124268, + "learning_rate": 4.575548334687247e-06, + "loss": 1.6604, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.6382815837860107, + "learning_rate": 4.573517465475224e-06, + "loss": 1.6773, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.6891751289367676, + "learning_rate": 4.571486596263201e-06, + "loss": 1.5849, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.7018136978149414, + "learning_rate": 4.569455727051178e-06, + "loss": 1.6358, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.838019847869873, + "eval_runtime": 37.0496, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.671952247619629, + "learning_rate": 4.567424857839156e-06, + "loss": 1.6207, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.882539749145508, + "learning_rate": 4.565393988627132e-06, + "loss": 1.6618, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.684070587158203, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.6502, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.5295751094818115, + "learning_rate": 4.561332250203087e-06, + "loss": 1.6078, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.0268194675445557, + "learning_rate": 4.559301380991065e-06, + "loss": 1.6312, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6895201206207275, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.5848, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.869950771331787, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.5951, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.909505605697632, + "learning_rate": 4.553208773354996e-06, + "loss": 1.6589, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.7604334354400635, + "learning_rate": 4.551177904142974e-06, + "loss": 1.5767, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.8792080879211426, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.6864, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.836847186088562, + "eval_runtime": 36.9142, + "eval_samples_per_second": 27.09, + "eval_steps_per_second": 1.138, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.607037305831909, + "learning_rate": 4.547116165718929e-06, + "loss": 1.5925, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.9338927268981934, + "learning_rate": 4.545085296506905e-06, + "loss": 1.6433, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.7723233699798584, + "learning_rate": 4.543054427294882e-06, + "loss": 1.5963, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.4122188091278076, + "learning_rate": 4.54102355808286e-06, + "loss": 1.5978, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.7582709789276123, + "learning_rate": 4.538992688870837e-06, + "loss": 1.581, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.6983039379119873, + "learning_rate": 4.536961819658814e-06, + "loss": 1.6253, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.6036365032196045, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.5957, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.759645700454712, + "learning_rate": 4.532900081234769e-06, + "loss": 1.6199, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.797744035720825, + "learning_rate": 4.530869212022746e-06, + "loss": 1.5854, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.752868890762329, + "learning_rate": 4.528838342810723e-06, + "loss": 1.619, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.8305871486663818, + "eval_runtime": 37.0307, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.7421889305114746, + "learning_rate": 4.526807473598701e-06, + "loss": 1.6182, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.6852707862854004, + "learning_rate": 4.524776604386678e-06, + "loss": 1.6135, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.861720323562622, + "learning_rate": 4.522745735174655e-06, + "loss": 1.6099, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.706851005554199, + "learning_rate": 4.520714865962632e-06, + "loss": 1.6597, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7992191314697266, + "learning_rate": 4.51868399675061e-06, + "loss": 1.6077, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.6248183250427246, + "learning_rate": 4.516653127538587e-06, + "loss": 1.6485, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.696040391921997, + "learning_rate": 4.514622258326564e-06, + "loss": 1.5853, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.5958261489868164, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.571, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.9715874195098877, + "learning_rate": 4.510560519902519e-06, + "loss": 1.6044, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.664834499359131, + "learning_rate": 4.508529650690496e-06, + "loss": 1.6568, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.8277279138565063, + "eval_runtime": 37.0489, + "eval_samples_per_second": 26.991, + "eval_steps_per_second": 1.134, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.869558811187744, + "learning_rate": 4.506498781478473e-06, + "loss": 1.5654, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.5814828872680664, + "learning_rate": 4.504467912266451e-06, + "loss": 1.5484, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.6158947944641113, + "learning_rate": 4.502437043054428e-06, + "loss": 1.6202, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.7424654960632324, + "learning_rate": 4.500406173842405e-06, + "loss": 1.6755, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.8670949935913086, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.6144, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.7914652824401855, + "learning_rate": 4.49634443541836e-06, + "loss": 1.6164, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.866380214691162, + "learning_rate": 4.494313566206336e-06, + "loss": 1.5961, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.966474771499634, + "learning_rate": 4.492282696994314e-06, + "loss": 1.6681, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.751291513442993, + "learning_rate": 4.490251827782291e-06, + "loss": 1.6338, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.58840012550354, + "learning_rate": 4.488220958570269e-06, + "loss": 1.5619, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.824889898300171, + "eval_runtime": 37.0821, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.786945104598999, + "learning_rate": 4.486190089358245e-06, + "loss": 1.5712, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4536845684051514, + "learning_rate": 4.484159220146223e-06, + "loss": 1.5487, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.780278205871582, + "learning_rate": 4.4821283509342e-06, + "loss": 1.6269, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.5427017211914062, + "learning_rate": 4.480097481722178e-06, + "loss": 1.6254, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.7540087699890137, + "learning_rate": 4.478066612510154e-06, + "loss": 1.6112, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.616352081298828, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.5637, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.686035633087158, + "learning_rate": 4.474004874086109e-06, + "loss": 1.5974, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.982365846633911, + "learning_rate": 4.471974004874087e-06, + "loss": 1.7034, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.6106085777282715, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.5482, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.7798519134521484, + "learning_rate": 4.467912266450041e-06, + "loss": 1.6475, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.820489764213562, + "eval_runtime": 37.0672, + "eval_samples_per_second": 26.978, + "eval_steps_per_second": 1.133, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.8634185791015625, + "learning_rate": 4.465881397238018e-06, + "loss": 1.6258, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.633626699447632, + "learning_rate": 4.463850528025995e-06, + "loss": 1.5806, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.719074249267578, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.6404, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.7545487880706787, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6404, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.6559669971466064, + "learning_rate": 4.457757920389927e-06, + "loss": 1.6104, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.9004898071289062, + "learning_rate": 4.455727051177904e-06, + "loss": 1.5513, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.6640021800994873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.5859, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.842026472091675, + "learning_rate": 4.451665312753859e-06, + "loss": 1.6241, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.784647226333618, + "learning_rate": 4.449634443541836e-06, + "loss": 1.627, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.7067618370056152, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.6276, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.8115602731704712, + "eval_runtime": 37.078, + "eval_samples_per_second": 26.97, + "eval_steps_per_second": 1.133, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.6582882404327393, + "learning_rate": 4.445572705117791e-06, + "loss": 1.5746, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7881102561950684, + "learning_rate": 4.443541835905768e-06, + "loss": 1.5749, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.9323439598083496, + "learning_rate": 4.441510966693745e-06, + "loss": 1.6421, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.63948655128479, + "learning_rate": 4.439480097481723e-06, + "loss": 1.5853, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.7746710777282715, + "learning_rate": 4.4374492282697e-06, + "loss": 1.6455, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.093895435333252, + "learning_rate": 4.435418359057677e-06, + "loss": 1.5311, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.526385545730591, + "learning_rate": 4.433387489845654e-06, + "loss": 1.6082, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.6604297161102295, + "learning_rate": 4.431356620633632e-06, + "loss": 1.61, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.6636247634887695, + "learning_rate": 4.429325751421609e-06, + "loss": 1.5815, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.5567514896392822, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6086, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.8170957565307617, + "eval_runtime": 37.4095, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.695563793182373, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.5836, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6009771823883057, + "learning_rate": 4.423233143785541e-06, + "loss": 1.5264, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.0490736961364746, + "learning_rate": 4.421202274573518e-06, + "loss": 1.5982, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.8609886169433594, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6265, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4415369033813477, + "learning_rate": 4.417140536149473e-06, + "loss": 1.4987, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.771657943725586, + "learning_rate": 4.415109666937449e-06, + "loss": 1.5494, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.739617347717285, + "learning_rate": 4.413078797725427e-06, + "loss": 1.6274, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.805098295211792, + "learning_rate": 4.411047928513404e-06, + "loss": 1.5686, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.7248735427856445, + "learning_rate": 4.409017059301382e-06, + "loss": 1.6027, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6653318405151367, + "learning_rate": 4.406986190089358e-06, + "loss": 1.5909, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.8076283931732178, + "eval_runtime": 37.5073, + "eval_samples_per_second": 26.661, + "eval_steps_per_second": 1.12, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.626783847808838, + "learning_rate": 4.404955320877336e-06, + "loss": 1.6205, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.6929268836975098, + "learning_rate": 4.402924451665313e-06, + "loss": 1.6146, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.6806800365448, + "learning_rate": 4.400893582453291e-06, + "loss": 1.5702, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.5407259464263916, + "learning_rate": 4.398862713241267e-06, + "loss": 1.5525, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.639122247695923, + "learning_rate": 4.396831844029245e-06, + "loss": 1.5826, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.669396162033081, + "learning_rate": 4.394800974817222e-06, + "loss": 1.5982, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.6377906799316406, + "learning_rate": 4.392770105605199e-06, + "loss": 1.5797, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6769638061523438, + "learning_rate": 4.390739236393176e-06, + "loss": 1.607, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.6347222328186035, + "learning_rate": 4.388708367181154e-06, + "loss": 1.6196, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.670680284500122, + "learning_rate": 4.386677497969131e-06, + "loss": 1.619, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.8124384880065918, + "eval_runtime": 37.4751, + "eval_samples_per_second": 26.684, + "eval_steps_per_second": 1.121, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.5886521339416504, + "learning_rate": 4.384646628757108e-06, + "loss": 1.6258, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.644566297531128, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6447, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.6432394981384277, + "learning_rate": 4.380584890333063e-06, + "loss": 1.5888, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.5602784156799316, + "learning_rate": 4.37855402112104e-06, + "loss": 1.5544, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.4874439239501953, + "learning_rate": 4.376523151909017e-06, + "loss": 1.5075, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.694748640060425, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.6498, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.582211494445801, + "learning_rate": 4.372461413484972e-06, + "loss": 1.5877, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.675713539123535, + "learning_rate": 4.370430544272949e-06, + "loss": 1.6045, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.7157161235809326, + "learning_rate": 4.368399675060926e-06, + "loss": 1.623, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.795076370239258, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6141, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.8062744140625, + "eval_runtime": 37.2435, + "eval_samples_per_second": 26.85, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6296374797821045, + "learning_rate": 4.364337936636881e-06, + "loss": 1.6087, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.362307067424858e-06, + "loss": 1.5182, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.7965915203094482, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.6025, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.807307481765747, + "learning_rate": 4.358245329000813e-06, + "loss": 1.574, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.584615468978882, + "learning_rate": 4.35621445978879e-06, + "loss": 1.5922, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.8472282886505127, + "learning_rate": 4.354183590576767e-06, + "loss": 1.5914, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.8471736907958984, + "learning_rate": 4.352152721364745e-06, + "loss": 1.5749, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.569225549697876, + "learning_rate": 4.350121852152722e-06, + "loss": 1.5591, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.688157796859741, + "learning_rate": 4.348090982940699e-06, + "loss": 1.6042, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.5865259170532227, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.6307, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.8040707111358643, + "eval_runtime": 36.9511, + "eval_samples_per_second": 27.063, + "eval_steps_per_second": 1.137, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.71850323677063, + "learning_rate": 4.344029244516654e-06, + "loss": 1.6043, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.7158093452453613, + "learning_rate": 4.341998375304631e-06, + "loss": 1.5907, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6879327297210693, + "learning_rate": 4.339967506092607e-06, + "loss": 1.6365, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.827277660369873, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.5928, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6709842681884766, + "learning_rate": 4.335905767668562e-06, + "loss": 1.5812, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.7071495056152344, + "learning_rate": 4.33387489845654e-06, + "loss": 1.5571, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6008732318878174, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.6048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.6238059997558594, + "learning_rate": 4.329813160032495e-06, + "loss": 1.5962, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.567065954208374, + "learning_rate": 4.327782290820471e-06, + "loss": 1.5713, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.573068857192993, + "learning_rate": 4.325751421608449e-06, + "loss": 1.536, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.8103471994400024, + "eval_runtime": 37.0956, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6245224475860596, + "learning_rate": 4.323720552396426e-06, + "loss": 1.5986, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.6243810653686523, + "learning_rate": 4.321689683184404e-06, + "loss": 1.589, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.5303313732147217, + "learning_rate": 4.31965881397238e-06, + "loss": 1.5793, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.5617806911468506, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.5704, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.7429656982421875, + "learning_rate": 4.315597075548335e-06, + "loss": 1.6174, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.6753251552581787, + "learning_rate": 4.313566206336312e-06, + "loss": 1.5879, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.5944488048553467, + "learning_rate": 4.311535337124289e-06, + "loss": 1.6082, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.9002904891967773, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.5521, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.5384247303009033, + "learning_rate": 4.307473598700244e-06, + "loss": 1.5751, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6015477180480957, + "learning_rate": 4.305442729488221e-06, + "loss": 1.5966, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8091154098510742, + "eval_runtime": 37.1075, + "eval_samples_per_second": 26.949, + "eval_steps_per_second": 1.132, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.6002886295318604, + "learning_rate": 4.303411860276198e-06, + "loss": 1.5403, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7755627632141113, + "learning_rate": 4.301380991064176e-06, + "loss": 1.6192, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.8351540565490723, + "learning_rate": 4.299350121852153e-06, + "loss": 1.6217, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.684007406234741, + "learning_rate": 4.29731925264013e-06, + "loss": 1.6242, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.8019614219665527, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6173, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.9122867584228516, + "learning_rate": 4.293257514216085e-06, + "loss": 1.5701, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.545732021331787, + "learning_rate": 4.291226645004062e-06, + "loss": 1.5494, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.9499566555023193, + "learning_rate": 4.289195775792039e-06, + "loss": 1.5848, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.8071892261505127, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6114, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.763864517211914, + "learning_rate": 4.285134037367994e-06, + "loss": 1.5768, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8057912588119507, + "eval_runtime": 37.0844, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 3.001190423965454, + "learning_rate": 4.283103168155971e-06, + "loss": 1.5601, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.7640297412872314, + "learning_rate": 4.281072298943948e-06, + "loss": 1.5619, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.7372794151306152, + "learning_rate": 4.279041429731926e-06, + "loss": 1.562, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 3.0390751361846924, + "learning_rate": 4.277010560519903e-06, + "loss": 1.5519, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.758798599243164, + "learning_rate": 4.27497969130788e-06, + "loss": 1.527, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.96443772315979, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.5899, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.7941088676452637, + "learning_rate": 4.270917952883835e-06, + "loss": 1.617, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 3.0385100841522217, + "learning_rate": 4.268887083671812e-06, + "loss": 1.5996, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.6946637630462646, + "learning_rate": 4.266856214459789e-06, + "loss": 1.5641, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.776723861694336, + "learning_rate": 4.264825345247766e-06, + "loss": 1.6451, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.7911298274993896, + "eval_runtime": 37.0623, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.851224184036255, + "learning_rate": 4.262794476035744e-06, + "loss": 1.5302, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9045867919921875, + "learning_rate": 4.26076360682372e-06, + "loss": 1.5808, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.6568782329559326, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.5233, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.94191837310791, + "learning_rate": 4.256701868399675e-06, + "loss": 1.5912, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.726778984069824, + "learning_rate": 4.254670999187653e-06, + "loss": 1.5823, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.8704583644866943, + "learning_rate": 4.252640129975629e-06, + "loss": 1.5931, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.5721547603607178, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6128, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6883509159088135, + "learning_rate": 4.248578391551584e-06, + "loss": 1.5479, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.7186272144317627, + "learning_rate": 4.246547522339562e-06, + "loss": 1.5203, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.8326518535614014, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.5547, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.7948570251464844, + "eval_runtime": 37.0971, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.7870523929595947, + "learning_rate": 4.242485783915517e-06, + "loss": 1.5891, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.866610288619995, + "learning_rate": 4.240454914703493e-06, + "loss": 1.5232, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.94161057472229, + "learning_rate": 4.23842404549147e-06, + "loss": 1.52, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.7331979274749756, + "learning_rate": 4.236393176279448e-06, + "loss": 1.5665, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.5375845432281494, + "learning_rate": 4.234362307067425e-06, + "loss": 1.5274, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.691927433013916, + "learning_rate": 4.232331437855402e-06, + "loss": 1.604, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7578394412994385, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.6, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.594538688659668, + "learning_rate": 4.228269699431357e-06, + "loss": 1.5928, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.5408968925476074, + "learning_rate": 4.226238830219334e-06, + "loss": 1.5908, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.689786672592163, + "learning_rate": 4.224207961007311e-06, + "loss": 1.5757, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.792494297027588, + "eval_runtime": 37.0822, + "eval_samples_per_second": 26.967, + "eval_steps_per_second": 1.133, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.735539197921753, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.5811, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.58634614944458, + "learning_rate": 4.220146222583266e-06, + "loss": 1.5808, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.7899138927459717, + "learning_rate": 4.218115353371243e-06, + "loss": 1.5801, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.6507744789123535, + "learning_rate": 4.21608448415922e-06, + "loss": 1.5585, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.6112217903137207, + "learning_rate": 4.214053614947198e-06, + "loss": 1.5329, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.729405164718628, + "learning_rate": 4.212022745735175e-06, + "loss": 1.6177, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.6431992053985596, + "learning_rate": 4.209991876523152e-06, + "loss": 1.5234, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.726243734359741, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.5926, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.6574180126190186, + "learning_rate": 4.205930138099107e-06, + "loss": 1.5424, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.6570956707000732, + "learning_rate": 4.203899268887084e-06, + "loss": 1.6119, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.781999945640564, + "eval_runtime": 37.3434, + "eval_samples_per_second": 26.778, + "eval_steps_per_second": 1.125, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.528052806854248, + "learning_rate": 4.201868399675061e-06, + "loss": 1.598, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.6075727939605713, + "learning_rate": 4.199837530463039e-06, + "loss": 1.5886, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.588737964630127, + "learning_rate": 4.197806661251016e-06, + "loss": 1.507, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.773618221282959, + "learning_rate": 4.195775792038993e-06, + "loss": 1.5794, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.624725580215454, + "learning_rate": 4.19374492282697e-06, + "loss": 1.5495, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.662397861480713, + "learning_rate": 4.191714053614948e-06, + "loss": 1.5203, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.6617255210876465, + "learning_rate": 4.189683184402924e-06, + "loss": 1.5885, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.5784451961517334, + "learning_rate": 4.187652315190902e-06, + "loss": 1.5689, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.726011037826538, + "learning_rate": 4.185621445978879e-06, + "loss": 1.5913, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.6117324829101562, + "learning_rate": 4.183590576766857e-06, + "loss": 1.5453, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.7893352508544922, + "eval_runtime": 37.4096, + "eval_samples_per_second": 26.731, + "eval_steps_per_second": 1.123, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7025442123413086, + "learning_rate": 4.181559707554833e-06, + "loss": 1.5868, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.656102180480957, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6141, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.741962194442749, + "learning_rate": 4.177497969130788e-06, + "loss": 1.5516, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5887229442596436, + "learning_rate": 4.175467099918766e-06, + "loss": 1.5525, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.542985439300537, + "learning_rate": 4.173436230706742e-06, + "loss": 1.592, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.7797129154205322, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.5349, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.632887363433838, + "learning_rate": 4.169374492282697e-06, + "loss": 1.5548, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.691690444946289, + "learning_rate": 4.167343623070675e-06, + "loss": 1.5352, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.8922817707061768, + "learning_rate": 4.165312753858652e-06, + "loss": 1.5717, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.5517446994781494, + "learning_rate": 4.163281884646629e-06, + "loss": 1.5658, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.780922770500183, + "eval_runtime": 37.3755, + "eval_samples_per_second": 26.755, + "eval_steps_per_second": 1.124, + "step": 700 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.182116586553344e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/zero_to_fp32.py b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_control_10M_seed0/runs/checkpoint-700/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_0.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..466c494b53a5cf3030ac585b29d1563c49b789d4 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a5075060d7b0d18a80b592cad740c3383144e4b5ba0bc7e7b9ff8ebb2afad8e5 +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_1.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..262d37671147896b5dfa4ab77f0e9578ea0d968c --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46766e2820279a98976091f452887a7ad876d6c74fa04b8df6451b77352f4c07 +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_2.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..b31f6e0c1e981c4c1dee9a4626e2a6c6ae5abb8e --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:deab7068d9411eb964839c7901679767e090930bd8dfb3faa9e266fdf4d5aed9 +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/scheduler.pt b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..8a9561293814a38b8c81e0bf7c0d1edb4bc5e59f --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d6e714958055feb5c2e0b60651507e978d59fb11b655f392b5b53afea5076078 +size 627 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/tokenizer.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..309789d5ccd06f418fbcfb7472a6a9affda68101 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d422bc7d90212ffb8070a743cdbe8ff2dfc5634b1fce8f2e3fa671ee441989eb +size 17210374 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/training_args.bin b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..6db59a81c34edebb00ef6a59f5669f3babfbb622 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:484dadb3af9420a36d698e37e41887c1bddbd89850c9b141ca33808576dc9968 +size 6011 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1500/model-00002-of-00002.safetensors b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1500/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5a6ac85260e5514e0691001840ceb980633b8586 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1500/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:71c276f8b124a573419358deeb972974dc5e6b8bf77a96c93b79e5a172df8538 +size 2247734920 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1900/model-00002-of-00002.safetensors b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1900/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3bcb0b4a921c22e34f82e5a7d3d288c23dbb890a --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1900/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fc8deafd7f7d7b0e3f51637b98210d68bf4a630642950873674ed397f36224ee +size 2247734920 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2500/model-00002-of-00002.safetensors b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2500/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..63a1e1640c284a5eea4249dd0650a614fdb862e9 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2500/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b38a392d587167dd21385ee406694fd106bc46727e4704d95c4d09e229f24e5 +size 2247734920 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2600/model-00002-of-00002.safetensors b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2600/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9d11ee4313205e9ec6ca4dcafa37dca528ac0699 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2600/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:34588dc8397442daa8d92f3e4af84c3273b8006ce30f6dff3f631cbb2aae4151 +size 2247734920 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_0.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..9d710ffc4a939f0f3079507659d737f4c9758140 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:732a42cf62774934ab4ef8d6e6f1cbe21e0aaa401b53fb48a60814cd815a5579 +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_1.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..ccc670a56d030f0a96c810818e5ec9f9819bd200 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f46b29c50087e92e8811493ebb75999bc8091f43308c94834e440988f96ae456 +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_2.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..56aa07e9a39fd447791b2cd09bf929cd86e351c3 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1fce192cc0c7cd38bc69f12c4ef2b4e1541c231484678601b065c9b17006ca7f +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/scheduler.pt b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..04704a1c6eb0de8ca4e3dd0fc77994b11b647ceb --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:377cf9bf003be07065b4ee960ea7de0ae6998486b85720accdfb9807691f8a91 +size 627 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/tokenizer.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..309789d5ccd06f418fbcfb7472a6a9affda68101 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d422bc7d90212ffb8070a743cdbe8ff2dfc5634b1fce8f2e3fa671ee441989eb +size 17210374 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/training_args.bin b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..6db59a81c34edebb00ef6a59f5669f3babfbb622 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:484dadb3af9420a36d698e37e41887c1bddbd89850c9b141ca33808576dc9968 +size 6011