diff --git a/.gitattributes b/.gitattributes index 24a035efab283f03c9f8e2c2c64be6deadbc9b36..319678a5361b9a3ae669c08dc10f71b791efea85 100644 --- a/.gitattributes +++ b/.gitattributes @@ -220,3 +220,5 @@ babylm_shuffle_local3_10M_seed0/artifacts/models--meta-llama--Llama-3.2-3B/blobs babylm_shuffle_local10_10M_seed0/artifacts/models--meta-llama--Llama-3.2-3B/blobs/584d8d3e3f82f7964955174dfe5e3b1cf117a9d859f022cfdf7fcb884856e002 filter=lfs diff=lfs merge=lfs -text babylm_shuffle_local3_10M_seed0/runs/checkpoint-2700/tokenizer.json filter=lfs diff=lfs merge=lfs -text babylm_shuffle_local3_10M_seed0/runs/checkpoint-200/tokenizer.json filter=lfs diff=lfs merge=lfs -text +babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/tokenizer.json filter=lfs diff=lfs merge=lfs -text +babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/babylm_shuffle_control_10M_seed0/artifacts/.locks/models--meta-llama--Llama-3.2-3B/2d73a6863086ff9d491c28e49df9fb697cd92c2b.lock b/babylm_shuffle_control_10M_seed0/artifacts/.locks/models--meta-llama--Llama-3.2-3B/2d73a6863086ff9d491c28e49df9fb697cd92c2b.lock new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/babylm_shuffle_control_10M_seed0/artifacts/.locks/models--meta-llama--Llama-3.2-3B/d3a1f0f5f401eeadca0c7a6786bd9e877fd42e58.lock b/babylm_shuffle_control_10M_seed0/artifacts/.locks/models--meta-llama--Llama-3.2-3B/d3a1f0f5f401eeadca0c7a6786bd9e877fd42e58.lock new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/babylm_shuffle_local10_10M_seed0/runs/checkpoint-1100/model-00002-of-00002.safetensors b/babylm_shuffle_local10_10M_seed0/runs/checkpoint-1100/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..89b3e9dc40633431024ab4a4cbe119a4d8412d77 --- /dev/null +++ b/babylm_shuffle_local10_10M_seed0/runs/checkpoint-1100/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57a0857f5f91e778b9296cb59af4e314f5305433df60a53f02dc2c881d3ec2d3 +size 2247734920 diff --git a/babylm_shuffle_local10_10M_seed0/runs/checkpoint-1400/model-00002-of-00002.safetensors b/babylm_shuffle_local10_10M_seed0/runs/checkpoint-1400/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a7ffb1a47f0e4d4f4cebb8a7a362c1338f3679b3 --- /dev/null +++ b/babylm_shuffle_local10_10M_seed0/runs/checkpoint-1400/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2864e479d5f813cf973b96bd96d089568a34635c77a40bcaf0bd5950aca5038f +size 2247734920 diff --git a/babylm_shuffle_local10_10M_seed0/runs/checkpoint-2500/model-00002-of-00002.safetensors b/babylm_shuffle_local10_10M_seed0/runs/checkpoint-2500/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f3084522c914abe1401d88b5355e03e61357f98a --- /dev/null +++ b/babylm_shuffle_local10_10M_seed0/runs/checkpoint-2500/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7affe3617d34a49cc1c389102c07a05ff03b5e247defa5f3c3c0da0f159fab44 +size 2247734920 diff --git a/babylm_shuffle_local10_10M_seed0/runs/checkpoint-2700/model-00002-of-00002.safetensors b/babylm_shuffle_local10_10M_seed0/runs/checkpoint-2700/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bd3e96d21e6afb0354a9f63d3e20bba64ebd15e8 --- /dev/null +++ b/babylm_shuffle_local10_10M_seed0/runs/checkpoint-2700/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:806637ad0286aeb7d8985cbea4abf621d16ea978e3d6d94a624df893bba74d1b +size 2247734920 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/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_local3_10M_seed0/runs/checkpoint-100/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/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_local3_10M_seed0/runs/checkpoint-100/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/latest new file mode 100644 index 0000000000000000000000000000000000000000..744ae7dbad571b6f37ec6c7066549494261bb59e --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/latest @@ -0,0 +1 @@ +global_step100 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-100/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/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_local3_10M_seed0/runs/checkpoint-100/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-100/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..5a2f9c05a30ce0eb8bb5204de33b8271a658434e --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/trainer_state.json @@ -0,0 +1,813 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.1095890410958904, + "eval_steps": 10, + "global_step": 100, + "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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + } + ], + "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": 3.11730940936192e+16, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-100/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1000/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/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_local3_10M_seed0/runs/checkpoint-1000/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/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_local3_10M_seed0/runs/checkpoint-1000/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/latest new file mode 100644 index 0000000000000000000000000000000000000000..e2d3435fb1acf8913e6bd6c51b01adfc69b11ac6 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/latest @@ -0,0 +1 @@ +global_step1000 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/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_local3_10M_seed0/runs/checkpoint-1000/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/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_local3_10M_seed0/runs/checkpoint-1000/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/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_local3_10M_seed0/runs/checkpoint-1000/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..ec31d2cc5cb8e52938ddc926caf75f20b015f4e2 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/trainer_state.json @@ -0,0 +1,7833 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.095890410958904, + "eval_steps": 10, + "global_step": 1000, + "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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.7659518718719482, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.279, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.433785915374756, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.3368, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 3.1486687660217285, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3896, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.391956090927124, + "learning_rate": 1.624087591240876e-06, + "loss": 2.297, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.6649584770202637, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.2415, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.5604891777038574, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2688, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.47900652885437, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2933, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.632723569869995, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.2143, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.8085947036743164, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2748, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.3927979469299316, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2685, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.355207920074463, + "eval_runtime": 42.3505, + "eval_samples_per_second": 23.612, + "eval_steps_per_second": 0.992, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.6170592308044434, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.3121, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.699814796447754, + "learning_rate": 1.77007299270073e-06, + "loss": 2.2853, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.3450610637664795, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3297, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.458219528198242, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2639, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.7101449966430664, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2134, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.5734996795654297, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2347, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 2.5200352668762207, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2299, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5022480487823486, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2253, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.190117359161377, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.2272, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.414524555206299, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2424, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.317634344100952, + "eval_runtime": 43.5742, + "eval_samples_per_second": 22.949, + "eval_steps_per_second": 0.964, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.3226804733276367, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1449, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.733006477355957, + "learning_rate": 1.952554744525548e-06, + "loss": 2.2027, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 2.5586233139038086, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2007, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.669914722442627, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2024, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.6148102283477783, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2496, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 2.331369638442993, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1814, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.3051886558532715, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2015, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.6890218257904053, + "learning_rate": 2.062043795620438e-06, + "loss": 2.184, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.4975883960723877, + "learning_rate": 2.08029197080292e-06, + "loss": 2.2439, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.5848891735076904, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1413, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.280966281890869, + "eval_runtime": 43.1001, + "eval_samples_per_second": 23.202, + "eval_steps_per_second": 0.974, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.2104, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.3356540203094482, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2312, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 2.3616602420806885, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.1399, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.581623077392578, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1456, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.547288656234741, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.184, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6386160850524902, + "learning_rate": 2.208029197080292e-06, + "loss": 2.1658, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.593065023422241, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1443, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.530632972717285, + "learning_rate": 2.244525547445256e-06, + "loss": 2.194, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.3553175926208496, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.1918, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 2.7536532878875732, + "learning_rate": 2.281021897810219e-06, + "loss": 2.1477, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2608063220977783, + "eval_runtime": 42.7938, + "eval_samples_per_second": 23.368, + "eval_steps_per_second": 0.981, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.160094738006592, + "learning_rate": 2.299270072992701e-06, + "loss": 2.133, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.4474735260009766, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1468, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.6654434204101562, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.1194, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.726611375808716, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1518, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 2.3560168743133545, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1537, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.664065361022949, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0888, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.3946099281311035, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.0998, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.7812652587890625, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1231, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.64595890045166, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1746, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.788524866104126, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0731, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.2340502738952637, + "eval_runtime": 42.3104, + "eval_samples_per_second": 23.635, + "eval_steps_per_second": 0.993, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 2.9249846935272217, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0564, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.5507798194885254, + "learning_rate": 2.5e-06, + "loss": 2.1714, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.1604716777801514, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0701, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.5429840087890625, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.0662, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8296737670898438, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.1378, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.9396755695343018, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1069, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.736814022064209, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1076, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 2.774909496307373, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1286, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.0861318111419678, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1499, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.624016284942627, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0989, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.215052604675293, + "eval_runtime": 43.9956, + "eval_samples_per_second": 22.73, + "eval_steps_per_second": 0.955, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.801500082015991, + "learning_rate": 2.664233576642336e-06, + "loss": 2.0795, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 2.2107114791870117, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.1035, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.030381441116333, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.0794, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 2.9719064235687256, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1496, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.688839912414551, + "learning_rate": 2.737226277372263e-06, + "loss": 2.1152, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 2.9350979328155518, + "learning_rate": 2.755474452554745e-06, + "loss": 2.0773, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.8958046436309814, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.1805, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.0781922340393066, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.1249, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.032059907913208, + "learning_rate": 2.81021897810219e-06, + "loss": 2.0632, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.798758029937744, + "learning_rate": 2.828467153284672e-06, + "loss": 2.1001, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2005105018615723, + "eval_runtime": 43.5111, + "eval_samples_per_second": 22.983, + "eval_steps_per_second": 0.965, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.7674927711486816, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.0211, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 2.6168575286865234, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0599, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 2.933353900909424, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.0169, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.81402587890625, + "learning_rate": 2.901459854014599e-06, + "loss": 2.122, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.5592572689056396, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.0957, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.738271951675415, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.0868, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.5362703800201416, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1237, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.6304945945739746, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.0825, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.040717601776123, + "learning_rate": 2.992700729927008e-06, + "loss": 2.0388, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.6226463317871094, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0163, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.186411142349243, + "eval_runtime": 44.625, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.5307068824768066, + "learning_rate": 3.029197080291971e-06, + "loss": 2.1097, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 2.9160444736480713, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.0946, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.5250024795532227, + "learning_rate": 3.065693430656935e-06, + "loss": 2.0709, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.6963064670562744, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.0937, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 2.845707416534424, + "learning_rate": 3.102189781021898e-06, + "loss": 2.0851, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.8985090255737305, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.0176, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 2.734377861022949, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.9931, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.5125555992126465, + "learning_rate": 3.156934306569343e-06, + "loss": 2.1133, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.0393502712249756, + "learning_rate": 3.175182481751825e-06, + "loss": 2.0636, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.6410164833068848, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0304, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.171868085861206, + "eval_runtime": 44.2249, + "eval_samples_per_second": 22.612, + "eval_steps_per_second": 0.95, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.7937498092651367, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0518, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7801201343536377, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.026, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7963638305664062, + "learning_rate": 3.248175182481752e-06, + "loss": 1.9642, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6330482959747314, + "learning_rate": 3.266423357664234e-06, + "loss": 2.0545, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.9771625995635986, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0445, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.6112730503082275, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0616, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.732865333557129, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0403, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.9113688468933105, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0382, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.4516773223876953, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.0416, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.781735897064209, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0348, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.160499334335327, + "eval_runtime": 44.5652, + "eval_samples_per_second": 22.439, + "eval_steps_per_second": 0.942, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.9990546703338623, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0324, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.783071279525757, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0518, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.659040689468384, + "learning_rate": 3.43065693430657e-06, + "loss": 1.9722, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 2.799798011779785, + "learning_rate": 3.4489051094890513e-06, + "loss": 2.0362, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 3.025458812713623, + "learning_rate": 3.467153284671533e-06, + "loss": 1.9863, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 2.686415195465088, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0201, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.716064929962158, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0249, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.637653350830078, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0403, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 2.7686848640441895, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0626, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.721956729888916, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.1077, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.152338981628418, + "eval_runtime": 43.5455, + "eval_samples_per_second": 22.964, + "eval_steps_per_second": 0.965, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.92571759223938, + "learning_rate": 3.576642335766424e-06, + "loss": 2.0928, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.9189202785491943, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9315, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 2.990233898162842, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9427, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.702348232269287, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9768, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6949572563171387, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.9954, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.108405113220215, + "learning_rate": 3.667883211678833e-06, + "loss": 1.9631, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.738922119140625, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9644, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.7419424057006836, + "learning_rate": 3.704379562043796e-06, + "loss": 2.0675, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9853813648223877, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0245, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.4614126682281494, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.9665, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.1342060565948486, + "eval_runtime": 46.3954, + "eval_samples_per_second": 21.554, + "eval_steps_per_second": 0.905, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.6787819862365723, + "learning_rate": 3.7591240875912412e-06, + "loss": 2.0063, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 2.563525438308716, + "learning_rate": 3.777372262773723e-06, + "loss": 1.998, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.4981701374053955, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.9738, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 2.5357978343963623, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0221, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.5365488529205322, + "learning_rate": 3.832116788321168e-06, + "loss": 1.9664, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 2.669468641281128, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9577, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 2.6950953006744385, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9786, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8249378204345703, + "learning_rate": 3.886861313868613e-06, + "loss": 2.0041, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.5924317836761475, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0338, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.1282334327697754, + "learning_rate": 3.923357664233577e-06, + "loss": 1.9698, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.1221420764923096, + "eval_runtime": 47.1127, + "eval_samples_per_second": 21.226, + "eval_steps_per_second": 0.891, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.7878193855285645, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0446, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.7137255668640137, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0575, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.286421060562134, + "learning_rate": 3.978102189781022e-06, + "loss": 1.9131, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.5466175079345703, + "learning_rate": 3.996350364963504e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.792483329772949, + "learning_rate": 4.014598540145986e-06, + "loss": 1.912, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.7533867359161377, + "learning_rate": 4.032846715328467e-06, + "loss": 1.9742, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 2.848405361175537, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9815, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.1034414768218994, + "learning_rate": 4.069343065693431e-06, + "loss": 1.9409, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.5109779834747314, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9676, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.486506938934326, + "learning_rate": 4.105839416058395e-06, + "loss": 2.0353, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1102187633514404, + "eval_runtime": 45.384, + "eval_samples_per_second": 22.034, + "eval_steps_per_second": 0.925, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.419269323348999, + "learning_rate": 4.124087591240876e-06, + "loss": 1.9622, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.4967262744903564, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.9523, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 2.7311460971832275, + "learning_rate": 4.16058394160584e-06, + "loss": 2.0234, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.241004467010498, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.907, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.821537733078003, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9761, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 2.5711426734924316, + "learning_rate": 4.215328467153285e-06, + "loss": 1.9411, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.936594009399414, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9432, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.6091108322143555, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9545, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.522674798965454, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9246, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.984649181365967, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9417, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1031947135925293, + "eval_runtime": 47.5409, + "eval_samples_per_second": 21.035, + "eval_steps_per_second": 0.883, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.8250324726104736, + "learning_rate": 4.306569343065693e-06, + "loss": 1.9061, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.7654290199279785, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9653, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.7091264724731445, + "learning_rate": 4.343065693430658e-06, + "loss": 1.9482, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 2.8019583225250244, + "learning_rate": 4.361313868613139e-06, + "loss": 1.9332, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.147368907928467, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9836, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.7050259113311768, + "learning_rate": 4.397810218978102e-06, + "loss": 1.9607, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.738743543624878, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9374, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.122777223587036, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9249, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.850745439529419, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9707, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.391998291015625, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.9703, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.0887911319732666, + "eval_runtime": 48.5149, + "eval_samples_per_second": 20.612, + "eval_steps_per_second": 0.866, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 2.9131698608398438, + "learning_rate": 4.489051094890512e-06, + "loss": 1.9989, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.1045937538146973, + "learning_rate": 4.507299270072993e-06, + "loss": 1.9718, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.7984158992767334, + "learning_rate": 4.525547445255475e-06, + "loss": 1.947, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.6223504543304443, + "learning_rate": 4.543795620437956e-06, + "loss": 1.8942, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.8221523761749268, + "learning_rate": 4.562043795620438e-06, + "loss": 2.0045, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.6657445430755615, + "learning_rate": 4.580291970802921e-06, + "loss": 2.0053, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 2.793455123901367, + "learning_rate": 4.598540145985402e-06, + "loss": 1.9577, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.551201820373535, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9256, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 2.739194631576538, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9265, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.2917332649230957, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9664, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.086106777191162, + "eval_runtime": 48.995, + "eval_samples_per_second": 20.41, + "eval_steps_per_second": 0.857, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 2.6489369869232178, + "learning_rate": 4.671532846715329e-06, + "loss": 1.9629, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 2.60017466545105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.932, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.569115400314331, + "learning_rate": 4.708029197080292e-06, + "loss": 1.9902, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 2.7225773334503174, + "learning_rate": 4.726277372262774e-06, + "loss": 1.9619, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.5064749717712402, + "learning_rate": 4.744525547445255e-06, + "loss": 1.8921, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.612931966781616, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9584, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.2524304389953613, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9675, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.8852503299713135, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9107, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 2.654776096343994, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9465, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.679086685180664, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9468, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.0842835903167725, + "eval_runtime": 51.7803, + "eval_samples_per_second": 19.312, + "eval_steps_per_second": 0.811, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.5110855102539062, + "learning_rate": 4.854014598540146e-06, + "loss": 1.88, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.1786117553710938, + "learning_rate": 4.872262773722628e-06, + "loss": 1.8994, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.388328790664673, + "learning_rate": 4.89051094890511e-06, + "loss": 1.8914, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 2.784834384918213, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.923, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.470426559448242, + "learning_rate": 4.927007299270074e-06, + "loss": 1.944, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.690580129623413, + "learning_rate": 4.945255474452555e-06, + "loss": 1.8978, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.869737386703491, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9964, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.9965591430664062, + "learning_rate": 4.981751824817518e-06, + "loss": 1.9706, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.6665658950805664, + "learning_rate": 5e-06, + "loss": 1.9313, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.9231722354888916, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9504, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.068364143371582, + "eval_runtime": 47.8334, + "eval_samples_per_second": 20.906, + "eval_steps_per_second": 0.878, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.6094462871551514, + "learning_rate": 4.995938261575955e-06, + "loss": 1.862, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.780848741531372, + "learning_rate": 4.993907392363932e-06, + "loss": 1.95, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.948662757873535, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.878, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 2.9622998237609863, + "learning_rate": 4.989845653939887e-06, + "loss": 1.8954, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 2.6401216983795166, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9021, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.029592275619507, + "learning_rate": 4.985783915515841e-06, + "loss": 1.8967, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.828392505645752, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9134, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.613926649093628, + "learning_rate": 4.981722177091796e-06, + "loss": 1.8829, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.6173207759857178, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9225, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.4657015800476074, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9142, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.059906482696533, + "eval_runtime": 49.6267, + "eval_samples_per_second": 20.15, + "eval_steps_per_second": 0.846, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7448673248291016, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9545, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.639096736907959, + "learning_rate": 4.973598700243705e-06, + "loss": 1.8664, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 2.6445884704589844, + "learning_rate": 4.971567831031681e-06, + "loss": 1.888, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.9130005836486816, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7717, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.48093318939209, + "learning_rate": 4.967506092607636e-06, + "loss": 1.8538, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.8918230533599854, + "learning_rate": 4.965475223395614e-06, + "loss": 1.886, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.511747121810913, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.8633, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.860029697418213, + "learning_rate": 4.961413484971569e-06, + "loss": 1.9063, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7304794788360596, + "learning_rate": 4.959382615759545e-06, + "loss": 1.8872, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.6569266319274902, + "learning_rate": 4.957351746547523e-06, + "loss": 1.8721, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0610575675964355, + "eval_runtime": 43.9292, + "eval_samples_per_second": 22.764, + "eval_steps_per_second": 0.956, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.1482994556427, + "learning_rate": 4.9553208773355e-06, + "loss": 2.0123, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.59568190574646, + "learning_rate": 4.953290008123478e-06, + "loss": 1.902, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.5897581577301025, + "learning_rate": 4.951259138911454e-06, + "loss": 1.8406, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.734032154083252, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.9605, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.8088371753692627, + "learning_rate": 4.947197400487409e-06, + "loss": 1.896, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.9678707122802734, + "learning_rate": 4.945166531275386e-06, + "loss": 1.8625, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.261200189590454, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8884, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 2.5811619758605957, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9311, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.6783649921417236, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9324, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.6533944606781006, + "learning_rate": 4.937043054427295e-06, + "loss": 1.8575, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.049668312072754, + "eval_runtime": 47.7347, + "eval_samples_per_second": 20.949, + "eval_steps_per_second": 0.88, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 3.0882792472839355, + "learning_rate": 4.935012185215272e-06, + "loss": 1.84, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.8259081840515137, + "learning_rate": 4.93298131600325e-06, + "loss": 1.9109, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.540908098220825, + "learning_rate": 4.930950446791227e-06, + "loss": 1.8906, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.9277396202087402, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9047, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.392449140548706, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.876, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 3.1890463829040527, + "learning_rate": 4.924857839155159e-06, + "loss": 1.8709, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.725710391998291, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8953, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.6287014484405518, + "learning_rate": 4.920796100731113e-06, + "loss": 1.8204, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.7316651344299316, + "learning_rate": 4.918765231519091e-06, + "loss": 1.882, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.5561037063598633, + "learning_rate": 4.916734362307068e-06, + "loss": 1.8331, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0412285327911377, + "eval_runtime": 50.237, + "eval_samples_per_second": 19.906, + "eval_steps_per_second": 0.836, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.974501132965088, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8414, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.4530839920043945, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8323, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.789569616317749, + "learning_rate": 4.910641754671e-06, + "loss": 1.8697, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.582704544067383, + "learning_rate": 4.908610885458977e-06, + "loss": 1.8592, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 2.776442289352417, + "learning_rate": 4.906580016246954e-06, + "loss": 1.912, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.670191764831543, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.8899, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.4534454345703125, + "learning_rate": 4.902518277822909e-06, + "loss": 1.8878, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.007469654083252, + "learning_rate": 4.900487408610886e-06, + "loss": 1.8977, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 2.856466293334961, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9062, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.817441463470459, + "learning_rate": 4.89642567018684e-06, + "loss": 1.9686, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.0344440937042236, + "eval_runtime": 47.2472, + "eval_samples_per_second": 21.165, + "eval_steps_per_second": 0.889, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.9732158184051514, + "learning_rate": 4.894394800974818e-06, + "loss": 1.8332, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 3.0413055419921875, + "learning_rate": 4.892363931762794e-06, + "loss": 1.9171, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.646749973297119, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.8403, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.1274876594543457, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8218, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.8122124671936035, + "learning_rate": 4.886271324126727e-06, + "loss": 1.9272, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.4608492851257324, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.8974, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.928083658218384, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.8646, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.6308164596557617, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8219, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7311816215515137, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8326, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.640162467956543, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8013, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0261101722717285, + "eval_runtime": 47.0605, + "eval_samples_per_second": 21.249, + "eval_steps_per_second": 0.892, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.8542518615722656, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8233, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.464303970336914, + "learning_rate": 4.872055239642567e-06, + "loss": 1.787, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8535103797912598, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8535, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.5060505867004395, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8733, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.6549222469329834, + "learning_rate": 4.865962632006499e-06, + "loss": 1.8047, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.8636224269866943, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8543, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.644256830215454, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8853, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.949490547180176, + "learning_rate": 4.859870024370431e-06, + "loss": 1.8951, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.4576282501220703, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9069, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.6140542030334473, + "learning_rate": 4.855808285946385e-06, + "loss": 1.8573, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0159800052642822, + "eval_runtime": 44.2447, + "eval_samples_per_second": 22.602, + "eval_steps_per_second": 0.949, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.525766134262085, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.844, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.5479896068573, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8572, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.628570795059204, + "learning_rate": 4.849715678310317e-06, + "loss": 1.865, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.2768666744232178, + "learning_rate": 4.847684809098294e-06, + "loss": 1.8355, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.635516881942749, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8603, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.5307435989379883, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8482, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.3688952922821045, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8894, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.6000115871429443, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.8045, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7782504558563232, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8283, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.6914777755737305, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8765, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0086047649383545, + "eval_runtime": 48.245, + "eval_samples_per_second": 20.728, + "eval_steps_per_second": 0.871, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.598365306854248, + "learning_rate": 4.833468724614135e-06, + "loss": 1.8461, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.6718530654907227, + "learning_rate": 4.831437855402113e-06, + "loss": 1.7453, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.393528699874878, + "learning_rate": 4.82940698619009e-06, + "loss": 1.8275, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 3.0622947216033936, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8658, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.431041955947876, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8601, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.939424514770508, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8649, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.8116796016693115, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8303, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.511960029602051, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8383, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 3.8372581005096436, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8111, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.511948347091675, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8448, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.9998281002044678, + "eval_runtime": 44.7624, + "eval_samples_per_second": 22.34, + "eval_steps_per_second": 0.938, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 3.5780487060546875, + "learning_rate": 4.813160032493907e-06, + "loss": 1.8161, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.81740403175354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.8633, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.8117666244506836, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8385, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.5791451930999756, + "learning_rate": 4.80706742485784e-06, + "loss": 1.896, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6816465854644775, + "learning_rate": 4.805036555645816e-06, + "loss": 1.8126, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.6419460773468018, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8753, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.666086196899414, + "learning_rate": 4.800974817221771e-06, + "loss": 1.8305, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8626585006713867, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8463, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.351825475692749, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.838, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.300717830657959, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8711, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.9904799461364746, + "eval_runtime": 43.9511, + "eval_samples_per_second": 22.753, + "eval_steps_per_second": 0.956, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.43884015083313, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8145, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.50593900680542, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8103, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.4643657207489014, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8162, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.4615302085876465, + "learning_rate": 4.786758732737612e-06, + "loss": 1.8637, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.6566433906555176, + "learning_rate": 4.784727863525589e-06, + "loss": 1.7805, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.5973432064056396, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8327, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.4996957778930664, + "learning_rate": 4.780666125101544e-06, + "loss": 1.7761, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 3.0266876220703125, + "learning_rate": 4.778635255889521e-06, + "loss": 1.8036, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.522369146347046, + "learning_rate": 4.776604386677498e-06, + "loss": 1.814, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.482727527618408, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8665, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.9888195991516113, + "eval_runtime": 44.0755, + "eval_samples_per_second": 22.688, + "eval_steps_per_second": 0.953, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 3.237544059753418, + "learning_rate": 4.772542648253453e-06, + "loss": 1.8642, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.393286943435669, + "learning_rate": 4.77051177904143e-06, + "loss": 1.7857, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.66072154045105, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8423, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.6725080013275146, + "learning_rate": 4.766450040617385e-06, + "loss": 1.815, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.7657666206359863, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8729, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.3153769969940186, + "learning_rate": 4.762388302193339e-06, + "loss": 1.8349, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5804553031921387, + "learning_rate": 4.760357432981316e-06, + "loss": 1.8155, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.7280807495117188, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8262, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.551940441131592, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8332, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.6137285232543945, + "learning_rate": 4.754264825345248e-06, + "loss": 1.7805, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.983920693397522, + "eval_runtime": 42.525, + "eval_samples_per_second": 23.516, + "eval_steps_per_second": 0.988, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.4460153579711914, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.7953, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.556032657623291, + "learning_rate": 4.750203086921203e-06, + "loss": 1.7393, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.335983991622925, + "learning_rate": 4.74817221770918e-06, + "loss": 1.8494, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.5057711601257324, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8663, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4447145462036133, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7532, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.769984483718872, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7755, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.472193956375122, + "learning_rate": 4.740048740861089e-06, + "loss": 1.8056, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.4286932945251465, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8094, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.0248074531555176, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8497, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.607700824737549, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8741, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.9779797792434692, + "eval_runtime": 43.7734, + "eval_samples_per_second": 22.845, + "eval_steps_per_second": 0.959, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 3.035045623779297, + "learning_rate": 4.731925264012998e-06, + "loss": 1.8438, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.786921977996826, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7666, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.4158425331115723, + "learning_rate": 4.727863525588953e-06, + "loss": 1.7464, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7772, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.8447, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 3.08266282081604, + "learning_rate": 4.723801787164907e-06, + "loss": 1.7812, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.791749954223633, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8047, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6457693576812744, + "learning_rate": 4.719740048740861e-06, + "loss": 1.836, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.69480037689209, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8145, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6356561183929443, + "learning_rate": 4.715678310316816e-06, + "loss": 1.7756, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.9723366498947144, + "eval_runtime": 45.5806, + "eval_samples_per_second": 21.939, + "eval_steps_per_second": 0.921, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.53474497795105, + "learning_rate": 4.713647441104794e-06, + "loss": 1.774, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.3631107807159424, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8517, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.594604730606079, + "learning_rate": 4.709585702680748e-06, + "loss": 1.8962, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.5280940532684326, + "learning_rate": 4.707554833468725e-06, + "loss": 1.7759, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.4596288204193115, + "learning_rate": 4.705523964256703e-06, + "loss": 1.7948, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.9123644828796387, + "learning_rate": 4.703493095044679e-06, + "loss": 1.7399, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.4354264736175537, + "learning_rate": 4.701462225832657e-06, + "loss": 1.8659, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.28625226020813, + "learning_rate": 4.699431356620634e-06, + "loss": 1.831, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.5445950031280518, + "learning_rate": 4.697400487408612e-06, + "loss": 1.8121, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.767825126647949, + "learning_rate": 4.695369618196588e-06, + "loss": 1.7874, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9653255939483643, + "eval_runtime": 45.2453, + "eval_samples_per_second": 22.102, + "eval_steps_per_second": 0.928, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4052765369415283, + "learning_rate": 4.693338748984566e-06, + "loss": 1.7539, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.498098373413086, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8343, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.821753740310669, + "learning_rate": 4.68927701056052e-06, + "loss": 1.8119, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.4158449172973633, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.7982, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.5229434967041016, + "learning_rate": 4.685215272136475e-06, + "loss": 1.806, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.729199171066284, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8768, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.5102570056915283, + "learning_rate": 4.681153533712429e-06, + "loss": 1.775, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4463515281677246, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7493, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.505733013153076, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8462, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.5784912109375, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7381, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9640673398971558, + "eval_runtime": 44.2714, + "eval_samples_per_second": 22.588, + "eval_steps_per_second": 0.949, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.7864291667938232, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7612, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.507809638977051, + "learning_rate": 4.670999187652316e-06, + "loss": 1.757, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.5499937534332275, + "learning_rate": 4.668968318440293e-06, + "loss": 1.8243, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.435472011566162, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7269, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.541031837463379, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.7959, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5052099227905273, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7191, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6285109519958496, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8423, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.7825605869293213, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7178, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.677319288253784, + "learning_rate": 4.656783103168157e-06, + "loss": 1.7939, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.6830129623413086, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7842, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.960136890411377, + "eval_runtime": 45.9379, + "eval_samples_per_second": 21.769, + "eval_steps_per_second": 0.914, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.5368666648864746, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8352, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.8922412395477295, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.8252, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.8332598209381104, + "learning_rate": 4.648659626320066e-06, + "loss": 1.7805, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.420780658721924, + "learning_rate": 4.646628757108043e-06, + "loss": 1.7499, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.6684062480926514, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.7857, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.8547494411468506, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.8115, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.59279203414917, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8254, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.80961275100708, + "learning_rate": 4.638505280259952e-06, + "loss": 1.7901, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.9017655849456787, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8265, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.595454216003418, + "learning_rate": 4.634443541835907e-06, + "loss": 1.7833, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9576054811477661, + "eval_runtime": 50.7265, + "eval_samples_per_second": 19.714, + "eval_steps_per_second": 0.828, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 4.260303020477295, + "learning_rate": 4.632412672623883e-06, + "loss": 1.6741, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.366147756576538, + "learning_rate": 4.630381803411861e-06, + "loss": 1.7778, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.894336462020874, + "learning_rate": 4.628350934199838e-06, + "loss": 1.8268, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.7675349712371826, + "learning_rate": 4.626320064987816e-06, + "loss": 1.7748, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.461413621902466, + "learning_rate": 4.624289195775792e-06, + "loss": 1.7599, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.670187473297119, + "learning_rate": 4.62225832656377e-06, + "loss": 1.7736, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.8119454383850098, + "learning_rate": 4.620227457351747e-06, + "loss": 1.7837, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.8020036220550537, + "learning_rate": 4.618196588139724e-06, + "loss": 1.7508, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.5977094173431396, + "learning_rate": 4.616165718927701e-06, + "loss": 1.834, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.731919765472412, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.8137, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.9465162754058838, + "eval_runtime": 48.9829, + "eval_samples_per_second": 20.415, + "eval_steps_per_second": 0.857, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.604386568069458, + "learning_rate": 4.612103980503656e-06, + "loss": 1.7854, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.601107120513916, + "learning_rate": 4.610073111291633e-06, + "loss": 1.7749, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6758370399475098, + "learning_rate": 4.60804224207961e-06, + "loss": 1.7826, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.631549119949341, + "learning_rate": 4.606011372867588e-06, + "loss": 1.7723, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.6704764366149902, + "learning_rate": 4.603980503655565e-06, + "loss": 1.7088, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 3.078573703765869, + "learning_rate": 4.601949634443542e-06, + "loss": 1.7254, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7492356300354004, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7756, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 3.0763585567474365, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7806, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.5379223823547363, + "learning_rate": 4.595857026807474e-06, + "loss": 1.7516, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8343496322631836, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8111, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.945749282836914, + "eval_runtime": 46.5453, + "eval_samples_per_second": 21.484, + "eval_steps_per_second": 0.902, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.809753179550171, + "learning_rate": 4.591795288383429e-06, + "loss": 1.7877, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.4196107387542725, + "learning_rate": 4.589764419171406e-06, + "loss": 1.7457, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.409287452697754, + "learning_rate": 4.587733549959383e-06, + "loss": 1.7338, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.414706230163574, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7185, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.591365337371826, + "learning_rate": 4.583671811535338e-06, + "loss": 1.8083, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.4648489952087402, + "learning_rate": 4.581640942323315e-06, + "loss": 1.8192, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.3804972171783447, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8077, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.283082962036133, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8294, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.61795711517334, + "learning_rate": 4.575548334687247e-06, + "loss": 1.7322, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.4035353660583496, + "learning_rate": 4.573517465475224e-06, + "loss": 1.7798, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9422866106033325, + "eval_runtime": 45.3068, + "eval_samples_per_second": 22.072, + "eval_steps_per_second": 0.927, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.7253756523132324, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7876, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.738295078277588, + "learning_rate": 4.569455727051178e-06, + "loss": 1.7933, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 3.1418697834014893, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8145, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.941603660583496, + "learning_rate": 4.565393988627132e-06, + "loss": 1.7671, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.063924551010132, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.7809, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 4.957335472106934, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7306, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.651716709136963, + "learning_rate": 4.559301380991065e-06, + "loss": 1.7428, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 4.777034282684326, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.8118, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.5549471378326416, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7294, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.474388599395752, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8338, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.9402854442596436, + "eval_runtime": 43.166, + "eval_samples_per_second": 23.166, + "eval_steps_per_second": 0.973, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.8968348503112793, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7432, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.6996726989746094, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.7837, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.4599392414093018, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7388, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.745487928390503, + "learning_rate": 4.545085296506905e-06, + "loss": 1.7439, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.6960670948028564, + "learning_rate": 4.543054427294882e-06, + "loss": 1.737, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.596876621246338, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7708, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.598267078399658, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7541, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.6069958209991455, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7756, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.490694761276245, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.7372, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.44179368019104, + "learning_rate": 4.532900081234769e-06, + "loss": 1.7652, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.934715986251831, + "eval_runtime": 44.688, + "eval_samples_per_second": 22.377, + "eval_steps_per_second": 0.94, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 3.018480062484741, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7596, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.5316643714904785, + "learning_rate": 4.528838342810723e-06, + "loss": 1.7715, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4893109798431396, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7549, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.524834156036377, + "learning_rate": 4.524776604386678e-06, + "loss": 1.8046, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7527239322662354, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7583, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.5210752487182617, + "learning_rate": 4.520714865962632e-06, + "loss": 1.7896, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.5787770748138428, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7355, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.537195920944214, + "learning_rate": 4.516653127538587e-06, + "loss": 1.7272, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.6810014247894287, + "learning_rate": 4.514622258326564e-06, + "loss": 1.753, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.5868756771087646, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7964, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.934499740600586, + "eval_runtime": 45.6556, + "eval_samples_per_second": 21.903, + "eval_steps_per_second": 0.92, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.6830856800079346, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7072, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.6574594974517822, + "learning_rate": 4.508529650690496e-06, + "loss": 1.7004, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.4415769577026367, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7667, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.825115442276001, + "learning_rate": 4.504467912266451e-06, + "loss": 1.8356, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.949010133743286, + "learning_rate": 4.502437043054428e-06, + "loss": 1.7579, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.690986394882202, + "learning_rate": 4.500406173842405e-06, + "loss": 1.7657, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 3.1015520095825195, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.7432, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.608424186706543, + "learning_rate": 4.49634443541836e-06, + "loss": 1.8112, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.5238773822784424, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7724, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.7712161540985107, + "learning_rate": 4.492282696994314e-06, + "loss": 1.7111, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9326083660125732, + "eval_runtime": 48.5359, + "eval_samples_per_second": 20.603, + "eval_steps_per_second": 0.865, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.4763033390045166, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7268, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.6310081481933594, + "learning_rate": 4.488220958570269e-06, + "loss": 1.6885, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.5950350761413574, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7775, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.769026517868042, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7737, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.4898502826690674, + "learning_rate": 4.4821283509342e-06, + "loss": 1.751, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.668081521987915, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7178, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.441615104675293, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7511, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.727665901184082, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.8639, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.7964634895324707, + "learning_rate": 4.474004874086109e-06, + "loss": 1.6945, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.6142687797546387, + "learning_rate": 4.471974004874087e-06, + "loss": 1.8028, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9281476736068726, + "eval_runtime": 48.078, + "eval_samples_per_second": 20.8, + "eval_steps_per_second": 0.874, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.6655778884887695, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.7632, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.931162118911743, + "learning_rate": 4.467912266450041e-06, + "loss": 1.7244, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.469895124435425, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7773, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.4948830604553223, + "learning_rate": 4.463850528025995e-06, + "loss": 1.788, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5961945056915283, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.756, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.452946424484253, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6968, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.9352009296417236, + "learning_rate": 4.457757920389927e-06, + "loss": 1.7414, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.7044763565063477, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7848, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.7200324535369873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7699, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.6185197830200195, + "learning_rate": 4.451665312753859e-06, + "loss": 1.7816, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9209110736846924, + "eval_runtime": 47.3321, + "eval_samples_per_second": 21.127, + "eval_steps_per_second": 0.887, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.3639564514160156, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7164, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7073519229888916, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.727, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.637080430984497, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7951, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.8013885021209717, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7333, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.512484312057495, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8048, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.3851466178894043, + "learning_rate": 4.439480097481723e-06, + "loss": 1.6667, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.8105392456054688, + "learning_rate": 4.4374492282697e-06, + "loss": 1.7622, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.555773973464966, + "learning_rate": 4.435418359057677e-06, + "loss": 1.7519, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.688596725463867, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7198, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.828646421432495, + "learning_rate": 4.431356620633632e-06, + "loss": 1.7478, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.919256567955017, + "eval_runtime": 50.5993, + "eval_samples_per_second": 19.763, + "eval_steps_per_second": 0.83, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.4523651599884033, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7344, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.4237422943115234, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6664, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.2746193408966064, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.7446, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.502819061279297, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7691, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.7032017707824707, + "learning_rate": 4.421202274573518e-06, + "loss": 1.6448, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.5879440307617188, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6908, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 3.2481982707977295, + "learning_rate": 4.417140536149473e-06, + "loss": 1.7725, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.4844820499420166, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7141, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.5596253871917725, + "learning_rate": 4.413078797725427e-06, + "loss": 1.7485, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 3.142839193344116, + "learning_rate": 4.411047928513404e-06, + "loss": 1.734, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9129430055618286, + "eval_runtime": 49.2608, + "eval_samples_per_second": 20.3, + "eval_steps_per_second": 0.853, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.5837552547454834, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7668, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.7043633460998535, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7497, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.9886765480041504, + "learning_rate": 4.404955320877336e-06, + "loss": 1.7082, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.3199501037597656, + "learning_rate": 4.402924451665313e-06, + "loss": 1.693, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.606682777404785, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7242, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.6967978477478027, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7516, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.351726531982422, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7338, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.590869188308716, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7462, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.3498220443725586, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7609, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.281002998352051, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7584, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.908841609954834, + "eval_runtime": 50.7771, + "eval_samples_per_second": 19.694, + "eval_steps_per_second": 0.827, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.468755006790161, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7654, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.3467307090759277, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7923, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.3723301887512207, + "learning_rate": 4.384646628757108e-06, + "loss": 1.74, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.4527792930603027, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6988, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.3923704624176025, + "learning_rate": 4.380584890333063e-06, + "loss": 1.6406, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.740556240081787, + "learning_rate": 4.37855402112104e-06, + "loss": 1.8032, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.3013832569122314, + "learning_rate": 4.376523151909017e-06, + "loss": 1.7259, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.4632391929626465, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.7462, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.4878177642822266, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7665, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.610898494720459, + "learning_rate": 4.370430544272949e-06, + "loss": 1.7646, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9056707620620728, + "eval_runtime": 51.8738, + "eval_samples_per_second": 19.278, + "eval_steps_per_second": 0.81, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.387721300125122, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7454, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.9966042041778564, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6789, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.664689064025879, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7486, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.4788103103637695, + "learning_rate": 4.362307067424858e-06, + "loss": 1.6981, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.4413328170776367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7279, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.6818156242370605, + "learning_rate": 4.358245329000813e-06, + "loss": 1.74, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.650348424911499, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7212, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.442805767059326, + "learning_rate": 4.354183590576767e-06, + "loss": 1.6951, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.6806585788726807, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7499, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.586792230606079, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7709, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.901674747467041, + "eval_runtime": 42.3186, + "eval_samples_per_second": 23.63, + "eval_steps_per_second": 0.992, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.547452688217163, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7458, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.5820348262786865, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.7232, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.4208321571350098, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7981, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.8791956901550293, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7372, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6537868976593018, + "learning_rate": 4.339967506092607e-06, + "loss": 1.7216, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.416097402572632, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.701, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.5565483570098877, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7558, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.5875823497772217, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7439, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5459702014923096, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.7058, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.4195566177368164, + "learning_rate": 4.329813160032495e-06, + "loss": 1.6749, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9035066366195679, + "eval_runtime": 43.7962, + "eval_samples_per_second": 22.833, + "eval_steps_per_second": 0.959, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6019818782806396, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7549, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.8627779483795166, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7257, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.582158088684082, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7138, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.4528613090515137, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7162, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 3.1956026554107666, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7556, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.3531901836395264, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.746, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.441988706588745, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7664, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.599506378173828, + "learning_rate": 4.313566206336312e-06, + "loss": 1.6955, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.4121932983398438, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7194, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.4803290367126465, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7386, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8993914127349854, + "eval_runtime": 43.0525, + "eval_samples_per_second": 23.227, + "eval_steps_per_second": 0.976, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.701594114303589, + "learning_rate": 4.307473598700244e-06, + "loss": 1.6796, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7144737243652344, + "learning_rate": 4.305442729488221e-06, + "loss": 1.763, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.395312786102295, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7522, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.3926732540130615, + "learning_rate": 4.301380991064176e-06, + "loss": 1.7705, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.6293702125549316, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7693, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.3746306896209717, + "learning_rate": 4.29731925264013e-06, + "loss": 1.72, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.5421926975250244, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6868, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.6991186141967773, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7378, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.463897705078125, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7628, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5139036178588867, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7125, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8909698724746704, + "eval_runtime": 42.5917, + "eval_samples_per_second": 23.479, + "eval_steps_per_second": 0.986, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.5220465660095215, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6975, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.432988405227661, + "learning_rate": 4.285134037367994e-06, + "loss": 1.6904, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.5894253253936768, + "learning_rate": 4.283103168155971e-06, + "loss": 1.6919, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5334458351135254, + "learning_rate": 4.281072298943948e-06, + "loss": 1.6968, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.510019302368164, + "learning_rate": 4.279041429731926e-06, + "loss": 1.6791, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.531538248062134, + "learning_rate": 4.277010560519903e-06, + "loss": 1.723, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.5621440410614014, + "learning_rate": 4.27497969130788e-06, + "loss": 1.749, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.594329833984375, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7579, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.321031093597412, + "learning_rate": 4.270917952883835e-06, + "loss": 1.6983, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.6325764656066895, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7872, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.886696219444275, + "eval_runtime": 43.4143, + "eval_samples_per_second": 23.034, + "eval_steps_per_second": 0.967, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5342679023742676, + "learning_rate": 4.266856214459789e-06, + "loss": 1.6655, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.5810720920562744, + "learning_rate": 4.264825345247766e-06, + "loss": 1.7182, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.7768406867980957, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6534, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.5680463314056396, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7243, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7162973880767822, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.7146, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.697925090789795, + "learning_rate": 4.256701868399675e-06, + "loss": 1.7296, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.4458425045013428, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7466, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6330654621124268, + "learning_rate": 4.252640129975629e-06, + "loss": 1.6949, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.350430727005005, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6574, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.505995512008667, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7111, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.8887447118759155, + "eval_runtime": 43.9304, + "eval_samples_per_second": 22.763, + "eval_steps_per_second": 0.956, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 3.036954879760742, + "learning_rate": 4.246547522339562e-06, + "loss": 1.7182, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.563606023788452, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.6566, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.516315460205078, + "learning_rate": 4.242485783915517e-06, + "loss": 1.6543, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.6289877891540527, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6996, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.7563092708587646, + "learning_rate": 4.23842404549147e-06, + "loss": 1.666, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5743000507354736, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7338, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 3.1975982189178467, + "learning_rate": 4.234362307067425e-06, + "loss": 1.7444, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.6022894382476807, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7367, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.4466192722320557, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.726, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.579015016555786, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7146, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.8904188871383667, + "eval_runtime": 43.8091, + "eval_samples_per_second": 22.826, + "eval_steps_per_second": 0.959, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.719646453857422, + "learning_rate": 4.226238830219334e-06, + "loss": 1.7209, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.4423880577087402, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7288, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5883753299713135, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7089, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.447049140930176, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7095, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.453739881515503, + "learning_rate": 4.218115353371243e-06, + "loss": 1.6747, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4540836811065674, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7578, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.5468671321868896, + "learning_rate": 4.214053614947198e-06, + "loss": 1.6556, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.443814277648926, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7307, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.4998748302459717, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6768, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.667025566101074, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7493, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.8830010890960693, + "eval_runtime": 44.7027, + "eval_samples_per_second": 22.37, + "eval_steps_per_second": 0.94, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.416248321533203, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7381, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.378025770187378, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7335, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.633173704147339, + "learning_rate": 4.201868399675061e-06, + "loss": 1.6484, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.549384117126465, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7143, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.6492598056793213, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6907, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.444086790084839, + "learning_rate": 4.195775792038993e-06, + "loss": 1.636, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.628814935684204, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7342, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.83831524848938, + "learning_rate": 4.191714053614948e-06, + "loss": 1.7028, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.467339515686035, + "learning_rate": 4.189683184402924e-06, + "loss": 1.749, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.4529521465301514, + "learning_rate": 4.187652315190902e-06, + "loss": 1.682, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.8826342821121216, + "eval_runtime": 44.7757, + "eval_samples_per_second": 22.334, + "eval_steps_per_second": 0.938, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7340762615203857, + "learning_rate": 4.185621445978879e-06, + "loss": 1.723, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.6028647422790527, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7566, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.3967695236206055, + "learning_rate": 4.181559707554833e-06, + "loss": 1.683, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5475900173187256, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6976, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.5411484241485596, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7489, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 3.458749294281006, + "learning_rate": 4.175467099918766e-06, + "loss": 1.6828, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.2977802753448486, + "learning_rate": 4.173436230706742e-06, + "loss": 1.6963, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.875307559967041, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.6857, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 4.097696304321289, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7225, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.6774847507476807, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7031, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.8783056735992432, + "eval_runtime": 44.0219, + "eval_samples_per_second": 22.716, + "eval_steps_per_second": 0.954, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.6787376403808594, + "learning_rate": 4.165312753858652e-06, + "loss": 1.6865, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.3803224563598633, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7237, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.9285683631896973, + "learning_rate": 4.161251015434607e-06, + "loss": 1.6575, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.5534448623657227, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6799, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4105820655822754, + "learning_rate": 4.157189277010561e-06, + "loss": 1.7128, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.6571805477142334, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7123, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.891935348510742, + "learning_rate": 4.153127538586516e-06, + "loss": 1.6949, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.8692519664764404, + "learning_rate": 4.151096669374492e-06, + "loss": 1.699, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.780078649520874, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7114, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.4797518253326416, + "learning_rate": 4.147034930950447e-06, + "loss": 1.6943, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.8726727962493896, + "eval_runtime": 43.5664, + "eval_samples_per_second": 22.953, + "eval_steps_per_second": 0.964, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.7010464668273926, + "learning_rate": 4.145004061738425e-06, + "loss": 1.6207, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8812713623046875, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.6608, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.6700279712677, + "learning_rate": 4.140942323314379e-06, + "loss": 1.677, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.796513557434082, + "learning_rate": 4.138911454102356e-06, + "loss": 1.7291, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.4274256229400635, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7058, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5528790950775146, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7064, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.4288926124572754, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7163, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.4939630031585693, + "learning_rate": 4.130787977254265e-06, + "loss": 1.721, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7909305095672607, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7777, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.3382463455200195, + "learning_rate": 4.12672623883022e-06, + "loss": 1.6975, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.8738009929656982, + "eval_runtime": 44.8142, + "eval_samples_per_second": 22.314, + "eval_steps_per_second": 0.937, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6876718997955322, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6851, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.450212240219116, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7209, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.4365246295928955, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.6545, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.3236255645751953, + "learning_rate": 4.118602761982129e-06, + "loss": 1.6434, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.4753000736236572, + "learning_rate": 4.116571892770106e-06, + "loss": 1.7471, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.84885835647583, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6777, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.4361016750335693, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7079, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.5868160724639893, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7288, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.6499862670898438, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7191, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.333289384841919, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5938, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.8746954202651978, + "eval_runtime": 46.8039, + "eval_samples_per_second": 21.366, + "eval_steps_per_second": 0.897, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.1299779415130615, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7585, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 3.6410014629364014, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6803, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.801745653152466, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6957, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 3.0350706577301025, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7031, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 3.5306341648101807, + "learning_rate": 4.096263200649879e-06, + "loss": 1.6598, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.591485023498535, + "learning_rate": 4.094232331437856e-06, + "loss": 1.6659, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 3.0916528701782227, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6951, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.459009885787964, + "learning_rate": 4.090170593013811e-06, + "loss": 1.6597, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.932072162628174, + "learning_rate": 4.088139723801787e-06, + "loss": 1.6698, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.773566484451294, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6824, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.8707821369171143, + "eval_runtime": 46.7886, + "eval_samples_per_second": 21.373, + "eval_steps_per_second": 0.898, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6008529663085938, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7352, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.502838134765625, + "learning_rate": 4.08204711616572e-06, + "loss": 1.6342, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 3.081169605255127, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7307, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.841200828552246, + "learning_rate": 4.077985377741674e-06, + "loss": 1.7401, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.5010390281677246, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7252, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.468873977661133, + "learning_rate": 4.073923639317629e-06, + "loss": 1.6627, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.829737663269043, + "learning_rate": 4.071892770105605e-06, + "loss": 1.6419, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.8112833499908447, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6629, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6793017387390137, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6667, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.327993869781494, + "learning_rate": 4.065800162469538e-06, + "loss": 1.71, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.874659538269043, + "eval_runtime": 49.0634, + "eval_samples_per_second": 20.382, + "eval_steps_per_second": 0.856, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.689023971557617, + "learning_rate": 4.063769293257514e-06, + "loss": 1.6881, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.612842082977295, + "learning_rate": 4.061738424045492e-06, + "loss": 1.6971, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.319627046585083, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7469, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.588916540145874, + "learning_rate": 4.057676685621446e-06, + "loss": 1.707, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.270096778869629, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6284, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.4404375553131104, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7007, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5532679557800293, + "learning_rate": 4.051584077985378e-06, + "loss": 1.72, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.5105185508728027, + "learning_rate": 4.049553208773355e-06, + "loss": 1.6772, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.577929973602295, + "learning_rate": 4.047522339561333e-06, + "loss": 1.6747, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.8271372318267822, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7951, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.8684895038604736, + "eval_runtime": 48.0288, + "eval_samples_per_second": 20.821, + "eval_steps_per_second": 0.874, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.3306479454040527, + "learning_rate": 4.043460601137287e-06, + "loss": 1.6791, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.5543673038482666, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6761, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.3913605213165283, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7579, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.3837356567382812, + "learning_rate": 4.037367993501219e-06, + "loss": 1.6636, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.6936402320861816, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6488, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.6400210857391357, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.6926, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.574533462524414, + "learning_rate": 4.031275385865151e-06, + "loss": 1.7066, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5130677223205566, + "learning_rate": 4.029244516653128e-06, + "loss": 1.6431, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.578166961669922, + "learning_rate": 4.027213647441105e-06, + "loss": 1.648, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.4736695289611816, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6822, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.8754831552505493, + "eval_runtime": 50.0971, + "eval_samples_per_second": 19.961, + "eval_steps_per_second": 0.838, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.544315814971924, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6853, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8765335083007812, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7587, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.4749417304992676, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.6536, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.493164300918579, + "learning_rate": 4.017059301380992e-06, + "loss": 1.7436, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.6274492740631104, + "learning_rate": 4.015028432168969e-06, + "loss": 1.632, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.4162979125976562, + "learning_rate": 4.012997562956945e-06, + "loss": 1.6663, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5683343410491943, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6771, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.3710429668426514, + "learning_rate": 4.0089358245329e-06, + "loss": 1.6644, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.5356292724609375, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7339, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.480884075164795, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.716, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.867702841758728, + "eval_runtime": 48.0038, + "eval_samples_per_second": 20.832, + "eval_steps_per_second": 0.875, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4768476486206055, + "learning_rate": 4.002843216896833e-06, + "loss": 1.6667, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5727357864379883, + "learning_rate": 4.000812347684809e-06, + "loss": 1.6786, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6243903636932373, + "learning_rate": 3.998781478472787e-06, + "loss": 1.6328, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.712955951690674, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7852, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.756272554397583, + "learning_rate": 3.994719740048742e-06, + "loss": 1.7006, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.919201135635376, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6624, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.471566915512085, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7122, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.46273136138916, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6621, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.5338189601898193, + "learning_rate": 3.98659626320065e-06, + "loss": 1.6869, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.3850855827331543, + "learning_rate": 3.984565393988627e-06, + "loss": 1.6925, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.8650325536727905, + "eval_runtime": 49.4167, + "eval_samples_per_second": 20.236, + "eval_steps_per_second": 0.85, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.654726028442383, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7113, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.5065038204193115, + "learning_rate": 3.980503655564582e-06, + "loss": 1.7078, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.51668381690979, + "learning_rate": 3.978472786352559e-06, + "loss": 1.6742, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4586126804351807, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7339, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.452857732772827, + "learning_rate": 3.974411047928514e-06, + "loss": 1.6993, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.608631134033203, + "learning_rate": 3.972380178716491e-06, + "loss": 1.7047, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.34022855758667, + "learning_rate": 3.970349309504468e-06, + "loss": 1.6423, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.329612970352173, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.6443, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.3301169872283936, + "learning_rate": 3.966287571080423e-06, + "loss": 1.6937, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.3010976314544678, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6896, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.863387942314148, + "eval_runtime": 47.6761, + "eval_samples_per_second": 20.975, + "eval_steps_per_second": 0.881, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.296764612197876, + "learning_rate": 3.962225832656377e-06, + "loss": 1.6098, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.437114953994751, + "learning_rate": 3.960194963444355e-06, + "loss": 1.6727, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4908952713012695, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6816, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.81123685836792, + "learning_rate": 3.956133225020309e-06, + "loss": 1.6795, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.7088773250579834, + "learning_rate": 3.954102355808286e-06, + "loss": 1.6838, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.5021824836730957, + "learning_rate": 3.952071486596264e-06, + "loss": 1.671, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.3021023273468018, + "learning_rate": 3.950040617384241e-06, + "loss": 1.651, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.866119861602783, + "learning_rate": 3.948009748172218e-06, + "loss": 1.7117, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.3471503257751465, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.6883, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4612059593200684, + "learning_rate": 3.943948009748173e-06, + "loss": 1.6898, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.8587063550949097, + "eval_runtime": 43.4462, + "eval_samples_per_second": 23.017, + "eval_steps_per_second": 0.967, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.376680374145508, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7253, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.4701383113861084, + "learning_rate": 3.939886271324127e-06, + "loss": 1.6306, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.574384927749634, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6287, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.8596720695495605, + "learning_rate": 3.935824532900082e-06, + "loss": 1.7031, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.7144267559051514, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7219, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.5807480812072754, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.6364, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 3.3094675540924072, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6526, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.390503168106079, + "learning_rate": 3.927701056051991e-06, + "loss": 1.6795, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.5525588989257812, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.6432, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.522307872772217, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6218, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.8604834079742432, + "eval_runtime": 46.6465, + "eval_samples_per_second": 21.438, + "eval_steps_per_second": 0.9, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.438201904296875, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6788, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.639164686203003, + "learning_rate": 3.9195775792039e-06, + "loss": 1.6552, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.608038902282715, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.69, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.4357974529266357, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7031, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.4964048862457275, + "learning_rate": 3.913484971567831e-06, + "loss": 1.6126, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.5115931034088135, + "learning_rate": 3.911454102355808e-06, + "loss": 1.684, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.54923152923584, + "learning_rate": 3.909423233143786e-06, + "loss": 1.6815, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.849602222442627, + "learning_rate": 3.907392363931763e-06, + "loss": 1.744, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.7209551334381104, + "learning_rate": 3.90536149471974e-06, + "loss": 1.755, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.602299451828003, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.7835, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.8590067625045776, + "eval_runtime": 49.5182, + "eval_samples_per_second": 20.195, + "eval_steps_per_second": 0.848, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.504567861557007, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6595, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.456721305847168, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6692, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.319553852081299, + "learning_rate": 3.897238017871649e-06, + "loss": 1.6664, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.45987606048584, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.65, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 3.0049448013305664, + "learning_rate": 3.893176279447604e-06, + "loss": 1.7366, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.522472858428955, + "learning_rate": 3.891145410235581e-06, + "loss": 1.7022, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.3914318084716797, + "learning_rate": 3.889114541023558e-06, + "loss": 1.6591, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.462167501449585, + "learning_rate": 3.887083671811536e-06, + "loss": 1.6158, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.6707093715667725, + "learning_rate": 3.885052802599513e-06, + "loss": 1.7453, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.493745803833008, + "learning_rate": 3.88302193338749e-06, + "loss": 1.6836, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.8536967039108276, + "eval_runtime": 50.3456, + "eval_samples_per_second": 19.863, + "eval_steps_per_second": 0.834, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.5338919162750244, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.65, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.6954877376556396, + "learning_rate": 3.878960194963445e-06, + "loss": 1.7156, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.4878089427948, + "learning_rate": 3.876929325751422e-06, + "loss": 1.6053, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.4329843521118164, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6642, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 3.118771553039551, + "learning_rate": 3.872867587327377e-06, + "loss": 1.6392, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4644217491149902, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6566, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.4643497467041016, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6654, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.7067277431488037, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.682, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.8748488426208496, + "learning_rate": 3.864744110479286e-06, + "loss": 1.6886, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.632155179977417, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7069, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8592487573623657, + "eval_runtime": 47.0773, + "eval_samples_per_second": 21.242, + "eval_steps_per_second": 0.892, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.5379488468170166, + "learning_rate": 3.86068237205524e-06, + "loss": 1.6695, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.571742057800293, + "learning_rate": 3.858651502843217e-06, + "loss": 1.6235, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.5355312824249268, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6528, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.715510606765747, + "learning_rate": 3.854589764419171e-06, + "loss": 1.7379, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.635364532470703, + "learning_rate": 3.852558895207149e-06, + "loss": 1.6958, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.4713191986083984, + "learning_rate": 3.850528025995126e-06, + "loss": 1.6471, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.584200382232666, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6548, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.4639663696289062, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6614, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.4823923110961914, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.6523, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.953594923019409, + "learning_rate": 3.842404549147035e-06, + "loss": 1.6281, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8549022674560547, + "eval_runtime": 47.2215, + "eval_samples_per_second": 21.177, + "eval_steps_per_second": 0.889, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.524352550506592, + "learning_rate": 3.840373679935013e-06, + "loss": 1.6741, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.411515235900879, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.6571, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.6827003955841064, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6927, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 3.3486077785491943, + "learning_rate": 3.834281072298944e-06, + "loss": 1.6347, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.4518957138061523, + "learning_rate": 3.832250203086921e-06, + "loss": 1.7011, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.623842239379883, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7459, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.689602851867676, + "learning_rate": 3.828188464662876e-06, + "loss": 1.6812, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.8425960540771484, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6258, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.6639413833618164, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6769, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.4269535541534424, + "learning_rate": 3.822095857026808e-06, + "loss": 1.6363, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8489640951156616, + "eval_runtime": 45.897, + "eval_samples_per_second": 21.788, + "eval_steps_per_second": 0.915, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.3430261611938477, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6563, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.322094678878784, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6588, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.6172726154327393, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.681, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.574084997177124, + "learning_rate": 3.813972380178717e-06, + "loss": 1.6977, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.437417984008789, + "learning_rate": 3.811941510966694e-06, + "loss": 1.6204, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.319276809692383, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5858, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5540425777435303, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6622, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.567634344100952, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.635, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4815967082977295, + "learning_rate": 3.803818034118603e-06, + "loss": 1.7146, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.475162982940674, + "learning_rate": 3.80178716490658e-06, + "loss": 1.6911, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.848276138305664, + "eval_runtime": 44.1011, + "eval_samples_per_second": 22.675, + "eval_steps_per_second": 0.952, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5385165214538574, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.6523, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.443129301071167, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6239, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.696885347366333, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.7318, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.8570339679718018, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7198, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.475338935852051, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7052, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.5250487327575684, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.7015, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.812121868133545, + "learning_rate": 3.787571080422421e-06, + "loss": 1.6876, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.486117362976074, + "learning_rate": 3.785540211210398e-06, + "loss": 1.6695, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.5146725177764893, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.6585, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.5808019638061523, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.7608, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8411076068878174, + "eval_runtime": 48.0798, + "eval_samples_per_second": 20.799, + "eval_steps_per_second": 0.874, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.509993076324463, + "learning_rate": 3.77944760357433e-06, + "loss": 1.7101, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.492285966873169, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6417, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.809333086013794, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6963, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.524418354034424, + "learning_rate": 3.773354995938262e-06, + "loss": 1.6957, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.5848705768585205, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7193, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.8617639541625977, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.6721, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.4655680656433105, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.586, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 3.0633299350738525, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6792, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.3882107734680176, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.6229, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.5860483646392822, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6753, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8393676280975342, + "eval_runtime": 44.3838, + "eval_samples_per_second": 22.531, + "eval_steps_per_second": 0.946, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.930290460586548, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.6955, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.8278939723968506, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.6455, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.3822925090789795, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.7163, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.6904139518737793, + "learning_rate": 3.753046303818034e-06, + "loss": 1.6942, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.535284996032715, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7237, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.669031858444214, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.6391, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.5162742137908936, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6663, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.5937418937683105, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7115, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.4892170429229736, + "learning_rate": 3.742891957757921e-06, + "loss": 1.6727, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.7901015281677246, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7179, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8420112133026123, + "eval_runtime": 42.0904, + "eval_samples_per_second": 23.758, + "eval_steps_per_second": 0.998, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.510568618774414, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.6408, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.5084173679351807, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6495, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.3693647384643555, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.6064, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.520414352416992, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5668, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.544624090194702, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.6231, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.6257777214050293, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6093, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.0384273529052734, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.7291, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.718588352203369, + "learning_rate": 3.724614134849716e-06, + "loss": 1.604, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.655592918395996, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6993, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.569904088973999, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.6595, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8419173955917358, + "eval_runtime": 42.5535, + "eval_samples_per_second": 23.5, + "eval_steps_per_second": 0.987, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.643686056137085, + "learning_rate": 3.718521527213648e-06, + "loss": 1.6483, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.7431893348693848, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5418, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 3.0191054344177246, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.6425, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.695777654647827, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.651, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.512838840484619, + "learning_rate": 3.710398050365557e-06, + "loss": 1.6181, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.9957621097564697, + "learning_rate": 3.708367181153534e-06, + "loss": 1.6454, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.5258326530456543, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.5995, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.6610865592956543, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5816, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.825767993927002, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6402, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.605870008468628, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6282, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8433754444122314, + "eval_runtime": 45.2127, + "eval_samples_per_second": 22.118, + "eval_steps_per_second": 0.929, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.7334296703338623, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5958, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.9007229804992676, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.7128, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8827497959136963, + "learning_rate": 3.694151096669375e-06, + "loss": 1.6157, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6470541954040527, + "learning_rate": 3.692120227457352e-06, + "loss": 1.643, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.5725178718566895, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.5481, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.7779407501220703, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.6241, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 3.161653995513916, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.644, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.5995497703552246, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6475, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.607166290283203, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6573, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.4532339572906494, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.6044, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.8344279527664185, + "eval_runtime": 43.2125, + "eval_samples_per_second": 23.141, + "eval_steps_per_second": 0.972, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.540632486343384, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6308, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.40977144241333, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6134, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.6345880031585693, + "learning_rate": 3.673842404549147e-06, + "loss": 1.615, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.565157413482666, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6351, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.7868876457214355, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.7186, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.7061610221862793, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.6454, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.587785482406616, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6867, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.4618277549743652, + "learning_rate": 3.663688058489034e-06, + "loss": 1.6189, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.4425477981567383, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5772, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.7514469623565674, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6368, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8366374969482422, + "eval_runtime": 44.4679, + "eval_samples_per_second": 22.488, + "eval_steps_per_second": 0.945, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.529805898666382, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.6229, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6875312328338623, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6445, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.715970993041992, + "learning_rate": 3.65353371242892e-06, + "loss": 1.5318, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.8685660362243652, + "learning_rate": 3.651502843216897e-06, + "loss": 1.6465, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.6801602840423584, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.5959, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.61037015914917, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6561, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.478137493133545, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6435, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5708508491516113, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6206, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8857522010803223, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6269, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.6102395057678223, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.648, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.836978793144226, + "eval_runtime": 44.9452, + "eval_samples_per_second": 22.249, + "eval_steps_per_second": 0.934, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7924158573150635, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6802, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.784693479537964, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.6515, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.747894525527954, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6093, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.635850191116333, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6791, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.5760908126831055, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.5721, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5258584022521973, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5644, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.919623613357544, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.6258, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.8330483436584473, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6077, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8145744800567627, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6463, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.9187726974487305, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6348, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.837236762046814, + "eval_runtime": 45.4421, + "eval_samples_per_second": 22.006, + "eval_steps_per_second": 0.924, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.6741979122161865, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6441, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.556701898574829, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5523, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.6352901458740234, + "learning_rate": 3.612916328188465e-06, + "loss": 1.6545, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.5784106254577637, + "learning_rate": 3.610885458976442e-06, + "loss": 1.6649, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.7136101722717285, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.596, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.608877182006836, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.5923, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.783069372177124, + "learning_rate": 3.604792851340374e-06, + "loss": 1.6685, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.621882438659668, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6404, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.672128915786743, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5713, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.645395040512085, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.609, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8348923921585083, + "eval_runtime": 44.8379, + "eval_samples_per_second": 22.303, + "eval_steps_per_second": 0.937, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 3.610335111618042, + "learning_rate": 3.596669374492283e-06, + "loss": 1.6693, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.7391626834869385, + "learning_rate": 3.59463850528026e-06, + "loss": 1.5847, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.658418655395508, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6552, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.6508545875549316, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6201, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 4.237835884094238, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.6725, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.672144889831543, + "learning_rate": 3.586515028432169e-06, + "loss": 1.6666, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.619480609893799, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5671, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 3.150592565536499, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6118, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.946467161178589, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6499, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.6372971534729004, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5516, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8352437019348145, + "eval_runtime": 46.1704, + "eval_samples_per_second": 21.659, + "eval_steps_per_second": 0.91, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.6211962699890137, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6414, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.567579746246338, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.6632, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.4393808841705322, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6323, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.7005021572113037, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6016, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7400808334350586, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6327, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.922241449356079, + "learning_rate": 3.566206336311942e-06, + "loss": 1.5866, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.524975299835205, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5707, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.637805461883545, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6259, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7430527210235596, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.595, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.5548908710479736, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.625, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8332648277282715, + "eval_runtime": 48.5515, + "eval_samples_per_second": 20.597, + "eval_steps_per_second": 0.865, + "step": 1000 + } + ], + "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": 3.11730940936192e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1000/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-1300/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1300/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1300/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/latest new file mode 100644 index 0000000000000000000000000000000000000000..d71c4c79bfb8e74842ee005fbfc27bedb903c39b --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/latest @@ -0,0 +1 @@ +global_step1300 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1300/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1300/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1300/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..9a5ea0a8f530c7cadae96f990054dbe39716ad46 --- /dev/null +++ b/babylm_shuffle_local3_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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.7659518718719482, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.279, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.433785915374756, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.3368, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 3.1486687660217285, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3896, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.391956090927124, + "learning_rate": 1.624087591240876e-06, + "loss": 2.297, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.6649584770202637, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.2415, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.5604891777038574, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2688, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.47900652885437, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2933, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.632723569869995, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.2143, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.8085947036743164, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2748, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.3927979469299316, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2685, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.355207920074463, + "eval_runtime": 42.3505, + "eval_samples_per_second": 23.612, + "eval_steps_per_second": 0.992, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.6170592308044434, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.3121, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.699814796447754, + "learning_rate": 1.77007299270073e-06, + "loss": 2.2853, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.3450610637664795, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3297, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.458219528198242, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2639, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.7101449966430664, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2134, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.5734996795654297, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2347, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 2.5200352668762207, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2299, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5022480487823486, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2253, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.190117359161377, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.2272, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.414524555206299, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2424, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.317634344100952, + "eval_runtime": 43.5742, + "eval_samples_per_second": 22.949, + "eval_steps_per_second": 0.964, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.3226804733276367, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1449, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.733006477355957, + "learning_rate": 1.952554744525548e-06, + "loss": 2.2027, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 2.5586233139038086, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2007, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.669914722442627, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2024, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.6148102283477783, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2496, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 2.331369638442993, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1814, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.3051886558532715, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2015, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.6890218257904053, + "learning_rate": 2.062043795620438e-06, + "loss": 2.184, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.4975883960723877, + "learning_rate": 2.08029197080292e-06, + "loss": 2.2439, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.5848891735076904, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1413, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.280966281890869, + "eval_runtime": 43.1001, + "eval_samples_per_second": 23.202, + "eval_steps_per_second": 0.974, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.2104, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.3356540203094482, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2312, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 2.3616602420806885, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.1399, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.581623077392578, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1456, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.547288656234741, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.184, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6386160850524902, + "learning_rate": 2.208029197080292e-06, + "loss": 2.1658, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.593065023422241, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1443, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.530632972717285, + "learning_rate": 2.244525547445256e-06, + "loss": 2.194, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.3553175926208496, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.1918, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 2.7536532878875732, + "learning_rate": 2.281021897810219e-06, + "loss": 2.1477, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2608063220977783, + "eval_runtime": 42.7938, + "eval_samples_per_second": 23.368, + "eval_steps_per_second": 0.981, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.160094738006592, + "learning_rate": 2.299270072992701e-06, + "loss": 2.133, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.4474735260009766, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1468, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.6654434204101562, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.1194, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.726611375808716, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1518, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 2.3560168743133545, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1537, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.664065361022949, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0888, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.3946099281311035, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.0998, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.7812652587890625, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1231, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.64595890045166, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1746, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.788524866104126, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0731, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.2340502738952637, + "eval_runtime": 42.3104, + "eval_samples_per_second": 23.635, + "eval_steps_per_second": 0.993, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 2.9249846935272217, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0564, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.5507798194885254, + "learning_rate": 2.5e-06, + "loss": 2.1714, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.1604716777801514, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0701, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.5429840087890625, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.0662, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8296737670898438, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.1378, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.9396755695343018, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1069, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.736814022064209, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1076, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 2.774909496307373, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1286, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.0861318111419678, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1499, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.624016284942627, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0989, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.215052604675293, + "eval_runtime": 43.9956, + "eval_samples_per_second": 22.73, + "eval_steps_per_second": 0.955, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.801500082015991, + "learning_rate": 2.664233576642336e-06, + "loss": 2.0795, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 2.2107114791870117, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.1035, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.030381441116333, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.0794, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 2.9719064235687256, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1496, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.688839912414551, + "learning_rate": 2.737226277372263e-06, + "loss": 2.1152, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 2.9350979328155518, + "learning_rate": 2.755474452554745e-06, + "loss": 2.0773, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.8958046436309814, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.1805, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.0781922340393066, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.1249, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.032059907913208, + "learning_rate": 2.81021897810219e-06, + "loss": 2.0632, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.798758029937744, + "learning_rate": 2.828467153284672e-06, + "loss": 2.1001, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2005105018615723, + "eval_runtime": 43.5111, + "eval_samples_per_second": 22.983, + "eval_steps_per_second": 0.965, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.7674927711486816, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.0211, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 2.6168575286865234, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0599, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 2.933353900909424, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.0169, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.81402587890625, + "learning_rate": 2.901459854014599e-06, + "loss": 2.122, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.5592572689056396, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.0957, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.738271951675415, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.0868, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.5362703800201416, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1237, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.6304945945739746, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.0825, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.040717601776123, + "learning_rate": 2.992700729927008e-06, + "loss": 2.0388, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.6226463317871094, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0163, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.186411142349243, + "eval_runtime": 44.625, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.5307068824768066, + "learning_rate": 3.029197080291971e-06, + "loss": 2.1097, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 2.9160444736480713, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.0946, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.5250024795532227, + "learning_rate": 3.065693430656935e-06, + "loss": 2.0709, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.6963064670562744, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.0937, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 2.845707416534424, + "learning_rate": 3.102189781021898e-06, + "loss": 2.0851, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.8985090255737305, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.0176, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 2.734377861022949, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.9931, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.5125555992126465, + "learning_rate": 3.156934306569343e-06, + "loss": 2.1133, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.0393502712249756, + "learning_rate": 3.175182481751825e-06, + "loss": 2.0636, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.6410164833068848, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0304, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.171868085861206, + "eval_runtime": 44.2249, + "eval_samples_per_second": 22.612, + "eval_steps_per_second": 0.95, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.7937498092651367, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0518, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7801201343536377, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.026, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7963638305664062, + "learning_rate": 3.248175182481752e-06, + "loss": 1.9642, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6330482959747314, + "learning_rate": 3.266423357664234e-06, + "loss": 2.0545, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.9771625995635986, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0445, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.6112730503082275, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0616, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.732865333557129, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0403, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.9113688468933105, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0382, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.4516773223876953, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.0416, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.781735897064209, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0348, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.160499334335327, + "eval_runtime": 44.5652, + "eval_samples_per_second": 22.439, + "eval_steps_per_second": 0.942, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.9990546703338623, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0324, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.783071279525757, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0518, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.659040689468384, + "learning_rate": 3.43065693430657e-06, + "loss": 1.9722, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 2.799798011779785, + "learning_rate": 3.4489051094890513e-06, + "loss": 2.0362, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 3.025458812713623, + "learning_rate": 3.467153284671533e-06, + "loss": 1.9863, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 2.686415195465088, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0201, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.716064929962158, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0249, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.637653350830078, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0403, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 2.7686848640441895, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0626, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.721956729888916, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.1077, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.152338981628418, + "eval_runtime": 43.5455, + "eval_samples_per_second": 22.964, + "eval_steps_per_second": 0.965, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.92571759223938, + "learning_rate": 3.576642335766424e-06, + "loss": 2.0928, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.9189202785491943, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9315, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 2.990233898162842, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9427, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.702348232269287, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9768, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6949572563171387, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.9954, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.108405113220215, + "learning_rate": 3.667883211678833e-06, + "loss": 1.9631, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.738922119140625, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9644, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.7419424057006836, + "learning_rate": 3.704379562043796e-06, + "loss": 2.0675, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9853813648223877, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0245, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.4614126682281494, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.9665, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.1342060565948486, + "eval_runtime": 46.3954, + "eval_samples_per_second": 21.554, + "eval_steps_per_second": 0.905, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.6787819862365723, + "learning_rate": 3.7591240875912412e-06, + "loss": 2.0063, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 2.563525438308716, + "learning_rate": 3.777372262773723e-06, + "loss": 1.998, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.4981701374053955, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.9738, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 2.5357978343963623, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0221, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.5365488529205322, + "learning_rate": 3.832116788321168e-06, + "loss": 1.9664, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 2.669468641281128, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9577, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 2.6950953006744385, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9786, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8249378204345703, + "learning_rate": 3.886861313868613e-06, + "loss": 2.0041, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.5924317836761475, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0338, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.1282334327697754, + "learning_rate": 3.923357664233577e-06, + "loss": 1.9698, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.1221420764923096, + "eval_runtime": 47.1127, + "eval_samples_per_second": 21.226, + "eval_steps_per_second": 0.891, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.7878193855285645, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0446, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.7137255668640137, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0575, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.286421060562134, + "learning_rate": 3.978102189781022e-06, + "loss": 1.9131, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.5466175079345703, + "learning_rate": 3.996350364963504e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.792483329772949, + "learning_rate": 4.014598540145986e-06, + "loss": 1.912, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.7533867359161377, + "learning_rate": 4.032846715328467e-06, + "loss": 1.9742, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 2.848405361175537, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9815, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.1034414768218994, + "learning_rate": 4.069343065693431e-06, + "loss": 1.9409, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.5109779834747314, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9676, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.486506938934326, + "learning_rate": 4.105839416058395e-06, + "loss": 2.0353, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1102187633514404, + "eval_runtime": 45.384, + "eval_samples_per_second": 22.034, + "eval_steps_per_second": 0.925, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.419269323348999, + "learning_rate": 4.124087591240876e-06, + "loss": 1.9622, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.4967262744903564, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.9523, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 2.7311460971832275, + "learning_rate": 4.16058394160584e-06, + "loss": 2.0234, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.241004467010498, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.907, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.821537733078003, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9761, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 2.5711426734924316, + "learning_rate": 4.215328467153285e-06, + "loss": 1.9411, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.936594009399414, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9432, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.6091108322143555, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9545, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.522674798965454, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9246, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.984649181365967, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9417, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1031947135925293, + "eval_runtime": 47.5409, + "eval_samples_per_second": 21.035, + "eval_steps_per_second": 0.883, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.8250324726104736, + "learning_rate": 4.306569343065693e-06, + "loss": 1.9061, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.7654290199279785, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9653, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.7091264724731445, + "learning_rate": 4.343065693430658e-06, + "loss": 1.9482, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 2.8019583225250244, + "learning_rate": 4.361313868613139e-06, + "loss": 1.9332, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.147368907928467, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9836, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.7050259113311768, + "learning_rate": 4.397810218978102e-06, + "loss": 1.9607, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.738743543624878, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9374, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.122777223587036, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9249, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.850745439529419, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9707, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.391998291015625, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.9703, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.0887911319732666, + "eval_runtime": 48.5149, + "eval_samples_per_second": 20.612, + "eval_steps_per_second": 0.866, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 2.9131698608398438, + "learning_rate": 4.489051094890512e-06, + "loss": 1.9989, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.1045937538146973, + "learning_rate": 4.507299270072993e-06, + "loss": 1.9718, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.7984158992767334, + "learning_rate": 4.525547445255475e-06, + "loss": 1.947, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.6223504543304443, + "learning_rate": 4.543795620437956e-06, + "loss": 1.8942, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.8221523761749268, + "learning_rate": 4.562043795620438e-06, + "loss": 2.0045, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.6657445430755615, + "learning_rate": 4.580291970802921e-06, + "loss": 2.0053, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 2.793455123901367, + "learning_rate": 4.598540145985402e-06, + "loss": 1.9577, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.551201820373535, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9256, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 2.739194631576538, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9265, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.2917332649230957, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9664, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.086106777191162, + "eval_runtime": 48.995, + "eval_samples_per_second": 20.41, + "eval_steps_per_second": 0.857, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 2.6489369869232178, + "learning_rate": 4.671532846715329e-06, + "loss": 1.9629, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 2.60017466545105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.932, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.569115400314331, + "learning_rate": 4.708029197080292e-06, + "loss": 1.9902, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 2.7225773334503174, + "learning_rate": 4.726277372262774e-06, + "loss": 1.9619, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.5064749717712402, + "learning_rate": 4.744525547445255e-06, + "loss": 1.8921, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.612931966781616, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9584, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.2524304389953613, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9675, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.8852503299713135, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9107, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 2.654776096343994, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9465, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.679086685180664, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9468, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.0842835903167725, + "eval_runtime": 51.7803, + "eval_samples_per_second": 19.312, + "eval_steps_per_second": 0.811, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.5110855102539062, + "learning_rate": 4.854014598540146e-06, + "loss": 1.88, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.1786117553710938, + "learning_rate": 4.872262773722628e-06, + "loss": 1.8994, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.388328790664673, + "learning_rate": 4.89051094890511e-06, + "loss": 1.8914, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 2.784834384918213, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.923, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.470426559448242, + "learning_rate": 4.927007299270074e-06, + "loss": 1.944, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.690580129623413, + "learning_rate": 4.945255474452555e-06, + "loss": 1.8978, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.869737386703491, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9964, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.9965591430664062, + "learning_rate": 4.981751824817518e-06, + "loss": 1.9706, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.6665658950805664, + "learning_rate": 5e-06, + "loss": 1.9313, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.9231722354888916, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9504, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.068364143371582, + "eval_runtime": 47.8334, + "eval_samples_per_second": 20.906, + "eval_steps_per_second": 0.878, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.6094462871551514, + "learning_rate": 4.995938261575955e-06, + "loss": 1.862, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.780848741531372, + "learning_rate": 4.993907392363932e-06, + "loss": 1.95, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.948662757873535, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.878, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 2.9622998237609863, + "learning_rate": 4.989845653939887e-06, + "loss": 1.8954, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 2.6401216983795166, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9021, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.029592275619507, + "learning_rate": 4.985783915515841e-06, + "loss": 1.8967, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.828392505645752, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9134, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.613926649093628, + "learning_rate": 4.981722177091796e-06, + "loss": 1.8829, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.6173207759857178, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9225, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.4657015800476074, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9142, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.059906482696533, + "eval_runtime": 49.6267, + "eval_samples_per_second": 20.15, + "eval_steps_per_second": 0.846, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7448673248291016, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9545, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.639096736907959, + "learning_rate": 4.973598700243705e-06, + "loss": 1.8664, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 2.6445884704589844, + "learning_rate": 4.971567831031681e-06, + "loss": 1.888, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.9130005836486816, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7717, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.48093318939209, + "learning_rate": 4.967506092607636e-06, + "loss": 1.8538, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.8918230533599854, + "learning_rate": 4.965475223395614e-06, + "loss": 1.886, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.511747121810913, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.8633, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.860029697418213, + "learning_rate": 4.961413484971569e-06, + "loss": 1.9063, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7304794788360596, + "learning_rate": 4.959382615759545e-06, + "loss": 1.8872, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.6569266319274902, + "learning_rate": 4.957351746547523e-06, + "loss": 1.8721, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0610575675964355, + "eval_runtime": 43.9292, + "eval_samples_per_second": 22.764, + "eval_steps_per_second": 0.956, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.1482994556427, + "learning_rate": 4.9553208773355e-06, + "loss": 2.0123, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.59568190574646, + "learning_rate": 4.953290008123478e-06, + "loss": 1.902, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.5897581577301025, + "learning_rate": 4.951259138911454e-06, + "loss": 1.8406, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.734032154083252, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.9605, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.8088371753692627, + "learning_rate": 4.947197400487409e-06, + "loss": 1.896, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.9678707122802734, + "learning_rate": 4.945166531275386e-06, + "loss": 1.8625, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.261200189590454, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8884, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 2.5811619758605957, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9311, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.6783649921417236, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9324, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.6533944606781006, + "learning_rate": 4.937043054427295e-06, + "loss": 1.8575, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.049668312072754, + "eval_runtime": 47.7347, + "eval_samples_per_second": 20.949, + "eval_steps_per_second": 0.88, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 3.0882792472839355, + "learning_rate": 4.935012185215272e-06, + "loss": 1.84, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.8259081840515137, + "learning_rate": 4.93298131600325e-06, + "loss": 1.9109, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.540908098220825, + "learning_rate": 4.930950446791227e-06, + "loss": 1.8906, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.9277396202087402, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9047, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.392449140548706, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.876, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 3.1890463829040527, + "learning_rate": 4.924857839155159e-06, + "loss": 1.8709, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.725710391998291, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8953, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.6287014484405518, + "learning_rate": 4.920796100731113e-06, + "loss": 1.8204, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.7316651344299316, + "learning_rate": 4.918765231519091e-06, + "loss": 1.882, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.5561037063598633, + "learning_rate": 4.916734362307068e-06, + "loss": 1.8331, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0412285327911377, + "eval_runtime": 50.237, + "eval_samples_per_second": 19.906, + "eval_steps_per_second": 0.836, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.974501132965088, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8414, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.4530839920043945, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8323, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.789569616317749, + "learning_rate": 4.910641754671e-06, + "loss": 1.8697, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.582704544067383, + "learning_rate": 4.908610885458977e-06, + "loss": 1.8592, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 2.776442289352417, + "learning_rate": 4.906580016246954e-06, + "loss": 1.912, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.670191764831543, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.8899, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.4534454345703125, + "learning_rate": 4.902518277822909e-06, + "loss": 1.8878, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.007469654083252, + "learning_rate": 4.900487408610886e-06, + "loss": 1.8977, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 2.856466293334961, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9062, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.817441463470459, + "learning_rate": 4.89642567018684e-06, + "loss": 1.9686, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.0344440937042236, + "eval_runtime": 47.2472, + "eval_samples_per_second": 21.165, + "eval_steps_per_second": 0.889, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.9732158184051514, + "learning_rate": 4.894394800974818e-06, + "loss": 1.8332, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 3.0413055419921875, + "learning_rate": 4.892363931762794e-06, + "loss": 1.9171, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.646749973297119, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.8403, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.1274876594543457, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8218, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.8122124671936035, + "learning_rate": 4.886271324126727e-06, + "loss": 1.9272, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.4608492851257324, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.8974, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.928083658218384, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.8646, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.6308164596557617, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8219, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7311816215515137, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8326, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.640162467956543, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8013, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0261101722717285, + "eval_runtime": 47.0605, + "eval_samples_per_second": 21.249, + "eval_steps_per_second": 0.892, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.8542518615722656, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8233, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.464303970336914, + "learning_rate": 4.872055239642567e-06, + "loss": 1.787, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8535103797912598, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8535, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.5060505867004395, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8733, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.6549222469329834, + "learning_rate": 4.865962632006499e-06, + "loss": 1.8047, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.8636224269866943, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8543, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.644256830215454, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8853, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.949490547180176, + "learning_rate": 4.859870024370431e-06, + "loss": 1.8951, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.4576282501220703, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9069, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.6140542030334473, + "learning_rate": 4.855808285946385e-06, + "loss": 1.8573, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0159800052642822, + "eval_runtime": 44.2447, + "eval_samples_per_second": 22.602, + "eval_steps_per_second": 0.949, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.525766134262085, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.844, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.5479896068573, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8572, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.628570795059204, + "learning_rate": 4.849715678310317e-06, + "loss": 1.865, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.2768666744232178, + "learning_rate": 4.847684809098294e-06, + "loss": 1.8355, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.635516881942749, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8603, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.5307435989379883, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8482, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.3688952922821045, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8894, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.6000115871429443, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.8045, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7782504558563232, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8283, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.6914777755737305, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8765, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0086047649383545, + "eval_runtime": 48.245, + "eval_samples_per_second": 20.728, + "eval_steps_per_second": 0.871, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.598365306854248, + "learning_rate": 4.833468724614135e-06, + "loss": 1.8461, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.6718530654907227, + "learning_rate": 4.831437855402113e-06, + "loss": 1.7453, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.393528699874878, + "learning_rate": 4.82940698619009e-06, + "loss": 1.8275, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 3.0622947216033936, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8658, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.431041955947876, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8601, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.939424514770508, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8649, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.8116796016693115, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8303, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.511960029602051, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8383, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 3.8372581005096436, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8111, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.511948347091675, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8448, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.9998281002044678, + "eval_runtime": 44.7624, + "eval_samples_per_second": 22.34, + "eval_steps_per_second": 0.938, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 3.5780487060546875, + "learning_rate": 4.813160032493907e-06, + "loss": 1.8161, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.81740403175354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.8633, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.8117666244506836, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8385, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.5791451930999756, + "learning_rate": 4.80706742485784e-06, + "loss": 1.896, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6816465854644775, + "learning_rate": 4.805036555645816e-06, + "loss": 1.8126, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.6419460773468018, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8753, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.666086196899414, + "learning_rate": 4.800974817221771e-06, + "loss": 1.8305, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8626585006713867, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8463, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.351825475692749, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.838, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.300717830657959, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8711, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.9904799461364746, + "eval_runtime": 43.9511, + "eval_samples_per_second": 22.753, + "eval_steps_per_second": 0.956, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.43884015083313, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8145, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.50593900680542, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8103, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.4643657207489014, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8162, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.4615302085876465, + "learning_rate": 4.786758732737612e-06, + "loss": 1.8637, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.6566433906555176, + "learning_rate": 4.784727863525589e-06, + "loss": 1.7805, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.5973432064056396, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8327, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.4996957778930664, + "learning_rate": 4.780666125101544e-06, + "loss": 1.7761, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 3.0266876220703125, + "learning_rate": 4.778635255889521e-06, + "loss": 1.8036, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.522369146347046, + "learning_rate": 4.776604386677498e-06, + "loss": 1.814, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.482727527618408, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8665, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.9888195991516113, + "eval_runtime": 44.0755, + "eval_samples_per_second": 22.688, + "eval_steps_per_second": 0.953, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 3.237544059753418, + "learning_rate": 4.772542648253453e-06, + "loss": 1.8642, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.393286943435669, + "learning_rate": 4.77051177904143e-06, + "loss": 1.7857, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.66072154045105, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8423, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.6725080013275146, + "learning_rate": 4.766450040617385e-06, + "loss": 1.815, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.7657666206359863, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8729, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.3153769969940186, + "learning_rate": 4.762388302193339e-06, + "loss": 1.8349, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5804553031921387, + "learning_rate": 4.760357432981316e-06, + "loss": 1.8155, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.7280807495117188, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8262, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.551940441131592, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8332, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.6137285232543945, + "learning_rate": 4.754264825345248e-06, + "loss": 1.7805, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.983920693397522, + "eval_runtime": 42.525, + "eval_samples_per_second": 23.516, + "eval_steps_per_second": 0.988, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.4460153579711914, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.7953, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.556032657623291, + "learning_rate": 4.750203086921203e-06, + "loss": 1.7393, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.335983991622925, + "learning_rate": 4.74817221770918e-06, + "loss": 1.8494, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.5057711601257324, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8663, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4447145462036133, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7532, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.769984483718872, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7755, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.472193956375122, + "learning_rate": 4.740048740861089e-06, + "loss": 1.8056, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.4286932945251465, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8094, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.0248074531555176, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8497, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.607700824737549, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8741, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.9779797792434692, + "eval_runtime": 43.7734, + "eval_samples_per_second": 22.845, + "eval_steps_per_second": 0.959, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 3.035045623779297, + "learning_rate": 4.731925264012998e-06, + "loss": 1.8438, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.786921977996826, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7666, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.4158425331115723, + "learning_rate": 4.727863525588953e-06, + "loss": 1.7464, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7772, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.8447, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 3.08266282081604, + "learning_rate": 4.723801787164907e-06, + "loss": 1.7812, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.791749954223633, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8047, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6457693576812744, + "learning_rate": 4.719740048740861e-06, + "loss": 1.836, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.69480037689209, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8145, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6356561183929443, + "learning_rate": 4.715678310316816e-06, + "loss": 1.7756, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.9723366498947144, + "eval_runtime": 45.5806, + "eval_samples_per_second": 21.939, + "eval_steps_per_second": 0.921, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.53474497795105, + "learning_rate": 4.713647441104794e-06, + "loss": 1.774, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.3631107807159424, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8517, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.594604730606079, + "learning_rate": 4.709585702680748e-06, + "loss": 1.8962, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.5280940532684326, + "learning_rate": 4.707554833468725e-06, + "loss": 1.7759, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.4596288204193115, + "learning_rate": 4.705523964256703e-06, + "loss": 1.7948, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.9123644828796387, + "learning_rate": 4.703493095044679e-06, + "loss": 1.7399, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.4354264736175537, + "learning_rate": 4.701462225832657e-06, + "loss": 1.8659, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.28625226020813, + "learning_rate": 4.699431356620634e-06, + "loss": 1.831, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.5445950031280518, + "learning_rate": 4.697400487408612e-06, + "loss": 1.8121, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.767825126647949, + "learning_rate": 4.695369618196588e-06, + "loss": 1.7874, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9653255939483643, + "eval_runtime": 45.2453, + "eval_samples_per_second": 22.102, + "eval_steps_per_second": 0.928, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4052765369415283, + "learning_rate": 4.693338748984566e-06, + "loss": 1.7539, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.498098373413086, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8343, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.821753740310669, + "learning_rate": 4.68927701056052e-06, + "loss": 1.8119, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.4158449172973633, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.7982, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.5229434967041016, + "learning_rate": 4.685215272136475e-06, + "loss": 1.806, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.729199171066284, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8768, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.5102570056915283, + "learning_rate": 4.681153533712429e-06, + "loss": 1.775, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4463515281677246, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7493, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.505733013153076, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8462, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.5784912109375, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7381, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9640673398971558, + "eval_runtime": 44.2714, + "eval_samples_per_second": 22.588, + "eval_steps_per_second": 0.949, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.7864291667938232, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7612, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.507809638977051, + "learning_rate": 4.670999187652316e-06, + "loss": 1.757, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.5499937534332275, + "learning_rate": 4.668968318440293e-06, + "loss": 1.8243, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.435472011566162, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7269, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.541031837463379, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.7959, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5052099227905273, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7191, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6285109519958496, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8423, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.7825605869293213, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7178, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.677319288253784, + "learning_rate": 4.656783103168157e-06, + "loss": 1.7939, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.6830129623413086, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7842, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.960136890411377, + "eval_runtime": 45.9379, + "eval_samples_per_second": 21.769, + "eval_steps_per_second": 0.914, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.5368666648864746, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8352, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.8922412395477295, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.8252, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.8332598209381104, + "learning_rate": 4.648659626320066e-06, + "loss": 1.7805, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.420780658721924, + "learning_rate": 4.646628757108043e-06, + "loss": 1.7499, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.6684062480926514, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.7857, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.8547494411468506, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.8115, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.59279203414917, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8254, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.80961275100708, + "learning_rate": 4.638505280259952e-06, + "loss": 1.7901, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.9017655849456787, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8265, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.595454216003418, + "learning_rate": 4.634443541835907e-06, + "loss": 1.7833, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9576054811477661, + "eval_runtime": 50.7265, + "eval_samples_per_second": 19.714, + "eval_steps_per_second": 0.828, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 4.260303020477295, + "learning_rate": 4.632412672623883e-06, + "loss": 1.6741, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.366147756576538, + "learning_rate": 4.630381803411861e-06, + "loss": 1.7778, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.894336462020874, + "learning_rate": 4.628350934199838e-06, + "loss": 1.8268, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.7675349712371826, + "learning_rate": 4.626320064987816e-06, + "loss": 1.7748, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.461413621902466, + "learning_rate": 4.624289195775792e-06, + "loss": 1.7599, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.670187473297119, + "learning_rate": 4.62225832656377e-06, + "loss": 1.7736, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.8119454383850098, + "learning_rate": 4.620227457351747e-06, + "loss": 1.7837, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.8020036220550537, + "learning_rate": 4.618196588139724e-06, + "loss": 1.7508, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.5977094173431396, + "learning_rate": 4.616165718927701e-06, + "loss": 1.834, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.731919765472412, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.8137, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.9465162754058838, + "eval_runtime": 48.9829, + "eval_samples_per_second": 20.415, + "eval_steps_per_second": 0.857, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.604386568069458, + "learning_rate": 4.612103980503656e-06, + "loss": 1.7854, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.601107120513916, + "learning_rate": 4.610073111291633e-06, + "loss": 1.7749, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6758370399475098, + "learning_rate": 4.60804224207961e-06, + "loss": 1.7826, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.631549119949341, + "learning_rate": 4.606011372867588e-06, + "loss": 1.7723, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.6704764366149902, + "learning_rate": 4.603980503655565e-06, + "loss": 1.7088, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 3.078573703765869, + "learning_rate": 4.601949634443542e-06, + "loss": 1.7254, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7492356300354004, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7756, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 3.0763585567474365, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7806, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.5379223823547363, + "learning_rate": 4.595857026807474e-06, + "loss": 1.7516, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8343496322631836, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8111, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.945749282836914, + "eval_runtime": 46.5453, + "eval_samples_per_second": 21.484, + "eval_steps_per_second": 0.902, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.809753179550171, + "learning_rate": 4.591795288383429e-06, + "loss": 1.7877, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.4196107387542725, + "learning_rate": 4.589764419171406e-06, + "loss": 1.7457, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.409287452697754, + "learning_rate": 4.587733549959383e-06, + "loss": 1.7338, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.414706230163574, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7185, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.591365337371826, + "learning_rate": 4.583671811535338e-06, + "loss": 1.8083, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.4648489952087402, + "learning_rate": 4.581640942323315e-06, + "loss": 1.8192, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.3804972171783447, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8077, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.283082962036133, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8294, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.61795711517334, + "learning_rate": 4.575548334687247e-06, + "loss": 1.7322, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.4035353660583496, + "learning_rate": 4.573517465475224e-06, + "loss": 1.7798, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9422866106033325, + "eval_runtime": 45.3068, + "eval_samples_per_second": 22.072, + "eval_steps_per_second": 0.927, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.7253756523132324, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7876, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.738295078277588, + "learning_rate": 4.569455727051178e-06, + "loss": 1.7933, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 3.1418697834014893, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8145, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.941603660583496, + "learning_rate": 4.565393988627132e-06, + "loss": 1.7671, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.063924551010132, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.7809, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 4.957335472106934, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7306, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.651716709136963, + "learning_rate": 4.559301380991065e-06, + "loss": 1.7428, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 4.777034282684326, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.8118, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.5549471378326416, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7294, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.474388599395752, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8338, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.9402854442596436, + "eval_runtime": 43.166, + "eval_samples_per_second": 23.166, + "eval_steps_per_second": 0.973, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.8968348503112793, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7432, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.6996726989746094, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.7837, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.4599392414093018, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7388, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.745487928390503, + "learning_rate": 4.545085296506905e-06, + "loss": 1.7439, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.6960670948028564, + "learning_rate": 4.543054427294882e-06, + "loss": 1.737, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.596876621246338, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7708, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.598267078399658, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7541, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.6069958209991455, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7756, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.490694761276245, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.7372, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.44179368019104, + "learning_rate": 4.532900081234769e-06, + "loss": 1.7652, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.934715986251831, + "eval_runtime": 44.688, + "eval_samples_per_second": 22.377, + "eval_steps_per_second": 0.94, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 3.018480062484741, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7596, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.5316643714904785, + "learning_rate": 4.528838342810723e-06, + "loss": 1.7715, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4893109798431396, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7549, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.524834156036377, + "learning_rate": 4.524776604386678e-06, + "loss": 1.8046, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7527239322662354, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7583, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.5210752487182617, + "learning_rate": 4.520714865962632e-06, + "loss": 1.7896, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.5787770748138428, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7355, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.537195920944214, + "learning_rate": 4.516653127538587e-06, + "loss": 1.7272, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.6810014247894287, + "learning_rate": 4.514622258326564e-06, + "loss": 1.753, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.5868756771087646, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7964, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.934499740600586, + "eval_runtime": 45.6556, + "eval_samples_per_second": 21.903, + "eval_steps_per_second": 0.92, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.6830856800079346, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7072, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.6574594974517822, + "learning_rate": 4.508529650690496e-06, + "loss": 1.7004, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.4415769577026367, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7667, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.825115442276001, + "learning_rate": 4.504467912266451e-06, + "loss": 1.8356, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.949010133743286, + "learning_rate": 4.502437043054428e-06, + "loss": 1.7579, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.690986394882202, + "learning_rate": 4.500406173842405e-06, + "loss": 1.7657, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 3.1015520095825195, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.7432, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.608424186706543, + "learning_rate": 4.49634443541836e-06, + "loss": 1.8112, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.5238773822784424, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7724, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.7712161540985107, + "learning_rate": 4.492282696994314e-06, + "loss": 1.7111, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9326083660125732, + "eval_runtime": 48.5359, + "eval_samples_per_second": 20.603, + "eval_steps_per_second": 0.865, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.4763033390045166, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7268, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.6310081481933594, + "learning_rate": 4.488220958570269e-06, + "loss": 1.6885, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.5950350761413574, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7775, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.769026517868042, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7737, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.4898502826690674, + "learning_rate": 4.4821283509342e-06, + "loss": 1.751, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.668081521987915, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7178, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.441615104675293, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7511, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.727665901184082, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.8639, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.7964634895324707, + "learning_rate": 4.474004874086109e-06, + "loss": 1.6945, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.6142687797546387, + "learning_rate": 4.471974004874087e-06, + "loss": 1.8028, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9281476736068726, + "eval_runtime": 48.078, + "eval_samples_per_second": 20.8, + "eval_steps_per_second": 0.874, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.6655778884887695, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.7632, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.931162118911743, + "learning_rate": 4.467912266450041e-06, + "loss": 1.7244, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.469895124435425, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7773, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.4948830604553223, + "learning_rate": 4.463850528025995e-06, + "loss": 1.788, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5961945056915283, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.756, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.452946424484253, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6968, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.9352009296417236, + "learning_rate": 4.457757920389927e-06, + "loss": 1.7414, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.7044763565063477, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7848, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.7200324535369873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7699, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.6185197830200195, + "learning_rate": 4.451665312753859e-06, + "loss": 1.7816, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9209110736846924, + "eval_runtime": 47.3321, + "eval_samples_per_second": 21.127, + "eval_steps_per_second": 0.887, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.3639564514160156, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7164, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7073519229888916, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.727, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.637080430984497, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7951, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.8013885021209717, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7333, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.512484312057495, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8048, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.3851466178894043, + "learning_rate": 4.439480097481723e-06, + "loss": 1.6667, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.8105392456054688, + "learning_rate": 4.4374492282697e-06, + "loss": 1.7622, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.555773973464966, + "learning_rate": 4.435418359057677e-06, + "loss": 1.7519, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.688596725463867, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7198, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.828646421432495, + "learning_rate": 4.431356620633632e-06, + "loss": 1.7478, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.919256567955017, + "eval_runtime": 50.5993, + "eval_samples_per_second": 19.763, + "eval_steps_per_second": 0.83, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.4523651599884033, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7344, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.4237422943115234, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6664, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.2746193408966064, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.7446, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.502819061279297, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7691, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.7032017707824707, + "learning_rate": 4.421202274573518e-06, + "loss": 1.6448, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.5879440307617188, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6908, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 3.2481982707977295, + "learning_rate": 4.417140536149473e-06, + "loss": 1.7725, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.4844820499420166, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7141, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.5596253871917725, + "learning_rate": 4.413078797725427e-06, + "loss": 1.7485, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 3.142839193344116, + "learning_rate": 4.411047928513404e-06, + "loss": 1.734, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9129430055618286, + "eval_runtime": 49.2608, + "eval_samples_per_second": 20.3, + "eval_steps_per_second": 0.853, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.5837552547454834, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7668, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.7043633460998535, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7497, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.9886765480041504, + "learning_rate": 4.404955320877336e-06, + "loss": 1.7082, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.3199501037597656, + "learning_rate": 4.402924451665313e-06, + "loss": 1.693, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.606682777404785, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7242, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.6967978477478027, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7516, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.351726531982422, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7338, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.590869188308716, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7462, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.3498220443725586, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7609, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.281002998352051, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7584, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.908841609954834, + "eval_runtime": 50.7771, + "eval_samples_per_second": 19.694, + "eval_steps_per_second": 0.827, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.468755006790161, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7654, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.3467307090759277, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7923, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.3723301887512207, + "learning_rate": 4.384646628757108e-06, + "loss": 1.74, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.4527792930603027, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6988, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.3923704624176025, + "learning_rate": 4.380584890333063e-06, + "loss": 1.6406, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.740556240081787, + "learning_rate": 4.37855402112104e-06, + "loss": 1.8032, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.3013832569122314, + "learning_rate": 4.376523151909017e-06, + "loss": 1.7259, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.4632391929626465, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.7462, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.4878177642822266, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7665, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.610898494720459, + "learning_rate": 4.370430544272949e-06, + "loss": 1.7646, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9056707620620728, + "eval_runtime": 51.8738, + "eval_samples_per_second": 19.278, + "eval_steps_per_second": 0.81, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.387721300125122, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7454, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.9966042041778564, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6789, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.664689064025879, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7486, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.4788103103637695, + "learning_rate": 4.362307067424858e-06, + "loss": 1.6981, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.4413328170776367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7279, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.6818156242370605, + "learning_rate": 4.358245329000813e-06, + "loss": 1.74, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.650348424911499, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7212, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.442805767059326, + "learning_rate": 4.354183590576767e-06, + "loss": 1.6951, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.6806585788726807, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7499, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.586792230606079, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7709, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.901674747467041, + "eval_runtime": 42.3186, + "eval_samples_per_second": 23.63, + "eval_steps_per_second": 0.992, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.547452688217163, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7458, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.5820348262786865, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.7232, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.4208321571350098, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7981, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.8791956901550293, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7372, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6537868976593018, + "learning_rate": 4.339967506092607e-06, + "loss": 1.7216, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.416097402572632, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.701, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.5565483570098877, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7558, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.5875823497772217, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7439, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5459702014923096, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.7058, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.4195566177368164, + "learning_rate": 4.329813160032495e-06, + "loss": 1.6749, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9035066366195679, + "eval_runtime": 43.7962, + "eval_samples_per_second": 22.833, + "eval_steps_per_second": 0.959, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6019818782806396, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7549, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.8627779483795166, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7257, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.582158088684082, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7138, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.4528613090515137, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7162, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 3.1956026554107666, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7556, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.3531901836395264, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.746, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.441988706588745, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7664, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.599506378173828, + "learning_rate": 4.313566206336312e-06, + "loss": 1.6955, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.4121932983398438, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7194, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.4803290367126465, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7386, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8993914127349854, + "eval_runtime": 43.0525, + "eval_samples_per_second": 23.227, + "eval_steps_per_second": 0.976, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.701594114303589, + "learning_rate": 4.307473598700244e-06, + "loss": 1.6796, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7144737243652344, + "learning_rate": 4.305442729488221e-06, + "loss": 1.763, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.395312786102295, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7522, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.3926732540130615, + "learning_rate": 4.301380991064176e-06, + "loss": 1.7705, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.6293702125549316, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7693, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.3746306896209717, + "learning_rate": 4.29731925264013e-06, + "loss": 1.72, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.5421926975250244, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6868, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.6991186141967773, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7378, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.463897705078125, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7628, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5139036178588867, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7125, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8909698724746704, + "eval_runtime": 42.5917, + "eval_samples_per_second": 23.479, + "eval_steps_per_second": 0.986, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.5220465660095215, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6975, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.432988405227661, + "learning_rate": 4.285134037367994e-06, + "loss": 1.6904, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.5894253253936768, + "learning_rate": 4.283103168155971e-06, + "loss": 1.6919, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5334458351135254, + "learning_rate": 4.281072298943948e-06, + "loss": 1.6968, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.510019302368164, + "learning_rate": 4.279041429731926e-06, + "loss": 1.6791, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.531538248062134, + "learning_rate": 4.277010560519903e-06, + "loss": 1.723, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.5621440410614014, + "learning_rate": 4.27497969130788e-06, + "loss": 1.749, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.594329833984375, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7579, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.321031093597412, + "learning_rate": 4.270917952883835e-06, + "loss": 1.6983, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.6325764656066895, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7872, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.886696219444275, + "eval_runtime": 43.4143, + "eval_samples_per_second": 23.034, + "eval_steps_per_second": 0.967, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5342679023742676, + "learning_rate": 4.266856214459789e-06, + "loss": 1.6655, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.5810720920562744, + "learning_rate": 4.264825345247766e-06, + "loss": 1.7182, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.7768406867980957, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6534, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.5680463314056396, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7243, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7162973880767822, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.7146, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.697925090789795, + "learning_rate": 4.256701868399675e-06, + "loss": 1.7296, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.4458425045013428, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7466, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6330654621124268, + "learning_rate": 4.252640129975629e-06, + "loss": 1.6949, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.350430727005005, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6574, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.505995512008667, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7111, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.8887447118759155, + "eval_runtime": 43.9304, + "eval_samples_per_second": 22.763, + "eval_steps_per_second": 0.956, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 3.036954879760742, + "learning_rate": 4.246547522339562e-06, + "loss": 1.7182, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.563606023788452, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.6566, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.516315460205078, + "learning_rate": 4.242485783915517e-06, + "loss": 1.6543, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.6289877891540527, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6996, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.7563092708587646, + "learning_rate": 4.23842404549147e-06, + "loss": 1.666, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5743000507354736, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7338, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 3.1975982189178467, + "learning_rate": 4.234362307067425e-06, + "loss": 1.7444, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.6022894382476807, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7367, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.4466192722320557, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.726, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.579015016555786, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7146, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.8904188871383667, + "eval_runtime": 43.8091, + "eval_samples_per_second": 22.826, + "eval_steps_per_second": 0.959, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.719646453857422, + "learning_rate": 4.226238830219334e-06, + "loss": 1.7209, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.4423880577087402, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7288, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5883753299713135, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7089, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.447049140930176, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7095, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.453739881515503, + "learning_rate": 4.218115353371243e-06, + "loss": 1.6747, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4540836811065674, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7578, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.5468671321868896, + "learning_rate": 4.214053614947198e-06, + "loss": 1.6556, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.443814277648926, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7307, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.4998748302459717, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6768, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.667025566101074, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7493, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.8830010890960693, + "eval_runtime": 44.7027, + "eval_samples_per_second": 22.37, + "eval_steps_per_second": 0.94, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.416248321533203, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7381, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.378025770187378, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7335, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.633173704147339, + "learning_rate": 4.201868399675061e-06, + "loss": 1.6484, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.549384117126465, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7143, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.6492598056793213, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6907, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.444086790084839, + "learning_rate": 4.195775792038993e-06, + "loss": 1.636, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.628814935684204, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7342, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.83831524848938, + "learning_rate": 4.191714053614948e-06, + "loss": 1.7028, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.467339515686035, + "learning_rate": 4.189683184402924e-06, + "loss": 1.749, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.4529521465301514, + "learning_rate": 4.187652315190902e-06, + "loss": 1.682, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.8826342821121216, + "eval_runtime": 44.7757, + "eval_samples_per_second": 22.334, + "eval_steps_per_second": 0.938, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7340762615203857, + "learning_rate": 4.185621445978879e-06, + "loss": 1.723, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.6028647422790527, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7566, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.3967695236206055, + "learning_rate": 4.181559707554833e-06, + "loss": 1.683, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5475900173187256, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6976, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.5411484241485596, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7489, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 3.458749294281006, + "learning_rate": 4.175467099918766e-06, + "loss": 1.6828, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.2977802753448486, + "learning_rate": 4.173436230706742e-06, + "loss": 1.6963, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.875307559967041, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.6857, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 4.097696304321289, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7225, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.6774847507476807, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7031, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.8783056735992432, + "eval_runtime": 44.0219, + "eval_samples_per_second": 22.716, + "eval_steps_per_second": 0.954, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.6787376403808594, + "learning_rate": 4.165312753858652e-06, + "loss": 1.6865, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.3803224563598633, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7237, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.9285683631896973, + "learning_rate": 4.161251015434607e-06, + "loss": 1.6575, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.5534448623657227, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6799, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4105820655822754, + "learning_rate": 4.157189277010561e-06, + "loss": 1.7128, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.6571805477142334, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7123, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.891935348510742, + "learning_rate": 4.153127538586516e-06, + "loss": 1.6949, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.8692519664764404, + "learning_rate": 4.151096669374492e-06, + "loss": 1.699, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.780078649520874, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7114, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.4797518253326416, + "learning_rate": 4.147034930950447e-06, + "loss": 1.6943, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.8726727962493896, + "eval_runtime": 43.5664, + "eval_samples_per_second": 22.953, + "eval_steps_per_second": 0.964, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.7010464668273926, + "learning_rate": 4.145004061738425e-06, + "loss": 1.6207, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8812713623046875, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.6608, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.6700279712677, + "learning_rate": 4.140942323314379e-06, + "loss": 1.677, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.796513557434082, + "learning_rate": 4.138911454102356e-06, + "loss": 1.7291, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.4274256229400635, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7058, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5528790950775146, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7064, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.4288926124572754, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7163, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.4939630031585693, + "learning_rate": 4.130787977254265e-06, + "loss": 1.721, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7909305095672607, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7777, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.3382463455200195, + "learning_rate": 4.12672623883022e-06, + "loss": 1.6975, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.8738009929656982, + "eval_runtime": 44.8142, + "eval_samples_per_second": 22.314, + "eval_steps_per_second": 0.937, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6876718997955322, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6851, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.450212240219116, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7209, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.4365246295928955, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.6545, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.3236255645751953, + "learning_rate": 4.118602761982129e-06, + "loss": 1.6434, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.4753000736236572, + "learning_rate": 4.116571892770106e-06, + "loss": 1.7471, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.84885835647583, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6777, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.4361016750335693, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7079, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.5868160724639893, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7288, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.6499862670898438, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7191, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.333289384841919, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5938, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.8746954202651978, + "eval_runtime": 46.8039, + "eval_samples_per_second": 21.366, + "eval_steps_per_second": 0.897, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.1299779415130615, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7585, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 3.6410014629364014, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6803, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.801745653152466, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6957, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 3.0350706577301025, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7031, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 3.5306341648101807, + "learning_rate": 4.096263200649879e-06, + "loss": 1.6598, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.591485023498535, + "learning_rate": 4.094232331437856e-06, + "loss": 1.6659, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 3.0916528701782227, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6951, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.459009885787964, + "learning_rate": 4.090170593013811e-06, + "loss": 1.6597, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.932072162628174, + "learning_rate": 4.088139723801787e-06, + "loss": 1.6698, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.773566484451294, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6824, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.8707821369171143, + "eval_runtime": 46.7886, + "eval_samples_per_second": 21.373, + "eval_steps_per_second": 0.898, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6008529663085938, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7352, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.502838134765625, + "learning_rate": 4.08204711616572e-06, + "loss": 1.6342, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 3.081169605255127, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7307, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.841200828552246, + "learning_rate": 4.077985377741674e-06, + "loss": 1.7401, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.5010390281677246, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7252, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.468873977661133, + "learning_rate": 4.073923639317629e-06, + "loss": 1.6627, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.829737663269043, + "learning_rate": 4.071892770105605e-06, + "loss": 1.6419, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.8112833499908447, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6629, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6793017387390137, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6667, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.327993869781494, + "learning_rate": 4.065800162469538e-06, + "loss": 1.71, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.874659538269043, + "eval_runtime": 49.0634, + "eval_samples_per_second": 20.382, + "eval_steps_per_second": 0.856, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.689023971557617, + "learning_rate": 4.063769293257514e-06, + "loss": 1.6881, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.612842082977295, + "learning_rate": 4.061738424045492e-06, + "loss": 1.6971, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.319627046585083, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7469, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.588916540145874, + "learning_rate": 4.057676685621446e-06, + "loss": 1.707, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.270096778869629, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6284, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.4404375553131104, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7007, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5532679557800293, + "learning_rate": 4.051584077985378e-06, + "loss": 1.72, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.5105185508728027, + "learning_rate": 4.049553208773355e-06, + "loss": 1.6772, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.577929973602295, + "learning_rate": 4.047522339561333e-06, + "loss": 1.6747, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.8271372318267822, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7951, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.8684895038604736, + "eval_runtime": 48.0288, + "eval_samples_per_second": 20.821, + "eval_steps_per_second": 0.874, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.3306479454040527, + "learning_rate": 4.043460601137287e-06, + "loss": 1.6791, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.5543673038482666, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6761, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.3913605213165283, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7579, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.3837356567382812, + "learning_rate": 4.037367993501219e-06, + "loss": 1.6636, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.6936402320861816, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6488, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.6400210857391357, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.6926, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.574533462524414, + "learning_rate": 4.031275385865151e-06, + "loss": 1.7066, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5130677223205566, + "learning_rate": 4.029244516653128e-06, + "loss": 1.6431, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.578166961669922, + "learning_rate": 4.027213647441105e-06, + "loss": 1.648, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.4736695289611816, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6822, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.8754831552505493, + "eval_runtime": 50.0971, + "eval_samples_per_second": 19.961, + "eval_steps_per_second": 0.838, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.544315814971924, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6853, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8765335083007812, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7587, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.4749417304992676, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.6536, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.493164300918579, + "learning_rate": 4.017059301380992e-06, + "loss": 1.7436, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.6274492740631104, + "learning_rate": 4.015028432168969e-06, + "loss": 1.632, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.4162979125976562, + "learning_rate": 4.012997562956945e-06, + "loss": 1.6663, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5683343410491943, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6771, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.3710429668426514, + "learning_rate": 4.0089358245329e-06, + "loss": 1.6644, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.5356292724609375, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7339, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.480884075164795, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.716, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.867702841758728, + "eval_runtime": 48.0038, + "eval_samples_per_second": 20.832, + "eval_steps_per_second": 0.875, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4768476486206055, + "learning_rate": 4.002843216896833e-06, + "loss": 1.6667, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5727357864379883, + "learning_rate": 4.000812347684809e-06, + "loss": 1.6786, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6243903636932373, + "learning_rate": 3.998781478472787e-06, + "loss": 1.6328, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.712955951690674, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7852, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.756272554397583, + "learning_rate": 3.994719740048742e-06, + "loss": 1.7006, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.919201135635376, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6624, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.471566915512085, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7122, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.46273136138916, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6621, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.5338189601898193, + "learning_rate": 3.98659626320065e-06, + "loss": 1.6869, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.3850855827331543, + "learning_rate": 3.984565393988627e-06, + "loss": 1.6925, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.8650325536727905, + "eval_runtime": 49.4167, + "eval_samples_per_second": 20.236, + "eval_steps_per_second": 0.85, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.654726028442383, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7113, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.5065038204193115, + "learning_rate": 3.980503655564582e-06, + "loss": 1.7078, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.51668381690979, + "learning_rate": 3.978472786352559e-06, + "loss": 1.6742, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4586126804351807, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7339, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.452857732772827, + "learning_rate": 3.974411047928514e-06, + "loss": 1.6993, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.608631134033203, + "learning_rate": 3.972380178716491e-06, + "loss": 1.7047, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.34022855758667, + "learning_rate": 3.970349309504468e-06, + "loss": 1.6423, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.329612970352173, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.6443, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.3301169872283936, + "learning_rate": 3.966287571080423e-06, + "loss": 1.6937, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.3010976314544678, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6896, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.863387942314148, + "eval_runtime": 47.6761, + "eval_samples_per_second": 20.975, + "eval_steps_per_second": 0.881, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.296764612197876, + "learning_rate": 3.962225832656377e-06, + "loss": 1.6098, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.437114953994751, + "learning_rate": 3.960194963444355e-06, + "loss": 1.6727, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4908952713012695, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6816, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.81123685836792, + "learning_rate": 3.956133225020309e-06, + "loss": 1.6795, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.7088773250579834, + "learning_rate": 3.954102355808286e-06, + "loss": 1.6838, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.5021824836730957, + "learning_rate": 3.952071486596264e-06, + "loss": 1.671, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.3021023273468018, + "learning_rate": 3.950040617384241e-06, + "loss": 1.651, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.866119861602783, + "learning_rate": 3.948009748172218e-06, + "loss": 1.7117, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.3471503257751465, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.6883, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4612059593200684, + "learning_rate": 3.943948009748173e-06, + "loss": 1.6898, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.8587063550949097, + "eval_runtime": 43.4462, + "eval_samples_per_second": 23.017, + "eval_steps_per_second": 0.967, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.376680374145508, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7253, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.4701383113861084, + "learning_rate": 3.939886271324127e-06, + "loss": 1.6306, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.574384927749634, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6287, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.8596720695495605, + "learning_rate": 3.935824532900082e-06, + "loss": 1.7031, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.7144267559051514, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7219, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.5807480812072754, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.6364, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 3.3094675540924072, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6526, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.390503168106079, + "learning_rate": 3.927701056051991e-06, + "loss": 1.6795, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.5525588989257812, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.6432, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.522307872772217, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6218, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.8604834079742432, + "eval_runtime": 46.6465, + "eval_samples_per_second": 21.438, + "eval_steps_per_second": 0.9, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.438201904296875, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6788, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.639164686203003, + "learning_rate": 3.9195775792039e-06, + "loss": 1.6552, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.608038902282715, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.69, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.4357974529266357, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7031, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.4964048862457275, + "learning_rate": 3.913484971567831e-06, + "loss": 1.6126, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.5115931034088135, + "learning_rate": 3.911454102355808e-06, + "loss": 1.684, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.54923152923584, + "learning_rate": 3.909423233143786e-06, + "loss": 1.6815, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.849602222442627, + "learning_rate": 3.907392363931763e-06, + "loss": 1.744, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.7209551334381104, + "learning_rate": 3.90536149471974e-06, + "loss": 1.755, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.602299451828003, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.7835, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.8590067625045776, + "eval_runtime": 49.5182, + "eval_samples_per_second": 20.195, + "eval_steps_per_second": 0.848, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.504567861557007, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6595, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.456721305847168, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6692, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.319553852081299, + "learning_rate": 3.897238017871649e-06, + "loss": 1.6664, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.45987606048584, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.65, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 3.0049448013305664, + "learning_rate": 3.893176279447604e-06, + "loss": 1.7366, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.522472858428955, + "learning_rate": 3.891145410235581e-06, + "loss": 1.7022, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.3914318084716797, + "learning_rate": 3.889114541023558e-06, + "loss": 1.6591, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.462167501449585, + "learning_rate": 3.887083671811536e-06, + "loss": 1.6158, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.6707093715667725, + "learning_rate": 3.885052802599513e-06, + "loss": 1.7453, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.493745803833008, + "learning_rate": 3.88302193338749e-06, + "loss": 1.6836, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.8536967039108276, + "eval_runtime": 50.3456, + "eval_samples_per_second": 19.863, + "eval_steps_per_second": 0.834, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.5338919162750244, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.65, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.6954877376556396, + "learning_rate": 3.878960194963445e-06, + "loss": 1.7156, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.4878089427948, + "learning_rate": 3.876929325751422e-06, + "loss": 1.6053, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.4329843521118164, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6642, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 3.118771553039551, + "learning_rate": 3.872867587327377e-06, + "loss": 1.6392, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4644217491149902, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6566, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.4643497467041016, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6654, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.7067277431488037, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.682, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.8748488426208496, + "learning_rate": 3.864744110479286e-06, + "loss": 1.6886, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.632155179977417, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7069, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8592487573623657, + "eval_runtime": 47.0773, + "eval_samples_per_second": 21.242, + "eval_steps_per_second": 0.892, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.5379488468170166, + "learning_rate": 3.86068237205524e-06, + "loss": 1.6695, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.571742057800293, + "learning_rate": 3.858651502843217e-06, + "loss": 1.6235, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.5355312824249268, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6528, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.715510606765747, + "learning_rate": 3.854589764419171e-06, + "loss": 1.7379, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.635364532470703, + "learning_rate": 3.852558895207149e-06, + "loss": 1.6958, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.4713191986083984, + "learning_rate": 3.850528025995126e-06, + "loss": 1.6471, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.584200382232666, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6548, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.4639663696289062, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6614, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.4823923110961914, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.6523, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.953594923019409, + "learning_rate": 3.842404549147035e-06, + "loss": 1.6281, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8549022674560547, + "eval_runtime": 47.2215, + "eval_samples_per_second": 21.177, + "eval_steps_per_second": 0.889, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.524352550506592, + "learning_rate": 3.840373679935013e-06, + "loss": 1.6741, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.411515235900879, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.6571, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.6827003955841064, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6927, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 3.3486077785491943, + "learning_rate": 3.834281072298944e-06, + "loss": 1.6347, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.4518957138061523, + "learning_rate": 3.832250203086921e-06, + "loss": 1.7011, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.623842239379883, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7459, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.689602851867676, + "learning_rate": 3.828188464662876e-06, + "loss": 1.6812, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.8425960540771484, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6258, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.6639413833618164, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6769, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.4269535541534424, + "learning_rate": 3.822095857026808e-06, + "loss": 1.6363, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8489640951156616, + "eval_runtime": 45.897, + "eval_samples_per_second": 21.788, + "eval_steps_per_second": 0.915, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.3430261611938477, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6563, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.322094678878784, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6588, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.6172726154327393, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.681, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.574084997177124, + "learning_rate": 3.813972380178717e-06, + "loss": 1.6977, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.437417984008789, + "learning_rate": 3.811941510966694e-06, + "loss": 1.6204, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.319276809692383, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5858, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5540425777435303, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6622, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.567634344100952, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.635, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4815967082977295, + "learning_rate": 3.803818034118603e-06, + "loss": 1.7146, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.475162982940674, + "learning_rate": 3.80178716490658e-06, + "loss": 1.6911, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.848276138305664, + "eval_runtime": 44.1011, + "eval_samples_per_second": 22.675, + "eval_steps_per_second": 0.952, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5385165214538574, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.6523, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.443129301071167, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6239, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.696885347366333, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.7318, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.8570339679718018, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7198, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.475338935852051, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7052, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.5250487327575684, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.7015, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.812121868133545, + "learning_rate": 3.787571080422421e-06, + "loss": 1.6876, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.486117362976074, + "learning_rate": 3.785540211210398e-06, + "loss": 1.6695, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.5146725177764893, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.6585, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.5808019638061523, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.7608, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8411076068878174, + "eval_runtime": 48.0798, + "eval_samples_per_second": 20.799, + "eval_steps_per_second": 0.874, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.509993076324463, + "learning_rate": 3.77944760357433e-06, + "loss": 1.7101, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.492285966873169, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6417, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.809333086013794, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6963, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.524418354034424, + "learning_rate": 3.773354995938262e-06, + "loss": 1.6957, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.5848705768585205, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7193, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.8617639541625977, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.6721, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.4655680656433105, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.586, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 3.0633299350738525, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6792, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.3882107734680176, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.6229, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.5860483646392822, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6753, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8393676280975342, + "eval_runtime": 44.3838, + "eval_samples_per_second": 22.531, + "eval_steps_per_second": 0.946, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.930290460586548, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.6955, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.8278939723968506, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.6455, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.3822925090789795, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.7163, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.6904139518737793, + "learning_rate": 3.753046303818034e-06, + "loss": 1.6942, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.535284996032715, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7237, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.669031858444214, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.6391, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.5162742137908936, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6663, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.5937418937683105, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7115, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.4892170429229736, + "learning_rate": 3.742891957757921e-06, + "loss": 1.6727, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.7901015281677246, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7179, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8420112133026123, + "eval_runtime": 42.0904, + "eval_samples_per_second": 23.758, + "eval_steps_per_second": 0.998, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.510568618774414, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.6408, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.5084173679351807, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6495, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.3693647384643555, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.6064, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.520414352416992, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5668, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.544624090194702, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.6231, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.6257777214050293, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6093, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.0384273529052734, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.7291, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.718588352203369, + "learning_rate": 3.724614134849716e-06, + "loss": 1.604, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.655592918395996, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6993, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.569904088973999, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.6595, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8419173955917358, + "eval_runtime": 42.5535, + "eval_samples_per_second": 23.5, + "eval_steps_per_second": 0.987, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.643686056137085, + "learning_rate": 3.718521527213648e-06, + "loss": 1.6483, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.7431893348693848, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5418, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 3.0191054344177246, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.6425, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.695777654647827, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.651, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.512838840484619, + "learning_rate": 3.710398050365557e-06, + "loss": 1.6181, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.9957621097564697, + "learning_rate": 3.708367181153534e-06, + "loss": 1.6454, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.5258326530456543, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.5995, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.6610865592956543, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5816, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.825767993927002, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6402, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.605870008468628, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6282, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8433754444122314, + "eval_runtime": 45.2127, + "eval_samples_per_second": 22.118, + "eval_steps_per_second": 0.929, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.7334296703338623, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5958, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.9007229804992676, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.7128, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8827497959136963, + "learning_rate": 3.694151096669375e-06, + "loss": 1.6157, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6470541954040527, + "learning_rate": 3.692120227457352e-06, + "loss": 1.643, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.5725178718566895, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.5481, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.7779407501220703, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.6241, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 3.161653995513916, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.644, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.5995497703552246, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6475, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.607166290283203, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6573, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.4532339572906494, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.6044, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.8344279527664185, + "eval_runtime": 43.2125, + "eval_samples_per_second": 23.141, + "eval_steps_per_second": 0.972, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.540632486343384, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6308, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.40977144241333, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6134, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.6345880031585693, + "learning_rate": 3.673842404549147e-06, + "loss": 1.615, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.565157413482666, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6351, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.7868876457214355, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.7186, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.7061610221862793, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.6454, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.587785482406616, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6867, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.4618277549743652, + "learning_rate": 3.663688058489034e-06, + "loss": 1.6189, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.4425477981567383, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5772, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.7514469623565674, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6368, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8366374969482422, + "eval_runtime": 44.4679, + "eval_samples_per_second": 22.488, + "eval_steps_per_second": 0.945, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.529805898666382, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.6229, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6875312328338623, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6445, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.715970993041992, + "learning_rate": 3.65353371242892e-06, + "loss": 1.5318, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.8685660362243652, + "learning_rate": 3.651502843216897e-06, + "loss": 1.6465, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.6801602840423584, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.5959, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.61037015914917, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6561, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.478137493133545, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6435, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5708508491516113, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6206, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8857522010803223, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6269, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.6102395057678223, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.648, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.836978793144226, + "eval_runtime": 44.9452, + "eval_samples_per_second": 22.249, + "eval_steps_per_second": 0.934, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7924158573150635, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6802, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.784693479537964, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.6515, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.747894525527954, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6093, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.635850191116333, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6791, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.5760908126831055, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.5721, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5258584022521973, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5644, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.919623613357544, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.6258, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.8330483436584473, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6077, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8145744800567627, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6463, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.9187726974487305, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6348, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.837236762046814, + "eval_runtime": 45.4421, + "eval_samples_per_second": 22.006, + "eval_steps_per_second": 0.924, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.6741979122161865, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6441, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.556701898574829, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5523, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.6352901458740234, + "learning_rate": 3.612916328188465e-06, + "loss": 1.6545, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.5784106254577637, + "learning_rate": 3.610885458976442e-06, + "loss": 1.6649, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.7136101722717285, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.596, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.608877182006836, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.5923, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.783069372177124, + "learning_rate": 3.604792851340374e-06, + "loss": 1.6685, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.621882438659668, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6404, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.672128915786743, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5713, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.645395040512085, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.609, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8348923921585083, + "eval_runtime": 44.8379, + "eval_samples_per_second": 22.303, + "eval_steps_per_second": 0.937, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 3.610335111618042, + "learning_rate": 3.596669374492283e-06, + "loss": 1.6693, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.7391626834869385, + "learning_rate": 3.59463850528026e-06, + "loss": 1.5847, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.658418655395508, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6552, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.6508545875549316, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6201, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 4.237835884094238, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.6725, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.672144889831543, + "learning_rate": 3.586515028432169e-06, + "loss": 1.6666, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.619480609893799, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5671, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 3.150592565536499, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6118, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.946467161178589, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6499, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.6372971534729004, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5516, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8352437019348145, + "eval_runtime": 46.1704, + "eval_samples_per_second": 21.659, + "eval_steps_per_second": 0.91, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.6211962699890137, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6414, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.567579746246338, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.6632, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.4393808841705322, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6323, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.7005021572113037, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6016, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7400808334350586, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6327, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.922241449356079, + "learning_rate": 3.566206336311942e-06, + "loss": 1.5866, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.524975299835205, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5707, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.637805461883545, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6259, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7430527210235596, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.595, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.5548908710479736, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.625, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8332648277282715, + "eval_runtime": 48.5515, + "eval_samples_per_second": 20.597, + "eval_steps_per_second": 0.865, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.5396368503570557, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.655, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.523167848587036, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6171, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.6028409004211426, + "learning_rate": 3.551990251827783e-06, + "loss": 1.5872, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.607464075088501, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6176, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.510603427886963, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.5999, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.444693088531494, + "learning_rate": 3.545897644191714e-06, + "loss": 1.5791, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6724541187286377, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6358, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.64925217628479, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5793, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.647118091583252, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6399, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8507258892059326, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6046, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8305917978286743, + "eval_runtime": 43.2738, + "eval_samples_per_second": 23.109, + "eval_steps_per_second": 0.971, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.5572781562805176, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6395, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.530216693878174, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6493, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.564739465713501, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6075, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.5772671699523926, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.64, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.828692674636841, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6584, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.700568675994873, + "learning_rate": 3.525588952071487e-06, + "loss": 1.6595, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.666743755340576, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6483, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.551187515258789, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.5946, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.553602933883667, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.5997, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.5355968475341797, + "learning_rate": 3.517465475223396e-06, + "loss": 1.571, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8321614265441895, + "eval_runtime": 46.1746, + "eval_samples_per_second": 21.657, + "eval_steps_per_second": 0.91, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.707101345062256, + "learning_rate": 3.515434606011373e-06, + "loss": 1.6389, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.8770599365234375, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.6828, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.5841145515441895, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6264, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.697502851486206, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6081, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.620100259780884, + "learning_rate": 3.507311129163282e-06, + "loss": 1.5617, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.6503913402557373, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6194, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.830296754837036, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.616, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.6960535049438477, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5523, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.59023118019104, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6378, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.582906484603882, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6368, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8257222175598145, + "eval_runtime": 47.9573, + "eval_samples_per_second": 20.852, + "eval_steps_per_second": 0.876, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.6826765537261963, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5685, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6153650283813477, + "learning_rate": 3.493095044679123e-06, + "loss": 1.6948, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.607027769088745, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5197, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.493180990219116, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6159, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.6291017532348633, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5404, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.90544056892395, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6164, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.673076629638672, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6486, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.5843210220336914, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.6127, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.55414080619812, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6255, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.496095657348633, + "learning_rate": 3.476848090982941e-06, + "loss": 1.5941, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8262794017791748, + "eval_runtime": 47.0577, + "eval_samples_per_second": 21.25, + "eval_steps_per_second": 0.893, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.597395658493042, + "learning_rate": 3.474817221770918e-06, + "loss": 1.6941, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.4487929344177246, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6417, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.63041353225708, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.611, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.481630325317383, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.575, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.7625951766967773, + "learning_rate": 3.466693744922827e-06, + "loss": 1.5988, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.663846492767334, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6326, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.594536781311035, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6004, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.5871670246124268, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6378, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.591907262802124, + "learning_rate": 3.458570268074736e-06, + "loss": 1.6093, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.456110954284668, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6014, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8266254663467407, + "eval_runtime": 47.4255, + "eval_samples_per_second": 21.086, + "eval_steps_per_second": 0.886, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.8052752017974854, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.5679, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.7458856105804443, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6239, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.4829750061035156, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.5692, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.5427238941192627, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6214, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.597419261932373, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6012, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.5192158222198486, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6469, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.4233906269073486, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5822, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.4213144779205322, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5345, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.763439655303955, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6092, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.228456497192383, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6591, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8253848552703857, + "eval_runtime": 50.1479, + "eval_samples_per_second": 19.941, + "eval_steps_per_second": 0.838, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.670700788497925, + "learning_rate": 3.434199837530463e-06, + "loss": 1.612, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.797170877456665, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5624, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8155477046966553, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6535, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.7093656063079834, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6479, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.6133127212524414, + "learning_rate": 3.426076360682372e-06, + "loss": 1.6669, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.471862554550171, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6136, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7078938484191895, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5382, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.6206846237182617, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6186, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.674201011657715, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.6805, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.7543511390686035, + "learning_rate": 3.415922014622259e-06, + "loss": 1.644, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8275290727615356, + "eval_runtime": 49.1382, + "eval_samples_per_second": 20.351, + "eval_steps_per_second": 0.855, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.934906244277954, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6502, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.6866018772125244, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6324, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.743281126022339, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6229, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.7543251514434814, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.5813, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8038830757141113, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6482, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.6368162631988525, + "learning_rate": 3.403736799350122e-06, + "loss": 1.618, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7447240352630615, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.5735, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.6714727878570557, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.686, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.7939751148223877, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6104, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.624718189239502, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6194, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8199249505996704, + "eval_runtime": 50.4903, + "eval_samples_per_second": 19.806, + "eval_steps_per_second": 0.832, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.563082695007324, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5669, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.6971375942230225, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6237, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.7873635292053223, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6241, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.570742130279541, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6304, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8116018772125244, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6068, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.6097946166992188, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.5996, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.593461275100708, + "learning_rate": 3.381397238017872e-06, + "loss": 1.5935, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.008805274963379, + "learning_rate": 3.379366368805849e-06, + "loss": 1.68, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7567968368530273, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6129, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.647456645965576, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.618, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8231184482574463, + "eval_runtime": 46.6144, + "eval_samples_per_second": 21.453, + "eval_steps_per_second": 0.901, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.842937707901001, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6118, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.794092893600464, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.5922, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.837575674057007, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6267, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.7084648609161377, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5597, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.626157283782959, + "learning_rate": 3.36515028432169e-06, + "loss": 1.5961, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.589062213897705, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6235, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.681427478790283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6568, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 3.3202719688415527, + "learning_rate": 3.359057676685622e-06, + "loss": 1.649, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.463998794555664, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5482, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.5475661754608154, + "learning_rate": 3.354995938261576e-06, + "loss": 1.586, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8188257217407227, + "eval_runtime": 51.7434, + "eval_samples_per_second": 19.326, + "eval_steps_per_second": 0.812, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.579566478729248, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5546, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6778531074523926, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.5831, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.98976731300354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6199, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.8345465660095215, + "learning_rate": 3.346872461413485e-06, + "loss": 1.5697, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.6166720390319824, + "learning_rate": 3.344841592201463e-06, + "loss": 1.5829, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.5504395961761475, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.5845, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.6515390872955322, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6154, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.589524269104004, + "learning_rate": 3.338748984565394e-06, + "loss": 1.639, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.8798511028289795, + "learning_rate": 3.336718115353372e-06, + "loss": 1.676, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.984379529953003, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6154, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.818778157234192, + "eval_runtime": 44.8621, + "eval_samples_per_second": 22.291, + "eval_steps_per_second": 0.936, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.5935728549957275, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.5895, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.7658374309539795, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6041, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.580352306365967, + "learning_rate": 3.32859463850528e-06, + "loss": 1.6127, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.689892292022705, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5624, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.708078622817993, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.5782, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.7298295497894287, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.5815, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8700411319732666, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5597, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.6671183109283447, + "learning_rate": 3.318440292445167e-06, + "loss": 1.5902, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.11922550201416, + "learning_rate": 3.316409423233144e-06, + "loss": 1.5678, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.72869610786438, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6568, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8164011240005493, + "eval_runtime": 47.4229, + "eval_samples_per_second": 21.087, + "eval_steps_per_second": 0.886, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.7037997245788574, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6191, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.6406683921813965, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6018, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.4749815464019775, + "learning_rate": 3.308285946385053e-06, + "loss": 1.679, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.5094757080078125, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5455, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.6685080528259277, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6123, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7547595500946045, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.6497, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.7807092666625977, + "learning_rate": 3.300162469536962e-06, + "loss": 1.5903, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.903127431869507, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6398, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.4666707515716553, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6194, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.9508626461029053, + "learning_rate": 3.294069861900894e-06, + "loss": 1.6774, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8119163513183594, + "eval_runtime": 51.0141, + "eval_samples_per_second": 19.602, + "eval_steps_per_second": 0.823, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.585411787033081, + "learning_rate": 3.292038992688871e-06, + "loss": 1.6086, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.7636921405792236, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.6386, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.72711181640625, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.6563, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.5698368549346924, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5501, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8764665126800537, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6229, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.653674840927124, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5343, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.896165609359741, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.6617, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.714388847351074, + "learning_rate": 3.277822908204712e-06, + "loss": 1.605, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.572202682495117, + "learning_rate": 3.275792038992689e-06, + "loss": 1.568, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.719407081604004, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6116, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.811676263809204, + "eval_runtime": 46.0392, + "eval_samples_per_second": 21.721, + "eval_steps_per_second": 0.912, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.752934694290161, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.5672, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.519256353378296, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6101, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.758317708969116, + "learning_rate": 3.267668562144598e-06, + "loss": 1.597, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.6348824501037598, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.5731, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.699477434158325, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6288, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.5104498863220215, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6024, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.680694341659546, + "learning_rate": 3.259545085296507e-06, + "loss": 1.6689, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.459789276123047, + "learning_rate": 3.257514216084484e-06, + "loss": 1.56, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.533125877380371, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.6734, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.640742301940918, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6214, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8131158351898193, + "eval_runtime": 47.2273, + "eval_samples_per_second": 21.174, + "eval_steps_per_second": 0.889, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.4593327045440674, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.5997, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.6157352924346924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6193, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.5011303424835205, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.5796, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.545781373977661, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.6128, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.5882468223571777, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6106, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.7500388622283936, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5491, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5432305335998535, + "learning_rate": 3.23923639317628e-06, + "loss": 1.5879, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.5674941539764404, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.5737, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.630491256713867, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6167, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.7517776489257812, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.6698, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8172497749328613, + "eval_runtime": 45.7554, + "eval_samples_per_second": 21.855, + "eval_steps_per_second": 0.918, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5155255794525146, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5647, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.5586776733398438, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6403, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7582011222839355, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5901, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.611339569091797, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.599, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.7254250049591064, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6167, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.5621025562286377, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5241, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849069118499756, + "learning_rate": 3.218927701056052e-06, + "loss": 1.5823, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.874762535095215, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5294, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.706822395324707, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.6529, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.7066354751586914, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5591, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.814163327217102, + "eval_runtime": 43.7518, + "eval_samples_per_second": 22.856, + "eval_steps_per_second": 0.96, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.584439516067505, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.5986, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.671645402908325, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6417, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6757681369781494, + "learning_rate": 3.206742485783916e-06, + "loss": 1.5594, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.6210391521453857, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5476, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.5632853507995605, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6165, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.64302921295166, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.631, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.4642136096954346, + "learning_rate": 3.198619008935825e-06, + "loss": 1.534, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.196588139723802e-06, + "loss": 1.598, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.5753607749938965, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6177, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.787393093109131, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.607, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.809692621231079, + "eval_runtime": 44.1276, + "eval_samples_per_second": 22.662, + "eval_steps_per_second": 0.952, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.4849026203155518, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6064, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.484736204147339, + "learning_rate": 3.188464662875711e-06, + "loss": 1.6591, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.652766466140747, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5931, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.6450769901275635, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5463, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.736511468887329, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5509, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.8212637901306152, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6244, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.7004551887512207, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6123, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.665766954421997, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.5904, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.7625415325164795, + "learning_rate": 3.174248578391552e-06, + "loss": 1.5843, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.5376832485198975, + "learning_rate": 3.172217709179529e-06, + "loss": 1.5758, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8079267740249634, + "eval_runtime": 45.139, + "eval_samples_per_second": 22.154, + "eval_steps_per_second": 0.93, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.704226016998291, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.5891, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6165294647216797, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5502, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.6848294734954834, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6134, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.778846502304077, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.611, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.647531747817993, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5531, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.605592966079712, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6011, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5820205211639404, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6121, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.7092363834381104, + "learning_rate": 3.155970755483347e-06, + "loss": 1.5928, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.7473032474517822, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6068, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.742840528488159, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6041, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8063452243804932, + "eval_runtime": 44.2112, + "eval_samples_per_second": 22.619, + "eval_steps_per_second": 0.95, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.7613816261291504, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6413, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.876913547515869, + "learning_rate": 3.147847278635256e-06, + "loss": 1.6848, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5945441722869873, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.5847, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.526456117630005, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5496, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.9870903491973877, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.5982, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6774742603302, + "learning_rate": 3.139723801787165e-06, + "loss": 1.5678, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.69058895111084, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6174, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.8981339931488037, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5529, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.522247552871704, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6061, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.709726333618164, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6267, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8052302598953247, + "eval_runtime": 42.6337, + "eval_samples_per_second": 23.456, + "eval_steps_per_second": 0.985, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.618793249130249, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6441, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5927956104278564, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6035, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.557023048400879, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6252, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5910511016845703, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.5883, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.6681272983551025, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6147, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.7024455070495605, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6317, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.6808700561523438, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.5871, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.585388422012329, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.6202, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.637092113494873, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6357, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.652611494064331, + "learning_rate": 3.111291632818847e-06, + "loss": 1.651, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.804667592048645, + "eval_runtime": 43.1932, + "eval_samples_per_second": 23.152, + "eval_steps_per_second": 0.972, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.770113945007324, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6005, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.570833206176758, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6169, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.537957191467285, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.5826, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.715669631958008, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6276, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.5450210571289062, + "learning_rate": 3.101137286758733e-06, + "loss": 1.6305, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.6358814239501953, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6367, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.7380471229553223, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.5984, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 3.030034303665161, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.6661, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.6777493953704834, + "learning_rate": 3.093013809910642e-06, + "loss": 1.6122, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8807458877563477, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.6514, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8064097166061401, + "eval_runtime": 43.3673, + "eval_samples_per_second": 23.059, + "eval_steps_per_second": 0.968, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.545091390609741, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6182, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.6389622688293457, + "learning_rate": 3.086921202274574e-06, + "loss": 1.5931, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.5784780979156494, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5579, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.67499041557312, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6128, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.558598756790161, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.5958, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6460611820220947, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5274, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.770334482192993, + "learning_rate": 3.07676685621446e-06, + "loss": 1.604, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6441917419433594, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.567, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.6142544746398926, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.5891, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.6242077350616455, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6059, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8036692142486572, + "eval_runtime": 45.6003, + "eval_samples_per_second": 21.93, + "eval_steps_per_second": 0.921, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.417332410812378, + "learning_rate": 3.068643379366369e-06, + "loss": 1.5991, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.4537513256073, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5603, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.745912551879883, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6171, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.636627674102783, + "learning_rate": 3.062550771730301e-06, + "loss": 1.5839, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.6001460552215576, + "learning_rate": 3.060519902518278e-06, + "loss": 1.5781, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.572185516357422, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5188, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.87540864944458, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.621, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.7670910358428955, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.584, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.672682523727417, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6293, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.655538558959961, + "learning_rate": 3.050365556458164e-06, + "loss": 1.5773, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8046640157699585, + "eval_runtime": 44.6243, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.582319974899292, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.5751, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.7160403728485107, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.5972, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.6800436973571777, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6311, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.8212592601776123, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6315, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.57470703125, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5414, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 3.3979809284210205, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.5732, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.606290102005005, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.5966, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.5241527557373047, + "learning_rate": 3.034118602761982e-06, + "loss": 1.588, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.7115085124969482, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5619, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.7072222232818604, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6346, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.804133415222168, + "eval_runtime": 44.5547, + "eval_samples_per_second": 22.444, + "eval_steps_per_second": 0.943, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.5874669551849365, + "learning_rate": 3.028025995125914e-06, + "loss": 1.623, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.644073247909546, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5484, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.553112506866455, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.5923, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6504719257354736, + "learning_rate": 3.021933387489846e-06, + "loss": 1.543, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.7608838081359863, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6001, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.690267562866211, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6093, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.8565986156463623, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5614, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.672799587249756, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5657, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5756146907806396, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.5664, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.8423633575439453, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.5936, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.800963044166565, + "eval_runtime": 46.5083, + "eval_samples_per_second": 21.502, + "eval_steps_per_second": 0.903, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.6263201236724854, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.5693, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 3.0006930828094482, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6084, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.7002620697021484, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.5667, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.722398519515991, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6285, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.738555908203125, + "learning_rate": 2.999593826157596e-06, + "loss": 1.5849, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.726271390914917, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.5951, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.6749048233032227, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.5807, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.774726629257202, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6355, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6740102767944336, + "learning_rate": 2.991470349309505e-06, + "loss": 1.5871, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.7571794986724854, + "learning_rate": 2.989439480097482e-06, + "loss": 1.578, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.7960126399993896, + "eval_runtime": 44.4777, + "eval_samples_per_second": 22.483, + "eval_steps_per_second": 0.944, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.67576265335083, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.5917, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.8302817344665527, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6313, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.874002456665039, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6212, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6992602348327637, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6011, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.726292371749878, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.5981, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.988934278488159, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6272, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.6781861782073975, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.5742, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.818401336669922, + "learning_rate": 2.9731925264013e-06, + "loss": 1.5713, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.6900908946990967, + "learning_rate": 2.971161657189277e-06, + "loss": 1.5754, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.9017207622528076, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.5746, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.7994096279144287, + "eval_runtime": 46.997, + "eval_samples_per_second": 21.278, + "eval_steps_per_second": 0.894, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.0305914878845215, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.5733, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.695298910140991, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6379, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.60552716255188, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5488, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.664668321609497, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6256, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.700955390930176, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6228, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.637998104095459, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.5565, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.664620876312256, + "learning_rate": 2.954914703493095e-06, + "loss": 1.5805, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.8120782375335693, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.5615, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.641434907913208, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.56, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.845808982849121, + "learning_rate": 2.948822095857027e-06, + "loss": 1.5954, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8004399538040161, + "eval_runtime": 47.1938, + "eval_samples_per_second": 21.189, + "eval_steps_per_second": 0.89, + "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_local3_10M_seed0/runs/checkpoint-1300/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1300/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1400/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1400/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1400/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/latest new file mode 100644 index 0000000000000000000000000000000000000000..c2a0b5b17b996514a9bea0debe9403a641844bce --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/latest @@ -0,0 +1 @@ +global_step1400 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1400/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1400/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1400/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..3b4b7c426483bc23d244697ca4016b621bc2a786 --- /dev/null +++ b/babylm_shuffle_local3_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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.7659518718719482, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.279, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.433785915374756, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.3368, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 3.1486687660217285, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3896, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.391956090927124, + "learning_rate": 1.624087591240876e-06, + "loss": 2.297, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.6649584770202637, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.2415, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.5604891777038574, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2688, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.47900652885437, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2933, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.632723569869995, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.2143, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.8085947036743164, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2748, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.3927979469299316, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2685, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.355207920074463, + "eval_runtime": 42.3505, + "eval_samples_per_second": 23.612, + "eval_steps_per_second": 0.992, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.6170592308044434, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.3121, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.699814796447754, + "learning_rate": 1.77007299270073e-06, + "loss": 2.2853, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.3450610637664795, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3297, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.458219528198242, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2639, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.7101449966430664, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2134, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.5734996795654297, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2347, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 2.5200352668762207, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2299, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5022480487823486, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2253, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.190117359161377, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.2272, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.414524555206299, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2424, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.317634344100952, + "eval_runtime": 43.5742, + "eval_samples_per_second": 22.949, + "eval_steps_per_second": 0.964, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.3226804733276367, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1449, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.733006477355957, + "learning_rate": 1.952554744525548e-06, + "loss": 2.2027, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 2.5586233139038086, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2007, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.669914722442627, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2024, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.6148102283477783, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2496, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 2.331369638442993, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1814, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.3051886558532715, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2015, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.6890218257904053, + "learning_rate": 2.062043795620438e-06, + "loss": 2.184, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.4975883960723877, + "learning_rate": 2.08029197080292e-06, + "loss": 2.2439, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.5848891735076904, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1413, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.280966281890869, + "eval_runtime": 43.1001, + "eval_samples_per_second": 23.202, + "eval_steps_per_second": 0.974, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.2104, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.3356540203094482, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2312, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 2.3616602420806885, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.1399, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.581623077392578, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1456, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.547288656234741, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.184, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6386160850524902, + "learning_rate": 2.208029197080292e-06, + "loss": 2.1658, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.593065023422241, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1443, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.530632972717285, + "learning_rate": 2.244525547445256e-06, + "loss": 2.194, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.3553175926208496, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.1918, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 2.7536532878875732, + "learning_rate": 2.281021897810219e-06, + "loss": 2.1477, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2608063220977783, + "eval_runtime": 42.7938, + "eval_samples_per_second": 23.368, + "eval_steps_per_second": 0.981, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.160094738006592, + "learning_rate": 2.299270072992701e-06, + "loss": 2.133, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.4474735260009766, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1468, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.6654434204101562, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.1194, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.726611375808716, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1518, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 2.3560168743133545, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1537, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.664065361022949, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0888, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.3946099281311035, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.0998, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.7812652587890625, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1231, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.64595890045166, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1746, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.788524866104126, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0731, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.2340502738952637, + "eval_runtime": 42.3104, + "eval_samples_per_second": 23.635, + "eval_steps_per_second": 0.993, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 2.9249846935272217, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0564, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.5507798194885254, + "learning_rate": 2.5e-06, + "loss": 2.1714, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.1604716777801514, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0701, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.5429840087890625, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.0662, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8296737670898438, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.1378, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.9396755695343018, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1069, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.736814022064209, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1076, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 2.774909496307373, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1286, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.0861318111419678, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1499, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.624016284942627, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0989, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.215052604675293, + "eval_runtime": 43.9956, + "eval_samples_per_second": 22.73, + "eval_steps_per_second": 0.955, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.801500082015991, + "learning_rate": 2.664233576642336e-06, + "loss": 2.0795, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 2.2107114791870117, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.1035, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.030381441116333, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.0794, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 2.9719064235687256, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1496, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.688839912414551, + "learning_rate": 2.737226277372263e-06, + "loss": 2.1152, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 2.9350979328155518, + "learning_rate": 2.755474452554745e-06, + "loss": 2.0773, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.8958046436309814, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.1805, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.0781922340393066, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.1249, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.032059907913208, + "learning_rate": 2.81021897810219e-06, + "loss": 2.0632, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.798758029937744, + "learning_rate": 2.828467153284672e-06, + "loss": 2.1001, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2005105018615723, + "eval_runtime": 43.5111, + "eval_samples_per_second": 22.983, + "eval_steps_per_second": 0.965, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.7674927711486816, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.0211, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 2.6168575286865234, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0599, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 2.933353900909424, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.0169, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.81402587890625, + "learning_rate": 2.901459854014599e-06, + "loss": 2.122, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.5592572689056396, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.0957, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.738271951675415, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.0868, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.5362703800201416, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1237, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.6304945945739746, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.0825, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.040717601776123, + "learning_rate": 2.992700729927008e-06, + "loss": 2.0388, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.6226463317871094, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0163, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.186411142349243, + "eval_runtime": 44.625, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.5307068824768066, + "learning_rate": 3.029197080291971e-06, + "loss": 2.1097, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 2.9160444736480713, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.0946, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.5250024795532227, + "learning_rate": 3.065693430656935e-06, + "loss": 2.0709, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.6963064670562744, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.0937, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 2.845707416534424, + "learning_rate": 3.102189781021898e-06, + "loss": 2.0851, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.8985090255737305, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.0176, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 2.734377861022949, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.9931, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.5125555992126465, + "learning_rate": 3.156934306569343e-06, + "loss": 2.1133, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.0393502712249756, + "learning_rate": 3.175182481751825e-06, + "loss": 2.0636, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.6410164833068848, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0304, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.171868085861206, + "eval_runtime": 44.2249, + "eval_samples_per_second": 22.612, + "eval_steps_per_second": 0.95, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.7937498092651367, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0518, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7801201343536377, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.026, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7963638305664062, + "learning_rate": 3.248175182481752e-06, + "loss": 1.9642, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6330482959747314, + "learning_rate": 3.266423357664234e-06, + "loss": 2.0545, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.9771625995635986, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0445, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.6112730503082275, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0616, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.732865333557129, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0403, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.9113688468933105, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0382, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.4516773223876953, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.0416, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.781735897064209, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0348, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.160499334335327, + "eval_runtime": 44.5652, + "eval_samples_per_second": 22.439, + "eval_steps_per_second": 0.942, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.9990546703338623, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0324, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.783071279525757, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0518, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.659040689468384, + "learning_rate": 3.43065693430657e-06, + "loss": 1.9722, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 2.799798011779785, + "learning_rate": 3.4489051094890513e-06, + "loss": 2.0362, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 3.025458812713623, + "learning_rate": 3.467153284671533e-06, + "loss": 1.9863, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 2.686415195465088, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0201, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.716064929962158, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0249, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.637653350830078, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0403, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 2.7686848640441895, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0626, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.721956729888916, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.1077, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.152338981628418, + "eval_runtime": 43.5455, + "eval_samples_per_second": 22.964, + "eval_steps_per_second": 0.965, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.92571759223938, + "learning_rate": 3.576642335766424e-06, + "loss": 2.0928, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.9189202785491943, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9315, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 2.990233898162842, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9427, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.702348232269287, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9768, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6949572563171387, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.9954, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.108405113220215, + "learning_rate": 3.667883211678833e-06, + "loss": 1.9631, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.738922119140625, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9644, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.7419424057006836, + "learning_rate": 3.704379562043796e-06, + "loss": 2.0675, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9853813648223877, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0245, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.4614126682281494, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.9665, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.1342060565948486, + "eval_runtime": 46.3954, + "eval_samples_per_second": 21.554, + "eval_steps_per_second": 0.905, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.6787819862365723, + "learning_rate": 3.7591240875912412e-06, + "loss": 2.0063, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 2.563525438308716, + "learning_rate": 3.777372262773723e-06, + "loss": 1.998, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.4981701374053955, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.9738, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 2.5357978343963623, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0221, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.5365488529205322, + "learning_rate": 3.832116788321168e-06, + "loss": 1.9664, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 2.669468641281128, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9577, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 2.6950953006744385, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9786, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8249378204345703, + "learning_rate": 3.886861313868613e-06, + "loss": 2.0041, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.5924317836761475, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0338, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.1282334327697754, + "learning_rate": 3.923357664233577e-06, + "loss": 1.9698, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.1221420764923096, + "eval_runtime": 47.1127, + "eval_samples_per_second": 21.226, + "eval_steps_per_second": 0.891, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.7878193855285645, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0446, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.7137255668640137, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0575, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.286421060562134, + "learning_rate": 3.978102189781022e-06, + "loss": 1.9131, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.5466175079345703, + "learning_rate": 3.996350364963504e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.792483329772949, + "learning_rate": 4.014598540145986e-06, + "loss": 1.912, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.7533867359161377, + "learning_rate": 4.032846715328467e-06, + "loss": 1.9742, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 2.848405361175537, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9815, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.1034414768218994, + "learning_rate": 4.069343065693431e-06, + "loss": 1.9409, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.5109779834747314, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9676, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.486506938934326, + "learning_rate": 4.105839416058395e-06, + "loss": 2.0353, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1102187633514404, + "eval_runtime": 45.384, + "eval_samples_per_second": 22.034, + "eval_steps_per_second": 0.925, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.419269323348999, + "learning_rate": 4.124087591240876e-06, + "loss": 1.9622, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.4967262744903564, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.9523, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 2.7311460971832275, + "learning_rate": 4.16058394160584e-06, + "loss": 2.0234, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.241004467010498, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.907, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.821537733078003, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9761, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 2.5711426734924316, + "learning_rate": 4.215328467153285e-06, + "loss": 1.9411, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.936594009399414, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9432, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.6091108322143555, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9545, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.522674798965454, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9246, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.984649181365967, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9417, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1031947135925293, + "eval_runtime": 47.5409, + "eval_samples_per_second": 21.035, + "eval_steps_per_second": 0.883, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.8250324726104736, + "learning_rate": 4.306569343065693e-06, + "loss": 1.9061, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.7654290199279785, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9653, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.7091264724731445, + "learning_rate": 4.343065693430658e-06, + "loss": 1.9482, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 2.8019583225250244, + "learning_rate": 4.361313868613139e-06, + "loss": 1.9332, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.147368907928467, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9836, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.7050259113311768, + "learning_rate": 4.397810218978102e-06, + "loss": 1.9607, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.738743543624878, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9374, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.122777223587036, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9249, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.850745439529419, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9707, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.391998291015625, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.9703, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.0887911319732666, + "eval_runtime": 48.5149, + "eval_samples_per_second": 20.612, + "eval_steps_per_second": 0.866, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 2.9131698608398438, + "learning_rate": 4.489051094890512e-06, + "loss": 1.9989, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.1045937538146973, + "learning_rate": 4.507299270072993e-06, + "loss": 1.9718, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.7984158992767334, + "learning_rate": 4.525547445255475e-06, + "loss": 1.947, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.6223504543304443, + "learning_rate": 4.543795620437956e-06, + "loss": 1.8942, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.8221523761749268, + "learning_rate": 4.562043795620438e-06, + "loss": 2.0045, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.6657445430755615, + "learning_rate": 4.580291970802921e-06, + "loss": 2.0053, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 2.793455123901367, + "learning_rate": 4.598540145985402e-06, + "loss": 1.9577, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.551201820373535, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9256, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 2.739194631576538, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9265, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.2917332649230957, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9664, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.086106777191162, + "eval_runtime": 48.995, + "eval_samples_per_second": 20.41, + "eval_steps_per_second": 0.857, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 2.6489369869232178, + "learning_rate": 4.671532846715329e-06, + "loss": 1.9629, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 2.60017466545105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.932, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.569115400314331, + "learning_rate": 4.708029197080292e-06, + "loss": 1.9902, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 2.7225773334503174, + "learning_rate": 4.726277372262774e-06, + "loss": 1.9619, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.5064749717712402, + "learning_rate": 4.744525547445255e-06, + "loss": 1.8921, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.612931966781616, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9584, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.2524304389953613, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9675, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.8852503299713135, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9107, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 2.654776096343994, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9465, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.679086685180664, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9468, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.0842835903167725, + "eval_runtime": 51.7803, + "eval_samples_per_second": 19.312, + "eval_steps_per_second": 0.811, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.5110855102539062, + "learning_rate": 4.854014598540146e-06, + "loss": 1.88, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.1786117553710938, + "learning_rate": 4.872262773722628e-06, + "loss": 1.8994, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.388328790664673, + "learning_rate": 4.89051094890511e-06, + "loss": 1.8914, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 2.784834384918213, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.923, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.470426559448242, + "learning_rate": 4.927007299270074e-06, + "loss": 1.944, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.690580129623413, + "learning_rate": 4.945255474452555e-06, + "loss": 1.8978, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.869737386703491, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9964, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.9965591430664062, + "learning_rate": 4.981751824817518e-06, + "loss": 1.9706, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.6665658950805664, + "learning_rate": 5e-06, + "loss": 1.9313, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.9231722354888916, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9504, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.068364143371582, + "eval_runtime": 47.8334, + "eval_samples_per_second": 20.906, + "eval_steps_per_second": 0.878, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.6094462871551514, + "learning_rate": 4.995938261575955e-06, + "loss": 1.862, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.780848741531372, + "learning_rate": 4.993907392363932e-06, + "loss": 1.95, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.948662757873535, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.878, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 2.9622998237609863, + "learning_rate": 4.989845653939887e-06, + "loss": 1.8954, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 2.6401216983795166, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9021, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.029592275619507, + "learning_rate": 4.985783915515841e-06, + "loss": 1.8967, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.828392505645752, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9134, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.613926649093628, + "learning_rate": 4.981722177091796e-06, + "loss": 1.8829, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.6173207759857178, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9225, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.4657015800476074, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9142, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.059906482696533, + "eval_runtime": 49.6267, + "eval_samples_per_second": 20.15, + "eval_steps_per_second": 0.846, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7448673248291016, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9545, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.639096736907959, + "learning_rate": 4.973598700243705e-06, + "loss": 1.8664, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 2.6445884704589844, + "learning_rate": 4.971567831031681e-06, + "loss": 1.888, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.9130005836486816, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7717, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.48093318939209, + "learning_rate": 4.967506092607636e-06, + "loss": 1.8538, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.8918230533599854, + "learning_rate": 4.965475223395614e-06, + "loss": 1.886, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.511747121810913, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.8633, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.860029697418213, + "learning_rate": 4.961413484971569e-06, + "loss": 1.9063, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7304794788360596, + "learning_rate": 4.959382615759545e-06, + "loss": 1.8872, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.6569266319274902, + "learning_rate": 4.957351746547523e-06, + "loss": 1.8721, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0610575675964355, + "eval_runtime": 43.9292, + "eval_samples_per_second": 22.764, + "eval_steps_per_second": 0.956, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.1482994556427, + "learning_rate": 4.9553208773355e-06, + "loss": 2.0123, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.59568190574646, + "learning_rate": 4.953290008123478e-06, + "loss": 1.902, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.5897581577301025, + "learning_rate": 4.951259138911454e-06, + "loss": 1.8406, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.734032154083252, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.9605, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.8088371753692627, + "learning_rate": 4.947197400487409e-06, + "loss": 1.896, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.9678707122802734, + "learning_rate": 4.945166531275386e-06, + "loss": 1.8625, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.261200189590454, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8884, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 2.5811619758605957, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9311, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.6783649921417236, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9324, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.6533944606781006, + "learning_rate": 4.937043054427295e-06, + "loss": 1.8575, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.049668312072754, + "eval_runtime": 47.7347, + "eval_samples_per_second": 20.949, + "eval_steps_per_second": 0.88, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 3.0882792472839355, + "learning_rate": 4.935012185215272e-06, + "loss": 1.84, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.8259081840515137, + "learning_rate": 4.93298131600325e-06, + "loss": 1.9109, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.540908098220825, + "learning_rate": 4.930950446791227e-06, + "loss": 1.8906, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.9277396202087402, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9047, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.392449140548706, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.876, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 3.1890463829040527, + "learning_rate": 4.924857839155159e-06, + "loss": 1.8709, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.725710391998291, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8953, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.6287014484405518, + "learning_rate": 4.920796100731113e-06, + "loss": 1.8204, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.7316651344299316, + "learning_rate": 4.918765231519091e-06, + "loss": 1.882, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.5561037063598633, + "learning_rate": 4.916734362307068e-06, + "loss": 1.8331, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0412285327911377, + "eval_runtime": 50.237, + "eval_samples_per_second": 19.906, + "eval_steps_per_second": 0.836, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.974501132965088, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8414, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.4530839920043945, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8323, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.789569616317749, + "learning_rate": 4.910641754671e-06, + "loss": 1.8697, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.582704544067383, + "learning_rate": 4.908610885458977e-06, + "loss": 1.8592, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 2.776442289352417, + "learning_rate": 4.906580016246954e-06, + "loss": 1.912, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.670191764831543, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.8899, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.4534454345703125, + "learning_rate": 4.902518277822909e-06, + "loss": 1.8878, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.007469654083252, + "learning_rate": 4.900487408610886e-06, + "loss": 1.8977, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 2.856466293334961, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9062, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.817441463470459, + "learning_rate": 4.89642567018684e-06, + "loss": 1.9686, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.0344440937042236, + "eval_runtime": 47.2472, + "eval_samples_per_second": 21.165, + "eval_steps_per_second": 0.889, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.9732158184051514, + "learning_rate": 4.894394800974818e-06, + "loss": 1.8332, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 3.0413055419921875, + "learning_rate": 4.892363931762794e-06, + "loss": 1.9171, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.646749973297119, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.8403, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.1274876594543457, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8218, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.8122124671936035, + "learning_rate": 4.886271324126727e-06, + "loss": 1.9272, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.4608492851257324, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.8974, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.928083658218384, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.8646, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.6308164596557617, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8219, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7311816215515137, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8326, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.640162467956543, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8013, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0261101722717285, + "eval_runtime": 47.0605, + "eval_samples_per_second": 21.249, + "eval_steps_per_second": 0.892, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.8542518615722656, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8233, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.464303970336914, + "learning_rate": 4.872055239642567e-06, + "loss": 1.787, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8535103797912598, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8535, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.5060505867004395, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8733, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.6549222469329834, + "learning_rate": 4.865962632006499e-06, + "loss": 1.8047, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.8636224269866943, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8543, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.644256830215454, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8853, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.949490547180176, + "learning_rate": 4.859870024370431e-06, + "loss": 1.8951, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.4576282501220703, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9069, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.6140542030334473, + "learning_rate": 4.855808285946385e-06, + "loss": 1.8573, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0159800052642822, + "eval_runtime": 44.2447, + "eval_samples_per_second": 22.602, + "eval_steps_per_second": 0.949, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.525766134262085, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.844, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.5479896068573, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8572, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.628570795059204, + "learning_rate": 4.849715678310317e-06, + "loss": 1.865, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.2768666744232178, + "learning_rate": 4.847684809098294e-06, + "loss": 1.8355, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.635516881942749, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8603, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.5307435989379883, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8482, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.3688952922821045, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8894, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.6000115871429443, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.8045, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7782504558563232, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8283, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.6914777755737305, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8765, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0086047649383545, + "eval_runtime": 48.245, + "eval_samples_per_second": 20.728, + "eval_steps_per_second": 0.871, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.598365306854248, + "learning_rate": 4.833468724614135e-06, + "loss": 1.8461, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.6718530654907227, + "learning_rate": 4.831437855402113e-06, + "loss": 1.7453, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.393528699874878, + "learning_rate": 4.82940698619009e-06, + "loss": 1.8275, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 3.0622947216033936, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8658, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.431041955947876, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8601, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.939424514770508, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8649, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.8116796016693115, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8303, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.511960029602051, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8383, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 3.8372581005096436, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8111, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.511948347091675, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8448, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.9998281002044678, + "eval_runtime": 44.7624, + "eval_samples_per_second": 22.34, + "eval_steps_per_second": 0.938, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 3.5780487060546875, + "learning_rate": 4.813160032493907e-06, + "loss": 1.8161, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.81740403175354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.8633, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.8117666244506836, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8385, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.5791451930999756, + "learning_rate": 4.80706742485784e-06, + "loss": 1.896, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6816465854644775, + "learning_rate": 4.805036555645816e-06, + "loss": 1.8126, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.6419460773468018, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8753, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.666086196899414, + "learning_rate": 4.800974817221771e-06, + "loss": 1.8305, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8626585006713867, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8463, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.351825475692749, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.838, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.300717830657959, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8711, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.9904799461364746, + "eval_runtime": 43.9511, + "eval_samples_per_second": 22.753, + "eval_steps_per_second": 0.956, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.43884015083313, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8145, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.50593900680542, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8103, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.4643657207489014, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8162, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.4615302085876465, + "learning_rate": 4.786758732737612e-06, + "loss": 1.8637, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.6566433906555176, + "learning_rate": 4.784727863525589e-06, + "loss": 1.7805, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.5973432064056396, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8327, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.4996957778930664, + "learning_rate": 4.780666125101544e-06, + "loss": 1.7761, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 3.0266876220703125, + "learning_rate": 4.778635255889521e-06, + "loss": 1.8036, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.522369146347046, + "learning_rate": 4.776604386677498e-06, + "loss": 1.814, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.482727527618408, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8665, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.9888195991516113, + "eval_runtime": 44.0755, + "eval_samples_per_second": 22.688, + "eval_steps_per_second": 0.953, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 3.237544059753418, + "learning_rate": 4.772542648253453e-06, + "loss": 1.8642, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.393286943435669, + "learning_rate": 4.77051177904143e-06, + "loss": 1.7857, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.66072154045105, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8423, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.6725080013275146, + "learning_rate": 4.766450040617385e-06, + "loss": 1.815, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.7657666206359863, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8729, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.3153769969940186, + "learning_rate": 4.762388302193339e-06, + "loss": 1.8349, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5804553031921387, + "learning_rate": 4.760357432981316e-06, + "loss": 1.8155, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.7280807495117188, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8262, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.551940441131592, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8332, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.6137285232543945, + "learning_rate": 4.754264825345248e-06, + "loss": 1.7805, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.983920693397522, + "eval_runtime": 42.525, + "eval_samples_per_second": 23.516, + "eval_steps_per_second": 0.988, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.4460153579711914, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.7953, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.556032657623291, + "learning_rate": 4.750203086921203e-06, + "loss": 1.7393, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.335983991622925, + "learning_rate": 4.74817221770918e-06, + "loss": 1.8494, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.5057711601257324, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8663, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4447145462036133, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7532, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.769984483718872, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7755, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.472193956375122, + "learning_rate": 4.740048740861089e-06, + "loss": 1.8056, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.4286932945251465, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8094, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.0248074531555176, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8497, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.607700824737549, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8741, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.9779797792434692, + "eval_runtime": 43.7734, + "eval_samples_per_second": 22.845, + "eval_steps_per_second": 0.959, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 3.035045623779297, + "learning_rate": 4.731925264012998e-06, + "loss": 1.8438, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.786921977996826, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7666, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.4158425331115723, + "learning_rate": 4.727863525588953e-06, + "loss": 1.7464, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7772, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.8447, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 3.08266282081604, + "learning_rate": 4.723801787164907e-06, + "loss": 1.7812, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.791749954223633, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8047, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6457693576812744, + "learning_rate": 4.719740048740861e-06, + "loss": 1.836, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.69480037689209, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8145, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6356561183929443, + "learning_rate": 4.715678310316816e-06, + "loss": 1.7756, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.9723366498947144, + "eval_runtime": 45.5806, + "eval_samples_per_second": 21.939, + "eval_steps_per_second": 0.921, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.53474497795105, + "learning_rate": 4.713647441104794e-06, + "loss": 1.774, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.3631107807159424, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8517, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.594604730606079, + "learning_rate": 4.709585702680748e-06, + "loss": 1.8962, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.5280940532684326, + "learning_rate": 4.707554833468725e-06, + "loss": 1.7759, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.4596288204193115, + "learning_rate": 4.705523964256703e-06, + "loss": 1.7948, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.9123644828796387, + "learning_rate": 4.703493095044679e-06, + "loss": 1.7399, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.4354264736175537, + "learning_rate": 4.701462225832657e-06, + "loss": 1.8659, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.28625226020813, + "learning_rate": 4.699431356620634e-06, + "loss": 1.831, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.5445950031280518, + "learning_rate": 4.697400487408612e-06, + "loss": 1.8121, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.767825126647949, + "learning_rate": 4.695369618196588e-06, + "loss": 1.7874, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9653255939483643, + "eval_runtime": 45.2453, + "eval_samples_per_second": 22.102, + "eval_steps_per_second": 0.928, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4052765369415283, + "learning_rate": 4.693338748984566e-06, + "loss": 1.7539, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.498098373413086, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8343, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.821753740310669, + "learning_rate": 4.68927701056052e-06, + "loss": 1.8119, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.4158449172973633, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.7982, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.5229434967041016, + "learning_rate": 4.685215272136475e-06, + "loss": 1.806, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.729199171066284, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8768, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.5102570056915283, + "learning_rate": 4.681153533712429e-06, + "loss": 1.775, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4463515281677246, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7493, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.505733013153076, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8462, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.5784912109375, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7381, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9640673398971558, + "eval_runtime": 44.2714, + "eval_samples_per_second": 22.588, + "eval_steps_per_second": 0.949, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.7864291667938232, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7612, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.507809638977051, + "learning_rate": 4.670999187652316e-06, + "loss": 1.757, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.5499937534332275, + "learning_rate": 4.668968318440293e-06, + "loss": 1.8243, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.435472011566162, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7269, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.541031837463379, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.7959, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5052099227905273, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7191, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6285109519958496, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8423, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.7825605869293213, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7178, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.677319288253784, + "learning_rate": 4.656783103168157e-06, + "loss": 1.7939, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.6830129623413086, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7842, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.960136890411377, + "eval_runtime": 45.9379, + "eval_samples_per_second": 21.769, + "eval_steps_per_second": 0.914, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.5368666648864746, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8352, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.8922412395477295, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.8252, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.8332598209381104, + "learning_rate": 4.648659626320066e-06, + "loss": 1.7805, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.420780658721924, + "learning_rate": 4.646628757108043e-06, + "loss": 1.7499, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.6684062480926514, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.7857, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.8547494411468506, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.8115, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.59279203414917, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8254, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.80961275100708, + "learning_rate": 4.638505280259952e-06, + "loss": 1.7901, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.9017655849456787, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8265, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.595454216003418, + "learning_rate": 4.634443541835907e-06, + "loss": 1.7833, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9576054811477661, + "eval_runtime": 50.7265, + "eval_samples_per_second": 19.714, + "eval_steps_per_second": 0.828, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 4.260303020477295, + "learning_rate": 4.632412672623883e-06, + "loss": 1.6741, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.366147756576538, + "learning_rate": 4.630381803411861e-06, + "loss": 1.7778, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.894336462020874, + "learning_rate": 4.628350934199838e-06, + "loss": 1.8268, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.7675349712371826, + "learning_rate": 4.626320064987816e-06, + "loss": 1.7748, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.461413621902466, + "learning_rate": 4.624289195775792e-06, + "loss": 1.7599, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.670187473297119, + "learning_rate": 4.62225832656377e-06, + "loss": 1.7736, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.8119454383850098, + "learning_rate": 4.620227457351747e-06, + "loss": 1.7837, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.8020036220550537, + "learning_rate": 4.618196588139724e-06, + "loss": 1.7508, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.5977094173431396, + "learning_rate": 4.616165718927701e-06, + "loss": 1.834, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.731919765472412, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.8137, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.9465162754058838, + "eval_runtime": 48.9829, + "eval_samples_per_second": 20.415, + "eval_steps_per_second": 0.857, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.604386568069458, + "learning_rate": 4.612103980503656e-06, + "loss": 1.7854, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.601107120513916, + "learning_rate": 4.610073111291633e-06, + "loss": 1.7749, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6758370399475098, + "learning_rate": 4.60804224207961e-06, + "loss": 1.7826, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.631549119949341, + "learning_rate": 4.606011372867588e-06, + "loss": 1.7723, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.6704764366149902, + "learning_rate": 4.603980503655565e-06, + "loss": 1.7088, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 3.078573703765869, + "learning_rate": 4.601949634443542e-06, + "loss": 1.7254, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7492356300354004, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7756, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 3.0763585567474365, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7806, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.5379223823547363, + "learning_rate": 4.595857026807474e-06, + "loss": 1.7516, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8343496322631836, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8111, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.945749282836914, + "eval_runtime": 46.5453, + "eval_samples_per_second": 21.484, + "eval_steps_per_second": 0.902, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.809753179550171, + "learning_rate": 4.591795288383429e-06, + "loss": 1.7877, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.4196107387542725, + "learning_rate": 4.589764419171406e-06, + "loss": 1.7457, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.409287452697754, + "learning_rate": 4.587733549959383e-06, + "loss": 1.7338, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.414706230163574, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7185, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.591365337371826, + "learning_rate": 4.583671811535338e-06, + "loss": 1.8083, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.4648489952087402, + "learning_rate": 4.581640942323315e-06, + "loss": 1.8192, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.3804972171783447, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8077, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.283082962036133, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8294, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.61795711517334, + "learning_rate": 4.575548334687247e-06, + "loss": 1.7322, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.4035353660583496, + "learning_rate": 4.573517465475224e-06, + "loss": 1.7798, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9422866106033325, + "eval_runtime": 45.3068, + "eval_samples_per_second": 22.072, + "eval_steps_per_second": 0.927, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.7253756523132324, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7876, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.738295078277588, + "learning_rate": 4.569455727051178e-06, + "loss": 1.7933, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 3.1418697834014893, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8145, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.941603660583496, + "learning_rate": 4.565393988627132e-06, + "loss": 1.7671, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.063924551010132, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.7809, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 4.957335472106934, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7306, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.651716709136963, + "learning_rate": 4.559301380991065e-06, + "loss": 1.7428, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 4.777034282684326, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.8118, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.5549471378326416, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7294, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.474388599395752, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8338, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.9402854442596436, + "eval_runtime": 43.166, + "eval_samples_per_second": 23.166, + "eval_steps_per_second": 0.973, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.8968348503112793, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7432, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.6996726989746094, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.7837, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.4599392414093018, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7388, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.745487928390503, + "learning_rate": 4.545085296506905e-06, + "loss": 1.7439, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.6960670948028564, + "learning_rate": 4.543054427294882e-06, + "loss": 1.737, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.596876621246338, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7708, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.598267078399658, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7541, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.6069958209991455, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7756, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.490694761276245, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.7372, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.44179368019104, + "learning_rate": 4.532900081234769e-06, + "loss": 1.7652, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.934715986251831, + "eval_runtime": 44.688, + "eval_samples_per_second": 22.377, + "eval_steps_per_second": 0.94, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 3.018480062484741, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7596, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.5316643714904785, + "learning_rate": 4.528838342810723e-06, + "loss": 1.7715, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4893109798431396, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7549, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.524834156036377, + "learning_rate": 4.524776604386678e-06, + "loss": 1.8046, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7527239322662354, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7583, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.5210752487182617, + "learning_rate": 4.520714865962632e-06, + "loss": 1.7896, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.5787770748138428, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7355, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.537195920944214, + "learning_rate": 4.516653127538587e-06, + "loss": 1.7272, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.6810014247894287, + "learning_rate": 4.514622258326564e-06, + "loss": 1.753, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.5868756771087646, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7964, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.934499740600586, + "eval_runtime": 45.6556, + "eval_samples_per_second": 21.903, + "eval_steps_per_second": 0.92, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.6830856800079346, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7072, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.6574594974517822, + "learning_rate": 4.508529650690496e-06, + "loss": 1.7004, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.4415769577026367, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7667, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.825115442276001, + "learning_rate": 4.504467912266451e-06, + "loss": 1.8356, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.949010133743286, + "learning_rate": 4.502437043054428e-06, + "loss": 1.7579, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.690986394882202, + "learning_rate": 4.500406173842405e-06, + "loss": 1.7657, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 3.1015520095825195, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.7432, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.608424186706543, + "learning_rate": 4.49634443541836e-06, + "loss": 1.8112, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.5238773822784424, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7724, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.7712161540985107, + "learning_rate": 4.492282696994314e-06, + "loss": 1.7111, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9326083660125732, + "eval_runtime": 48.5359, + "eval_samples_per_second": 20.603, + "eval_steps_per_second": 0.865, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.4763033390045166, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7268, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.6310081481933594, + "learning_rate": 4.488220958570269e-06, + "loss": 1.6885, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.5950350761413574, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7775, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.769026517868042, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7737, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.4898502826690674, + "learning_rate": 4.4821283509342e-06, + "loss": 1.751, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.668081521987915, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7178, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.441615104675293, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7511, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.727665901184082, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.8639, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.7964634895324707, + "learning_rate": 4.474004874086109e-06, + "loss": 1.6945, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.6142687797546387, + "learning_rate": 4.471974004874087e-06, + "loss": 1.8028, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9281476736068726, + "eval_runtime": 48.078, + "eval_samples_per_second": 20.8, + "eval_steps_per_second": 0.874, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.6655778884887695, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.7632, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.931162118911743, + "learning_rate": 4.467912266450041e-06, + "loss": 1.7244, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.469895124435425, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7773, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.4948830604553223, + "learning_rate": 4.463850528025995e-06, + "loss": 1.788, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5961945056915283, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.756, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.452946424484253, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6968, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.9352009296417236, + "learning_rate": 4.457757920389927e-06, + "loss": 1.7414, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.7044763565063477, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7848, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.7200324535369873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7699, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.6185197830200195, + "learning_rate": 4.451665312753859e-06, + "loss": 1.7816, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9209110736846924, + "eval_runtime": 47.3321, + "eval_samples_per_second": 21.127, + "eval_steps_per_second": 0.887, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.3639564514160156, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7164, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7073519229888916, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.727, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.637080430984497, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7951, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.8013885021209717, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7333, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.512484312057495, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8048, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.3851466178894043, + "learning_rate": 4.439480097481723e-06, + "loss": 1.6667, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.8105392456054688, + "learning_rate": 4.4374492282697e-06, + "loss": 1.7622, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.555773973464966, + "learning_rate": 4.435418359057677e-06, + "loss": 1.7519, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.688596725463867, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7198, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.828646421432495, + "learning_rate": 4.431356620633632e-06, + "loss": 1.7478, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.919256567955017, + "eval_runtime": 50.5993, + "eval_samples_per_second": 19.763, + "eval_steps_per_second": 0.83, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.4523651599884033, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7344, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.4237422943115234, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6664, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.2746193408966064, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.7446, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.502819061279297, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7691, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.7032017707824707, + "learning_rate": 4.421202274573518e-06, + "loss": 1.6448, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.5879440307617188, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6908, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 3.2481982707977295, + "learning_rate": 4.417140536149473e-06, + "loss": 1.7725, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.4844820499420166, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7141, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.5596253871917725, + "learning_rate": 4.413078797725427e-06, + "loss": 1.7485, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 3.142839193344116, + "learning_rate": 4.411047928513404e-06, + "loss": 1.734, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9129430055618286, + "eval_runtime": 49.2608, + "eval_samples_per_second": 20.3, + "eval_steps_per_second": 0.853, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.5837552547454834, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7668, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.7043633460998535, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7497, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.9886765480041504, + "learning_rate": 4.404955320877336e-06, + "loss": 1.7082, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.3199501037597656, + "learning_rate": 4.402924451665313e-06, + "loss": 1.693, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.606682777404785, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7242, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.6967978477478027, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7516, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.351726531982422, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7338, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.590869188308716, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7462, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.3498220443725586, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7609, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.281002998352051, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7584, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.908841609954834, + "eval_runtime": 50.7771, + "eval_samples_per_second": 19.694, + "eval_steps_per_second": 0.827, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.468755006790161, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7654, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.3467307090759277, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7923, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.3723301887512207, + "learning_rate": 4.384646628757108e-06, + "loss": 1.74, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.4527792930603027, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6988, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.3923704624176025, + "learning_rate": 4.380584890333063e-06, + "loss": 1.6406, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.740556240081787, + "learning_rate": 4.37855402112104e-06, + "loss": 1.8032, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.3013832569122314, + "learning_rate": 4.376523151909017e-06, + "loss": 1.7259, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.4632391929626465, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.7462, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.4878177642822266, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7665, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.610898494720459, + "learning_rate": 4.370430544272949e-06, + "loss": 1.7646, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9056707620620728, + "eval_runtime": 51.8738, + "eval_samples_per_second": 19.278, + "eval_steps_per_second": 0.81, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.387721300125122, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7454, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.9966042041778564, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6789, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.664689064025879, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7486, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.4788103103637695, + "learning_rate": 4.362307067424858e-06, + "loss": 1.6981, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.4413328170776367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7279, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.6818156242370605, + "learning_rate": 4.358245329000813e-06, + "loss": 1.74, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.650348424911499, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7212, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.442805767059326, + "learning_rate": 4.354183590576767e-06, + "loss": 1.6951, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.6806585788726807, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7499, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.586792230606079, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7709, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.901674747467041, + "eval_runtime": 42.3186, + "eval_samples_per_second": 23.63, + "eval_steps_per_second": 0.992, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.547452688217163, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7458, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.5820348262786865, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.7232, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.4208321571350098, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7981, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.8791956901550293, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7372, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6537868976593018, + "learning_rate": 4.339967506092607e-06, + "loss": 1.7216, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.416097402572632, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.701, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.5565483570098877, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7558, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.5875823497772217, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7439, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5459702014923096, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.7058, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.4195566177368164, + "learning_rate": 4.329813160032495e-06, + "loss": 1.6749, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9035066366195679, + "eval_runtime": 43.7962, + "eval_samples_per_second": 22.833, + "eval_steps_per_second": 0.959, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6019818782806396, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7549, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.8627779483795166, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7257, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.582158088684082, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7138, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.4528613090515137, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7162, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 3.1956026554107666, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7556, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.3531901836395264, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.746, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.441988706588745, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7664, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.599506378173828, + "learning_rate": 4.313566206336312e-06, + "loss": 1.6955, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.4121932983398438, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7194, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.4803290367126465, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7386, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8993914127349854, + "eval_runtime": 43.0525, + "eval_samples_per_second": 23.227, + "eval_steps_per_second": 0.976, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.701594114303589, + "learning_rate": 4.307473598700244e-06, + "loss": 1.6796, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7144737243652344, + "learning_rate": 4.305442729488221e-06, + "loss": 1.763, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.395312786102295, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7522, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.3926732540130615, + "learning_rate": 4.301380991064176e-06, + "loss": 1.7705, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.6293702125549316, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7693, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.3746306896209717, + "learning_rate": 4.29731925264013e-06, + "loss": 1.72, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.5421926975250244, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6868, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.6991186141967773, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7378, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.463897705078125, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7628, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5139036178588867, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7125, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8909698724746704, + "eval_runtime": 42.5917, + "eval_samples_per_second": 23.479, + "eval_steps_per_second": 0.986, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.5220465660095215, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6975, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.432988405227661, + "learning_rate": 4.285134037367994e-06, + "loss": 1.6904, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.5894253253936768, + "learning_rate": 4.283103168155971e-06, + "loss": 1.6919, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5334458351135254, + "learning_rate": 4.281072298943948e-06, + "loss": 1.6968, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.510019302368164, + "learning_rate": 4.279041429731926e-06, + "loss": 1.6791, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.531538248062134, + "learning_rate": 4.277010560519903e-06, + "loss": 1.723, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.5621440410614014, + "learning_rate": 4.27497969130788e-06, + "loss": 1.749, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.594329833984375, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7579, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.321031093597412, + "learning_rate": 4.270917952883835e-06, + "loss": 1.6983, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.6325764656066895, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7872, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.886696219444275, + "eval_runtime": 43.4143, + "eval_samples_per_second": 23.034, + "eval_steps_per_second": 0.967, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5342679023742676, + "learning_rate": 4.266856214459789e-06, + "loss": 1.6655, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.5810720920562744, + "learning_rate": 4.264825345247766e-06, + "loss": 1.7182, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.7768406867980957, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6534, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.5680463314056396, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7243, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7162973880767822, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.7146, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.697925090789795, + "learning_rate": 4.256701868399675e-06, + "loss": 1.7296, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.4458425045013428, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7466, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6330654621124268, + "learning_rate": 4.252640129975629e-06, + "loss": 1.6949, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.350430727005005, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6574, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.505995512008667, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7111, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.8887447118759155, + "eval_runtime": 43.9304, + "eval_samples_per_second": 22.763, + "eval_steps_per_second": 0.956, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 3.036954879760742, + "learning_rate": 4.246547522339562e-06, + "loss": 1.7182, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.563606023788452, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.6566, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.516315460205078, + "learning_rate": 4.242485783915517e-06, + "loss": 1.6543, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.6289877891540527, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6996, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.7563092708587646, + "learning_rate": 4.23842404549147e-06, + "loss": 1.666, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5743000507354736, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7338, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 3.1975982189178467, + "learning_rate": 4.234362307067425e-06, + "loss": 1.7444, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.6022894382476807, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7367, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.4466192722320557, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.726, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.579015016555786, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7146, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.8904188871383667, + "eval_runtime": 43.8091, + "eval_samples_per_second": 22.826, + "eval_steps_per_second": 0.959, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.719646453857422, + "learning_rate": 4.226238830219334e-06, + "loss": 1.7209, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.4423880577087402, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7288, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5883753299713135, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7089, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.447049140930176, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7095, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.453739881515503, + "learning_rate": 4.218115353371243e-06, + "loss": 1.6747, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4540836811065674, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7578, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.5468671321868896, + "learning_rate": 4.214053614947198e-06, + "loss": 1.6556, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.443814277648926, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7307, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.4998748302459717, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6768, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.667025566101074, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7493, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.8830010890960693, + "eval_runtime": 44.7027, + "eval_samples_per_second": 22.37, + "eval_steps_per_second": 0.94, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.416248321533203, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7381, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.378025770187378, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7335, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.633173704147339, + "learning_rate": 4.201868399675061e-06, + "loss": 1.6484, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.549384117126465, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7143, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.6492598056793213, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6907, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.444086790084839, + "learning_rate": 4.195775792038993e-06, + "loss": 1.636, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.628814935684204, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7342, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.83831524848938, + "learning_rate": 4.191714053614948e-06, + "loss": 1.7028, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.467339515686035, + "learning_rate": 4.189683184402924e-06, + "loss": 1.749, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.4529521465301514, + "learning_rate": 4.187652315190902e-06, + "loss": 1.682, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.8826342821121216, + "eval_runtime": 44.7757, + "eval_samples_per_second": 22.334, + "eval_steps_per_second": 0.938, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7340762615203857, + "learning_rate": 4.185621445978879e-06, + "loss": 1.723, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.6028647422790527, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7566, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.3967695236206055, + "learning_rate": 4.181559707554833e-06, + "loss": 1.683, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5475900173187256, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6976, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.5411484241485596, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7489, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 3.458749294281006, + "learning_rate": 4.175467099918766e-06, + "loss": 1.6828, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.2977802753448486, + "learning_rate": 4.173436230706742e-06, + "loss": 1.6963, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.875307559967041, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.6857, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 4.097696304321289, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7225, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.6774847507476807, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7031, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.8783056735992432, + "eval_runtime": 44.0219, + "eval_samples_per_second": 22.716, + "eval_steps_per_second": 0.954, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.6787376403808594, + "learning_rate": 4.165312753858652e-06, + "loss": 1.6865, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.3803224563598633, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7237, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.9285683631896973, + "learning_rate": 4.161251015434607e-06, + "loss": 1.6575, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.5534448623657227, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6799, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4105820655822754, + "learning_rate": 4.157189277010561e-06, + "loss": 1.7128, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.6571805477142334, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7123, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.891935348510742, + "learning_rate": 4.153127538586516e-06, + "loss": 1.6949, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.8692519664764404, + "learning_rate": 4.151096669374492e-06, + "loss": 1.699, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.780078649520874, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7114, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.4797518253326416, + "learning_rate": 4.147034930950447e-06, + "loss": 1.6943, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.8726727962493896, + "eval_runtime": 43.5664, + "eval_samples_per_second": 22.953, + "eval_steps_per_second": 0.964, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.7010464668273926, + "learning_rate": 4.145004061738425e-06, + "loss": 1.6207, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8812713623046875, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.6608, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.6700279712677, + "learning_rate": 4.140942323314379e-06, + "loss": 1.677, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.796513557434082, + "learning_rate": 4.138911454102356e-06, + "loss": 1.7291, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.4274256229400635, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7058, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5528790950775146, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7064, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.4288926124572754, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7163, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.4939630031585693, + "learning_rate": 4.130787977254265e-06, + "loss": 1.721, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7909305095672607, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7777, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.3382463455200195, + "learning_rate": 4.12672623883022e-06, + "loss": 1.6975, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.8738009929656982, + "eval_runtime": 44.8142, + "eval_samples_per_second": 22.314, + "eval_steps_per_second": 0.937, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6876718997955322, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6851, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.450212240219116, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7209, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.4365246295928955, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.6545, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.3236255645751953, + "learning_rate": 4.118602761982129e-06, + "loss": 1.6434, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.4753000736236572, + "learning_rate": 4.116571892770106e-06, + "loss": 1.7471, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.84885835647583, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6777, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.4361016750335693, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7079, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.5868160724639893, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7288, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.6499862670898438, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7191, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.333289384841919, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5938, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.8746954202651978, + "eval_runtime": 46.8039, + "eval_samples_per_second": 21.366, + "eval_steps_per_second": 0.897, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.1299779415130615, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7585, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 3.6410014629364014, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6803, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.801745653152466, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6957, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 3.0350706577301025, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7031, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 3.5306341648101807, + "learning_rate": 4.096263200649879e-06, + "loss": 1.6598, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.591485023498535, + "learning_rate": 4.094232331437856e-06, + "loss": 1.6659, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 3.0916528701782227, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6951, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.459009885787964, + "learning_rate": 4.090170593013811e-06, + "loss": 1.6597, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.932072162628174, + "learning_rate": 4.088139723801787e-06, + "loss": 1.6698, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.773566484451294, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6824, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.8707821369171143, + "eval_runtime": 46.7886, + "eval_samples_per_second": 21.373, + "eval_steps_per_second": 0.898, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6008529663085938, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7352, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.502838134765625, + "learning_rate": 4.08204711616572e-06, + "loss": 1.6342, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 3.081169605255127, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7307, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.841200828552246, + "learning_rate": 4.077985377741674e-06, + "loss": 1.7401, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.5010390281677246, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7252, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.468873977661133, + "learning_rate": 4.073923639317629e-06, + "loss": 1.6627, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.829737663269043, + "learning_rate": 4.071892770105605e-06, + "loss": 1.6419, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.8112833499908447, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6629, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6793017387390137, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6667, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.327993869781494, + "learning_rate": 4.065800162469538e-06, + "loss": 1.71, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.874659538269043, + "eval_runtime": 49.0634, + "eval_samples_per_second": 20.382, + "eval_steps_per_second": 0.856, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.689023971557617, + "learning_rate": 4.063769293257514e-06, + "loss": 1.6881, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.612842082977295, + "learning_rate": 4.061738424045492e-06, + "loss": 1.6971, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.319627046585083, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7469, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.588916540145874, + "learning_rate": 4.057676685621446e-06, + "loss": 1.707, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.270096778869629, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6284, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.4404375553131104, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7007, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5532679557800293, + "learning_rate": 4.051584077985378e-06, + "loss": 1.72, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.5105185508728027, + "learning_rate": 4.049553208773355e-06, + "loss": 1.6772, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.577929973602295, + "learning_rate": 4.047522339561333e-06, + "loss": 1.6747, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.8271372318267822, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7951, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.8684895038604736, + "eval_runtime": 48.0288, + "eval_samples_per_second": 20.821, + "eval_steps_per_second": 0.874, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.3306479454040527, + "learning_rate": 4.043460601137287e-06, + "loss": 1.6791, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.5543673038482666, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6761, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.3913605213165283, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7579, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.3837356567382812, + "learning_rate": 4.037367993501219e-06, + "loss": 1.6636, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.6936402320861816, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6488, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.6400210857391357, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.6926, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.574533462524414, + "learning_rate": 4.031275385865151e-06, + "loss": 1.7066, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5130677223205566, + "learning_rate": 4.029244516653128e-06, + "loss": 1.6431, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.578166961669922, + "learning_rate": 4.027213647441105e-06, + "loss": 1.648, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.4736695289611816, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6822, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.8754831552505493, + "eval_runtime": 50.0971, + "eval_samples_per_second": 19.961, + "eval_steps_per_second": 0.838, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.544315814971924, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6853, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8765335083007812, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7587, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.4749417304992676, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.6536, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.493164300918579, + "learning_rate": 4.017059301380992e-06, + "loss": 1.7436, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.6274492740631104, + "learning_rate": 4.015028432168969e-06, + "loss": 1.632, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.4162979125976562, + "learning_rate": 4.012997562956945e-06, + "loss": 1.6663, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5683343410491943, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6771, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.3710429668426514, + "learning_rate": 4.0089358245329e-06, + "loss": 1.6644, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.5356292724609375, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7339, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.480884075164795, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.716, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.867702841758728, + "eval_runtime": 48.0038, + "eval_samples_per_second": 20.832, + "eval_steps_per_second": 0.875, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4768476486206055, + "learning_rate": 4.002843216896833e-06, + "loss": 1.6667, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5727357864379883, + "learning_rate": 4.000812347684809e-06, + "loss": 1.6786, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6243903636932373, + "learning_rate": 3.998781478472787e-06, + "loss": 1.6328, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.712955951690674, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7852, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.756272554397583, + "learning_rate": 3.994719740048742e-06, + "loss": 1.7006, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.919201135635376, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6624, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.471566915512085, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7122, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.46273136138916, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6621, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.5338189601898193, + "learning_rate": 3.98659626320065e-06, + "loss": 1.6869, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.3850855827331543, + "learning_rate": 3.984565393988627e-06, + "loss": 1.6925, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.8650325536727905, + "eval_runtime": 49.4167, + "eval_samples_per_second": 20.236, + "eval_steps_per_second": 0.85, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.654726028442383, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7113, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.5065038204193115, + "learning_rate": 3.980503655564582e-06, + "loss": 1.7078, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.51668381690979, + "learning_rate": 3.978472786352559e-06, + "loss": 1.6742, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4586126804351807, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7339, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.452857732772827, + "learning_rate": 3.974411047928514e-06, + "loss": 1.6993, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.608631134033203, + "learning_rate": 3.972380178716491e-06, + "loss": 1.7047, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.34022855758667, + "learning_rate": 3.970349309504468e-06, + "loss": 1.6423, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.329612970352173, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.6443, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.3301169872283936, + "learning_rate": 3.966287571080423e-06, + "loss": 1.6937, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.3010976314544678, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6896, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.863387942314148, + "eval_runtime": 47.6761, + "eval_samples_per_second": 20.975, + "eval_steps_per_second": 0.881, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.296764612197876, + "learning_rate": 3.962225832656377e-06, + "loss": 1.6098, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.437114953994751, + "learning_rate": 3.960194963444355e-06, + "loss": 1.6727, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4908952713012695, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6816, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.81123685836792, + "learning_rate": 3.956133225020309e-06, + "loss": 1.6795, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.7088773250579834, + "learning_rate": 3.954102355808286e-06, + "loss": 1.6838, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.5021824836730957, + "learning_rate": 3.952071486596264e-06, + "loss": 1.671, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.3021023273468018, + "learning_rate": 3.950040617384241e-06, + "loss": 1.651, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.866119861602783, + "learning_rate": 3.948009748172218e-06, + "loss": 1.7117, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.3471503257751465, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.6883, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4612059593200684, + "learning_rate": 3.943948009748173e-06, + "loss": 1.6898, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.8587063550949097, + "eval_runtime": 43.4462, + "eval_samples_per_second": 23.017, + "eval_steps_per_second": 0.967, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.376680374145508, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7253, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.4701383113861084, + "learning_rate": 3.939886271324127e-06, + "loss": 1.6306, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.574384927749634, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6287, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.8596720695495605, + "learning_rate": 3.935824532900082e-06, + "loss": 1.7031, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.7144267559051514, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7219, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.5807480812072754, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.6364, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 3.3094675540924072, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6526, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.390503168106079, + "learning_rate": 3.927701056051991e-06, + "loss": 1.6795, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.5525588989257812, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.6432, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.522307872772217, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6218, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.8604834079742432, + "eval_runtime": 46.6465, + "eval_samples_per_second": 21.438, + "eval_steps_per_second": 0.9, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.438201904296875, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6788, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.639164686203003, + "learning_rate": 3.9195775792039e-06, + "loss": 1.6552, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.608038902282715, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.69, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.4357974529266357, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7031, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.4964048862457275, + "learning_rate": 3.913484971567831e-06, + "loss": 1.6126, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.5115931034088135, + "learning_rate": 3.911454102355808e-06, + "loss": 1.684, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.54923152923584, + "learning_rate": 3.909423233143786e-06, + "loss": 1.6815, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.849602222442627, + "learning_rate": 3.907392363931763e-06, + "loss": 1.744, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.7209551334381104, + "learning_rate": 3.90536149471974e-06, + "loss": 1.755, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.602299451828003, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.7835, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.8590067625045776, + "eval_runtime": 49.5182, + "eval_samples_per_second": 20.195, + "eval_steps_per_second": 0.848, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.504567861557007, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6595, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.456721305847168, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6692, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.319553852081299, + "learning_rate": 3.897238017871649e-06, + "loss": 1.6664, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.45987606048584, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.65, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 3.0049448013305664, + "learning_rate": 3.893176279447604e-06, + "loss": 1.7366, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.522472858428955, + "learning_rate": 3.891145410235581e-06, + "loss": 1.7022, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.3914318084716797, + "learning_rate": 3.889114541023558e-06, + "loss": 1.6591, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.462167501449585, + "learning_rate": 3.887083671811536e-06, + "loss": 1.6158, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.6707093715667725, + "learning_rate": 3.885052802599513e-06, + "loss": 1.7453, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.493745803833008, + "learning_rate": 3.88302193338749e-06, + "loss": 1.6836, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.8536967039108276, + "eval_runtime": 50.3456, + "eval_samples_per_second": 19.863, + "eval_steps_per_second": 0.834, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.5338919162750244, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.65, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.6954877376556396, + "learning_rate": 3.878960194963445e-06, + "loss": 1.7156, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.4878089427948, + "learning_rate": 3.876929325751422e-06, + "loss": 1.6053, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.4329843521118164, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6642, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 3.118771553039551, + "learning_rate": 3.872867587327377e-06, + "loss": 1.6392, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4644217491149902, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6566, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.4643497467041016, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6654, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.7067277431488037, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.682, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.8748488426208496, + "learning_rate": 3.864744110479286e-06, + "loss": 1.6886, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.632155179977417, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7069, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8592487573623657, + "eval_runtime": 47.0773, + "eval_samples_per_second": 21.242, + "eval_steps_per_second": 0.892, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.5379488468170166, + "learning_rate": 3.86068237205524e-06, + "loss": 1.6695, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.571742057800293, + "learning_rate": 3.858651502843217e-06, + "loss": 1.6235, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.5355312824249268, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6528, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.715510606765747, + "learning_rate": 3.854589764419171e-06, + "loss": 1.7379, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.635364532470703, + "learning_rate": 3.852558895207149e-06, + "loss": 1.6958, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.4713191986083984, + "learning_rate": 3.850528025995126e-06, + "loss": 1.6471, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.584200382232666, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6548, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.4639663696289062, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6614, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.4823923110961914, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.6523, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.953594923019409, + "learning_rate": 3.842404549147035e-06, + "loss": 1.6281, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8549022674560547, + "eval_runtime": 47.2215, + "eval_samples_per_second": 21.177, + "eval_steps_per_second": 0.889, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.524352550506592, + "learning_rate": 3.840373679935013e-06, + "loss": 1.6741, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.411515235900879, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.6571, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.6827003955841064, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6927, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 3.3486077785491943, + "learning_rate": 3.834281072298944e-06, + "loss": 1.6347, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.4518957138061523, + "learning_rate": 3.832250203086921e-06, + "loss": 1.7011, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.623842239379883, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7459, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.689602851867676, + "learning_rate": 3.828188464662876e-06, + "loss": 1.6812, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.8425960540771484, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6258, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.6639413833618164, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6769, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.4269535541534424, + "learning_rate": 3.822095857026808e-06, + "loss": 1.6363, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8489640951156616, + "eval_runtime": 45.897, + "eval_samples_per_second": 21.788, + "eval_steps_per_second": 0.915, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.3430261611938477, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6563, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.322094678878784, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6588, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.6172726154327393, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.681, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.574084997177124, + "learning_rate": 3.813972380178717e-06, + "loss": 1.6977, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.437417984008789, + "learning_rate": 3.811941510966694e-06, + "loss": 1.6204, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.319276809692383, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5858, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5540425777435303, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6622, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.567634344100952, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.635, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4815967082977295, + "learning_rate": 3.803818034118603e-06, + "loss": 1.7146, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.475162982940674, + "learning_rate": 3.80178716490658e-06, + "loss": 1.6911, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.848276138305664, + "eval_runtime": 44.1011, + "eval_samples_per_second": 22.675, + "eval_steps_per_second": 0.952, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5385165214538574, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.6523, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.443129301071167, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6239, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.696885347366333, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.7318, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.8570339679718018, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7198, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.475338935852051, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7052, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.5250487327575684, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.7015, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.812121868133545, + "learning_rate": 3.787571080422421e-06, + "loss": 1.6876, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.486117362976074, + "learning_rate": 3.785540211210398e-06, + "loss": 1.6695, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.5146725177764893, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.6585, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.5808019638061523, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.7608, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8411076068878174, + "eval_runtime": 48.0798, + "eval_samples_per_second": 20.799, + "eval_steps_per_second": 0.874, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.509993076324463, + "learning_rate": 3.77944760357433e-06, + "loss": 1.7101, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.492285966873169, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6417, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.809333086013794, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6963, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.524418354034424, + "learning_rate": 3.773354995938262e-06, + "loss": 1.6957, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.5848705768585205, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7193, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.8617639541625977, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.6721, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.4655680656433105, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.586, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 3.0633299350738525, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6792, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.3882107734680176, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.6229, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.5860483646392822, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6753, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8393676280975342, + "eval_runtime": 44.3838, + "eval_samples_per_second": 22.531, + "eval_steps_per_second": 0.946, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.930290460586548, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.6955, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.8278939723968506, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.6455, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.3822925090789795, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.7163, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.6904139518737793, + "learning_rate": 3.753046303818034e-06, + "loss": 1.6942, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.535284996032715, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7237, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.669031858444214, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.6391, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.5162742137908936, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6663, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.5937418937683105, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7115, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.4892170429229736, + "learning_rate": 3.742891957757921e-06, + "loss": 1.6727, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.7901015281677246, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7179, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8420112133026123, + "eval_runtime": 42.0904, + "eval_samples_per_second": 23.758, + "eval_steps_per_second": 0.998, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.510568618774414, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.6408, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.5084173679351807, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6495, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.3693647384643555, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.6064, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.520414352416992, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5668, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.544624090194702, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.6231, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.6257777214050293, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6093, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.0384273529052734, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.7291, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.718588352203369, + "learning_rate": 3.724614134849716e-06, + "loss": 1.604, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.655592918395996, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6993, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.569904088973999, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.6595, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8419173955917358, + "eval_runtime": 42.5535, + "eval_samples_per_second": 23.5, + "eval_steps_per_second": 0.987, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.643686056137085, + "learning_rate": 3.718521527213648e-06, + "loss": 1.6483, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.7431893348693848, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5418, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 3.0191054344177246, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.6425, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.695777654647827, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.651, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.512838840484619, + "learning_rate": 3.710398050365557e-06, + "loss": 1.6181, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.9957621097564697, + "learning_rate": 3.708367181153534e-06, + "loss": 1.6454, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.5258326530456543, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.5995, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.6610865592956543, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5816, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.825767993927002, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6402, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.605870008468628, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6282, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8433754444122314, + "eval_runtime": 45.2127, + "eval_samples_per_second": 22.118, + "eval_steps_per_second": 0.929, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.7334296703338623, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5958, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.9007229804992676, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.7128, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8827497959136963, + "learning_rate": 3.694151096669375e-06, + "loss": 1.6157, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6470541954040527, + "learning_rate": 3.692120227457352e-06, + "loss": 1.643, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.5725178718566895, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.5481, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.7779407501220703, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.6241, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 3.161653995513916, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.644, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.5995497703552246, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6475, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.607166290283203, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6573, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.4532339572906494, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.6044, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.8344279527664185, + "eval_runtime": 43.2125, + "eval_samples_per_second": 23.141, + "eval_steps_per_second": 0.972, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.540632486343384, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6308, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.40977144241333, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6134, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.6345880031585693, + "learning_rate": 3.673842404549147e-06, + "loss": 1.615, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.565157413482666, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6351, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.7868876457214355, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.7186, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.7061610221862793, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.6454, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.587785482406616, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6867, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.4618277549743652, + "learning_rate": 3.663688058489034e-06, + "loss": 1.6189, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.4425477981567383, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5772, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.7514469623565674, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6368, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8366374969482422, + "eval_runtime": 44.4679, + "eval_samples_per_second": 22.488, + "eval_steps_per_second": 0.945, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.529805898666382, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.6229, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6875312328338623, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6445, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.715970993041992, + "learning_rate": 3.65353371242892e-06, + "loss": 1.5318, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.8685660362243652, + "learning_rate": 3.651502843216897e-06, + "loss": 1.6465, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.6801602840423584, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.5959, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.61037015914917, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6561, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.478137493133545, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6435, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5708508491516113, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6206, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8857522010803223, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6269, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.6102395057678223, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.648, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.836978793144226, + "eval_runtime": 44.9452, + "eval_samples_per_second": 22.249, + "eval_steps_per_second": 0.934, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7924158573150635, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6802, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.784693479537964, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.6515, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.747894525527954, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6093, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.635850191116333, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6791, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.5760908126831055, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.5721, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5258584022521973, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5644, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.919623613357544, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.6258, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.8330483436584473, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6077, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8145744800567627, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6463, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.9187726974487305, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6348, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.837236762046814, + "eval_runtime": 45.4421, + "eval_samples_per_second": 22.006, + "eval_steps_per_second": 0.924, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.6741979122161865, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6441, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.556701898574829, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5523, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.6352901458740234, + "learning_rate": 3.612916328188465e-06, + "loss": 1.6545, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.5784106254577637, + "learning_rate": 3.610885458976442e-06, + "loss": 1.6649, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.7136101722717285, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.596, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.608877182006836, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.5923, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.783069372177124, + "learning_rate": 3.604792851340374e-06, + "loss": 1.6685, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.621882438659668, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6404, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.672128915786743, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5713, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.645395040512085, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.609, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8348923921585083, + "eval_runtime": 44.8379, + "eval_samples_per_second": 22.303, + "eval_steps_per_second": 0.937, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 3.610335111618042, + "learning_rate": 3.596669374492283e-06, + "loss": 1.6693, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.7391626834869385, + "learning_rate": 3.59463850528026e-06, + "loss": 1.5847, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.658418655395508, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6552, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.6508545875549316, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6201, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 4.237835884094238, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.6725, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.672144889831543, + "learning_rate": 3.586515028432169e-06, + "loss": 1.6666, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.619480609893799, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5671, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 3.150592565536499, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6118, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.946467161178589, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6499, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.6372971534729004, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5516, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8352437019348145, + "eval_runtime": 46.1704, + "eval_samples_per_second": 21.659, + "eval_steps_per_second": 0.91, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.6211962699890137, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6414, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.567579746246338, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.6632, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.4393808841705322, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6323, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.7005021572113037, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6016, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7400808334350586, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6327, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.922241449356079, + "learning_rate": 3.566206336311942e-06, + "loss": 1.5866, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.524975299835205, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5707, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.637805461883545, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6259, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7430527210235596, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.595, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.5548908710479736, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.625, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8332648277282715, + "eval_runtime": 48.5515, + "eval_samples_per_second": 20.597, + "eval_steps_per_second": 0.865, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.5396368503570557, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.655, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.523167848587036, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6171, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.6028409004211426, + "learning_rate": 3.551990251827783e-06, + "loss": 1.5872, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.607464075088501, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6176, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.510603427886963, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.5999, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.444693088531494, + "learning_rate": 3.545897644191714e-06, + "loss": 1.5791, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6724541187286377, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6358, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.64925217628479, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5793, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.647118091583252, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6399, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8507258892059326, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6046, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8305917978286743, + "eval_runtime": 43.2738, + "eval_samples_per_second": 23.109, + "eval_steps_per_second": 0.971, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.5572781562805176, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6395, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.530216693878174, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6493, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.564739465713501, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6075, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.5772671699523926, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.64, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.828692674636841, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6584, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.700568675994873, + "learning_rate": 3.525588952071487e-06, + "loss": 1.6595, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.666743755340576, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6483, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.551187515258789, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.5946, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.553602933883667, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.5997, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.5355968475341797, + "learning_rate": 3.517465475223396e-06, + "loss": 1.571, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8321614265441895, + "eval_runtime": 46.1746, + "eval_samples_per_second": 21.657, + "eval_steps_per_second": 0.91, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.707101345062256, + "learning_rate": 3.515434606011373e-06, + "loss": 1.6389, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.8770599365234375, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.6828, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.5841145515441895, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6264, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.697502851486206, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6081, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.620100259780884, + "learning_rate": 3.507311129163282e-06, + "loss": 1.5617, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.6503913402557373, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6194, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.830296754837036, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.616, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.6960535049438477, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5523, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.59023118019104, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6378, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.582906484603882, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6368, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8257222175598145, + "eval_runtime": 47.9573, + "eval_samples_per_second": 20.852, + "eval_steps_per_second": 0.876, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.6826765537261963, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5685, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6153650283813477, + "learning_rate": 3.493095044679123e-06, + "loss": 1.6948, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.607027769088745, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5197, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.493180990219116, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6159, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.6291017532348633, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5404, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.90544056892395, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6164, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.673076629638672, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6486, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.5843210220336914, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.6127, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.55414080619812, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6255, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.496095657348633, + "learning_rate": 3.476848090982941e-06, + "loss": 1.5941, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8262794017791748, + "eval_runtime": 47.0577, + "eval_samples_per_second": 21.25, + "eval_steps_per_second": 0.893, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.597395658493042, + "learning_rate": 3.474817221770918e-06, + "loss": 1.6941, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.4487929344177246, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6417, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.63041353225708, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.611, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.481630325317383, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.575, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.7625951766967773, + "learning_rate": 3.466693744922827e-06, + "loss": 1.5988, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.663846492767334, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6326, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.594536781311035, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6004, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.5871670246124268, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6378, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.591907262802124, + "learning_rate": 3.458570268074736e-06, + "loss": 1.6093, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.456110954284668, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6014, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8266254663467407, + "eval_runtime": 47.4255, + "eval_samples_per_second": 21.086, + "eval_steps_per_second": 0.886, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.8052752017974854, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.5679, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.7458856105804443, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6239, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.4829750061035156, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.5692, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.5427238941192627, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6214, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.597419261932373, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6012, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.5192158222198486, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6469, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.4233906269073486, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5822, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.4213144779205322, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5345, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.763439655303955, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6092, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.228456497192383, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6591, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8253848552703857, + "eval_runtime": 50.1479, + "eval_samples_per_second": 19.941, + "eval_steps_per_second": 0.838, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.670700788497925, + "learning_rate": 3.434199837530463e-06, + "loss": 1.612, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.797170877456665, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5624, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8155477046966553, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6535, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.7093656063079834, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6479, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.6133127212524414, + "learning_rate": 3.426076360682372e-06, + "loss": 1.6669, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.471862554550171, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6136, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7078938484191895, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5382, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.6206846237182617, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6186, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.674201011657715, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.6805, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.7543511390686035, + "learning_rate": 3.415922014622259e-06, + "loss": 1.644, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8275290727615356, + "eval_runtime": 49.1382, + "eval_samples_per_second": 20.351, + "eval_steps_per_second": 0.855, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.934906244277954, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6502, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.6866018772125244, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6324, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.743281126022339, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6229, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.7543251514434814, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.5813, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8038830757141113, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6482, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.6368162631988525, + "learning_rate": 3.403736799350122e-06, + "loss": 1.618, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7447240352630615, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.5735, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.6714727878570557, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.686, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.7939751148223877, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6104, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.624718189239502, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6194, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8199249505996704, + "eval_runtime": 50.4903, + "eval_samples_per_second": 19.806, + "eval_steps_per_second": 0.832, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.563082695007324, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5669, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.6971375942230225, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6237, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.7873635292053223, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6241, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.570742130279541, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6304, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8116018772125244, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6068, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.6097946166992188, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.5996, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.593461275100708, + "learning_rate": 3.381397238017872e-06, + "loss": 1.5935, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.008805274963379, + "learning_rate": 3.379366368805849e-06, + "loss": 1.68, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7567968368530273, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6129, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.647456645965576, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.618, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8231184482574463, + "eval_runtime": 46.6144, + "eval_samples_per_second": 21.453, + "eval_steps_per_second": 0.901, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.842937707901001, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6118, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.794092893600464, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.5922, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.837575674057007, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6267, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.7084648609161377, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5597, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.626157283782959, + "learning_rate": 3.36515028432169e-06, + "loss": 1.5961, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.589062213897705, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6235, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.681427478790283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6568, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 3.3202719688415527, + "learning_rate": 3.359057676685622e-06, + "loss": 1.649, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.463998794555664, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5482, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.5475661754608154, + "learning_rate": 3.354995938261576e-06, + "loss": 1.586, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8188257217407227, + "eval_runtime": 51.7434, + "eval_samples_per_second": 19.326, + "eval_steps_per_second": 0.812, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.579566478729248, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5546, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6778531074523926, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.5831, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.98976731300354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6199, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.8345465660095215, + "learning_rate": 3.346872461413485e-06, + "loss": 1.5697, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.6166720390319824, + "learning_rate": 3.344841592201463e-06, + "loss": 1.5829, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.5504395961761475, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.5845, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.6515390872955322, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6154, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.589524269104004, + "learning_rate": 3.338748984565394e-06, + "loss": 1.639, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.8798511028289795, + "learning_rate": 3.336718115353372e-06, + "loss": 1.676, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.984379529953003, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6154, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.818778157234192, + "eval_runtime": 44.8621, + "eval_samples_per_second": 22.291, + "eval_steps_per_second": 0.936, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.5935728549957275, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.5895, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.7658374309539795, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6041, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.580352306365967, + "learning_rate": 3.32859463850528e-06, + "loss": 1.6127, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.689892292022705, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5624, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.708078622817993, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.5782, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.7298295497894287, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.5815, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8700411319732666, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5597, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.6671183109283447, + "learning_rate": 3.318440292445167e-06, + "loss": 1.5902, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.11922550201416, + "learning_rate": 3.316409423233144e-06, + "loss": 1.5678, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.72869610786438, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6568, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8164011240005493, + "eval_runtime": 47.4229, + "eval_samples_per_second": 21.087, + "eval_steps_per_second": 0.886, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.7037997245788574, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6191, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.6406683921813965, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6018, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.4749815464019775, + "learning_rate": 3.308285946385053e-06, + "loss": 1.679, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.5094757080078125, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5455, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.6685080528259277, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6123, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7547595500946045, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.6497, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.7807092666625977, + "learning_rate": 3.300162469536962e-06, + "loss": 1.5903, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.903127431869507, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6398, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.4666707515716553, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6194, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.9508626461029053, + "learning_rate": 3.294069861900894e-06, + "loss": 1.6774, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8119163513183594, + "eval_runtime": 51.0141, + "eval_samples_per_second": 19.602, + "eval_steps_per_second": 0.823, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.585411787033081, + "learning_rate": 3.292038992688871e-06, + "loss": 1.6086, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.7636921405792236, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.6386, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.72711181640625, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.6563, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.5698368549346924, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5501, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8764665126800537, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6229, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.653674840927124, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5343, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.896165609359741, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.6617, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.714388847351074, + "learning_rate": 3.277822908204712e-06, + "loss": 1.605, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.572202682495117, + "learning_rate": 3.275792038992689e-06, + "loss": 1.568, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.719407081604004, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6116, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.811676263809204, + "eval_runtime": 46.0392, + "eval_samples_per_second": 21.721, + "eval_steps_per_second": 0.912, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.752934694290161, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.5672, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.519256353378296, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6101, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.758317708969116, + "learning_rate": 3.267668562144598e-06, + "loss": 1.597, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.6348824501037598, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.5731, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.699477434158325, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6288, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.5104498863220215, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6024, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.680694341659546, + "learning_rate": 3.259545085296507e-06, + "loss": 1.6689, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.459789276123047, + "learning_rate": 3.257514216084484e-06, + "loss": 1.56, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.533125877380371, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.6734, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.640742301940918, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6214, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8131158351898193, + "eval_runtime": 47.2273, + "eval_samples_per_second": 21.174, + "eval_steps_per_second": 0.889, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.4593327045440674, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.5997, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.6157352924346924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6193, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.5011303424835205, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.5796, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.545781373977661, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.6128, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.5882468223571777, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6106, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.7500388622283936, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5491, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5432305335998535, + "learning_rate": 3.23923639317628e-06, + "loss": 1.5879, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.5674941539764404, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.5737, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.630491256713867, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6167, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.7517776489257812, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.6698, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8172497749328613, + "eval_runtime": 45.7554, + "eval_samples_per_second": 21.855, + "eval_steps_per_second": 0.918, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5155255794525146, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5647, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.5586776733398438, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6403, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7582011222839355, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5901, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.611339569091797, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.599, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.7254250049591064, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6167, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.5621025562286377, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5241, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849069118499756, + "learning_rate": 3.218927701056052e-06, + "loss": 1.5823, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.874762535095215, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5294, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.706822395324707, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.6529, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.7066354751586914, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5591, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.814163327217102, + "eval_runtime": 43.7518, + "eval_samples_per_second": 22.856, + "eval_steps_per_second": 0.96, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.584439516067505, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.5986, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.671645402908325, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6417, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6757681369781494, + "learning_rate": 3.206742485783916e-06, + "loss": 1.5594, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.6210391521453857, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5476, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.5632853507995605, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6165, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.64302921295166, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.631, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.4642136096954346, + "learning_rate": 3.198619008935825e-06, + "loss": 1.534, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.196588139723802e-06, + "loss": 1.598, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.5753607749938965, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6177, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.787393093109131, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.607, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.809692621231079, + "eval_runtime": 44.1276, + "eval_samples_per_second": 22.662, + "eval_steps_per_second": 0.952, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.4849026203155518, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6064, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.484736204147339, + "learning_rate": 3.188464662875711e-06, + "loss": 1.6591, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.652766466140747, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5931, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.6450769901275635, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5463, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.736511468887329, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5509, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.8212637901306152, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6244, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.7004551887512207, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6123, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.665766954421997, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.5904, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.7625415325164795, + "learning_rate": 3.174248578391552e-06, + "loss": 1.5843, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.5376832485198975, + "learning_rate": 3.172217709179529e-06, + "loss": 1.5758, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8079267740249634, + "eval_runtime": 45.139, + "eval_samples_per_second": 22.154, + "eval_steps_per_second": 0.93, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.704226016998291, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.5891, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6165294647216797, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5502, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.6848294734954834, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6134, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.778846502304077, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.611, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.647531747817993, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5531, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.605592966079712, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6011, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5820205211639404, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6121, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.7092363834381104, + "learning_rate": 3.155970755483347e-06, + "loss": 1.5928, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.7473032474517822, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6068, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.742840528488159, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6041, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8063452243804932, + "eval_runtime": 44.2112, + "eval_samples_per_second": 22.619, + "eval_steps_per_second": 0.95, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.7613816261291504, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6413, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.876913547515869, + "learning_rate": 3.147847278635256e-06, + "loss": 1.6848, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5945441722869873, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.5847, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.526456117630005, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5496, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.9870903491973877, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.5982, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6774742603302, + "learning_rate": 3.139723801787165e-06, + "loss": 1.5678, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.69058895111084, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6174, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.8981339931488037, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5529, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.522247552871704, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6061, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.709726333618164, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6267, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8052302598953247, + "eval_runtime": 42.6337, + "eval_samples_per_second": 23.456, + "eval_steps_per_second": 0.985, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.618793249130249, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6441, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5927956104278564, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6035, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.557023048400879, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6252, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5910511016845703, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.5883, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.6681272983551025, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6147, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.7024455070495605, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6317, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.6808700561523438, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.5871, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.585388422012329, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.6202, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.637092113494873, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6357, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.652611494064331, + "learning_rate": 3.111291632818847e-06, + "loss": 1.651, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.804667592048645, + "eval_runtime": 43.1932, + "eval_samples_per_second": 23.152, + "eval_steps_per_second": 0.972, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.770113945007324, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6005, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.570833206176758, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6169, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.537957191467285, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.5826, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.715669631958008, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6276, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.5450210571289062, + "learning_rate": 3.101137286758733e-06, + "loss": 1.6305, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.6358814239501953, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6367, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.7380471229553223, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.5984, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 3.030034303665161, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.6661, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.6777493953704834, + "learning_rate": 3.093013809910642e-06, + "loss": 1.6122, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8807458877563477, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.6514, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8064097166061401, + "eval_runtime": 43.3673, + "eval_samples_per_second": 23.059, + "eval_steps_per_second": 0.968, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.545091390609741, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6182, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.6389622688293457, + "learning_rate": 3.086921202274574e-06, + "loss": 1.5931, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.5784780979156494, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5579, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.67499041557312, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6128, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.558598756790161, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.5958, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6460611820220947, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5274, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.770334482192993, + "learning_rate": 3.07676685621446e-06, + "loss": 1.604, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6441917419433594, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.567, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.6142544746398926, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.5891, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.6242077350616455, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6059, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8036692142486572, + "eval_runtime": 45.6003, + "eval_samples_per_second": 21.93, + "eval_steps_per_second": 0.921, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.417332410812378, + "learning_rate": 3.068643379366369e-06, + "loss": 1.5991, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.4537513256073, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5603, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.745912551879883, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6171, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.636627674102783, + "learning_rate": 3.062550771730301e-06, + "loss": 1.5839, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.6001460552215576, + "learning_rate": 3.060519902518278e-06, + "loss": 1.5781, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.572185516357422, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5188, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.87540864944458, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.621, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.7670910358428955, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.584, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.672682523727417, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6293, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.655538558959961, + "learning_rate": 3.050365556458164e-06, + "loss": 1.5773, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8046640157699585, + "eval_runtime": 44.6243, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.582319974899292, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.5751, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.7160403728485107, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.5972, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.6800436973571777, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6311, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.8212592601776123, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6315, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.57470703125, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5414, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 3.3979809284210205, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.5732, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.606290102005005, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.5966, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.5241527557373047, + "learning_rate": 3.034118602761982e-06, + "loss": 1.588, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.7115085124969482, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5619, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.7072222232818604, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6346, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.804133415222168, + "eval_runtime": 44.5547, + "eval_samples_per_second": 22.444, + "eval_steps_per_second": 0.943, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.5874669551849365, + "learning_rate": 3.028025995125914e-06, + "loss": 1.623, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.644073247909546, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5484, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.553112506866455, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.5923, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6504719257354736, + "learning_rate": 3.021933387489846e-06, + "loss": 1.543, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.7608838081359863, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6001, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.690267562866211, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6093, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.8565986156463623, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5614, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.672799587249756, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5657, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5756146907806396, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.5664, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.8423633575439453, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.5936, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.800963044166565, + "eval_runtime": 46.5083, + "eval_samples_per_second": 21.502, + "eval_steps_per_second": 0.903, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.6263201236724854, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.5693, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 3.0006930828094482, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6084, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.7002620697021484, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.5667, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.722398519515991, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6285, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.738555908203125, + "learning_rate": 2.999593826157596e-06, + "loss": 1.5849, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.726271390914917, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.5951, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.6749048233032227, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.5807, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.774726629257202, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6355, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6740102767944336, + "learning_rate": 2.991470349309505e-06, + "loss": 1.5871, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.7571794986724854, + "learning_rate": 2.989439480097482e-06, + "loss": 1.578, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.7960126399993896, + "eval_runtime": 44.4777, + "eval_samples_per_second": 22.483, + "eval_steps_per_second": 0.944, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.67576265335083, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.5917, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.8302817344665527, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6313, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.874002456665039, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6212, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6992602348327637, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6011, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.726292371749878, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.5981, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.988934278488159, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6272, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.6781861782073975, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.5742, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.818401336669922, + "learning_rate": 2.9731925264013e-06, + "loss": 1.5713, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.6900908946990967, + "learning_rate": 2.971161657189277e-06, + "loss": 1.5754, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.9017207622528076, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.5746, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.7994096279144287, + "eval_runtime": 46.997, + "eval_samples_per_second": 21.278, + "eval_steps_per_second": 0.894, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.0305914878845215, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.5733, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.695298910140991, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6379, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.60552716255188, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5488, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.664668321609497, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6256, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.700955390930176, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6228, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.637998104095459, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.5565, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.664620876312256, + "learning_rate": 2.954914703493095e-06, + "loss": 1.5805, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.8120782375335693, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.5615, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.641434907913208, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.56, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.845808982849121, + "learning_rate": 2.948822095857027e-06, + "loss": 1.5954, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8004399538040161, + "eval_runtime": 47.1938, + "eval_samples_per_second": 21.189, + "eval_steps_per_second": 0.89, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.713960886001587, + "learning_rate": 2.946791226645004e-06, + "loss": 1.5628, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.8967084884643555, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.5445, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 3.3338470458984375, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.5735, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.864295721054077, + "learning_rate": 2.940698619008936e-06, + "loss": 1.5892, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.704193353652954, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.5916, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.6473755836486816, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.6059, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.610670804977417, + "learning_rate": 2.934606011372868e-06, + "loss": 1.6012, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.6023919582366943, + "learning_rate": 2.932575142160845e-06, + "loss": 1.5973, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8217804431915283, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.5794, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.8418939113616943, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.5942, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.8003290891647339, + "eval_runtime": 44.3696, + "eval_samples_per_second": 22.538, + "eval_steps_per_second": 0.947, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.6487412452697754, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.6297, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 2.6183269023895264, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.5916, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.6346395015716553, + "learning_rate": 2.922420796100731e-06, + "loss": 1.5209, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.631850242614746, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.6268, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.7087175846099854, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.5348, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.70473575592041, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.5943, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 3.0979015827178955, + "learning_rate": 2.91429731925264e-06, + "loss": 1.6308, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.6136019229888916, + "learning_rate": 2.912266450040618e-06, + "loss": 1.5955, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.6788594722747803, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.5909, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 3.681739330291748, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.5794, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.7974433898925781, + "eval_runtime": 50.5413, + "eval_samples_per_second": 19.786, + "eval_steps_per_second": 0.831, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.543687582015991, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.5455, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 3.0568184852600098, + "learning_rate": 2.904142973192527e-06, + "loss": 1.6445, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.557318925857544, + "learning_rate": 2.902112103980504e-06, + "loss": 1.5933, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.4658801555633545, + "learning_rate": 2.900081234768481e-06, + "loss": 1.5899, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.610834836959839, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.5557, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.643775224685669, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.5523, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.8988914489746094, + "learning_rate": 2.893988627132413e-06, + "loss": 1.5787, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.686321973800659, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.5205, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.6282148361206055, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.5942, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.581450939178467, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.5779, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.7954648733139038, + "eval_runtime": 45.0426, + "eval_samples_per_second": 22.201, + "eval_steps_per_second": 0.932, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.6617720127105713, + "learning_rate": 2.885865150284322e-06, + "loss": 1.6168, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.643260955810547, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.5764, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.5922341346740723, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.5385, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.7868714332580566, + "learning_rate": 2.879772542648254e-06, + "loss": 1.5905, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.6874098777770996, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.516, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.9882795810699463, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.6497, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.574932336807251, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.6217, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 2.625720739364624, + "learning_rate": 2.871649065800163e-06, + "loss": 1.6436, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 2.6322028636932373, + "learning_rate": 2.86961819658814e-06, + "loss": 1.5743, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.753891944885254, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.594, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.7947734594345093, + "eval_runtime": 47.0036, + "eval_samples_per_second": 21.275, + "eval_steps_per_second": 0.894, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.6878788471221924, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.5801, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.7409934997558594, + "learning_rate": 2.863525588952072e-06, + "loss": 1.6251, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.517815351486206, + "learning_rate": 2.861494719740049e-06, + "loss": 1.5055, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.7027196884155273, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.5536, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.8066604137420654, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.5819, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.7119011878967285, + "learning_rate": 2.855402112103981e-06, + "loss": 1.5776, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.6952199935913086, + "learning_rate": 2.853371242891958e-06, + "loss": 1.5682, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.7582387924194336, + "learning_rate": 2.851340373679935e-06, + "loss": 1.5863, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.8032495975494385, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.6204, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.6014182567596436, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.5816, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.7986392974853516, + "eval_runtime": 45.7419, + "eval_samples_per_second": 21.862, + "eval_steps_per_second": 0.918, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.611240863800049, + "learning_rate": 2.845247766043867e-06, + "loss": 1.5563, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.604238510131836, + "learning_rate": 2.843216896831844e-06, + "loss": 1.5143, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.531214475631714, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.5579, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.6469998359680176, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.6049, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 3.1636979579925537, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.6072, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.6377527713775635, + "learning_rate": 2.835093419983753e-06, + "loss": 1.574, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.6896755695343018, + "learning_rate": 2.833062550771731e-06, + "loss": 1.5498, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.6172642707824707, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.5413, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.7607314586639404, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.5816, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.5872104167938232, + "learning_rate": 2.826969943135662e-06, + "loss": 1.5868, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.7989377975463867, + "eval_runtime": 46.6112, + "eval_samples_per_second": 21.454, + "eval_steps_per_second": 0.901, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.517178773880005, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.5875, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.531451940536499, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.5583, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.6367409229278564, + "learning_rate": 2.820877335499594e-06, + "loss": 1.5605, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.496851682662964, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.5227, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.583904981613159, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.5434, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.7455286979675293, + "learning_rate": 2.814784727863526e-06, + "loss": 1.5732, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.6893272399902344, + "learning_rate": 2.812753858651503e-06, + "loss": 1.5807, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 2.6912310123443604, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.5484, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 2.6918725967407227, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.535, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 2.808300256729126, + "learning_rate": 2.806661251015435e-06, + "loss": 1.6278, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.8001086711883545, + "eval_runtime": 50.3556, + "eval_samples_per_second": 19.859, + "eval_steps_per_second": 0.834, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.702766180038452, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.6223, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 2.890214681625366, + "learning_rate": 2.802599512591389e-06, + "loss": 1.68, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.6849842071533203, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.5836, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.678279161453247, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.529, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 2.782949686050415, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.5814, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.8183817863464355, + "learning_rate": 2.794476035743298e-06, + "loss": 1.6401, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.917726516723633, + "learning_rate": 2.792445166531276e-06, + "loss": 1.5936, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 2.8861331939697266, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.6213, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.7103989124298096, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.5537, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.876448392868042, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.5464, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.7939362525939941, + "eval_runtime": 47.8308, + "eval_samples_per_second": 20.907, + "eval_steps_per_second": 0.878, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.613130569458008, + "learning_rate": 2.784321689683185e-06, + "loss": 1.5427, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.553765058517456, + "learning_rate": 2.782290820471162e-06, + "loss": 1.5496, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.7586376667022705, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.6221, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.6437292098999023, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.6075, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.6939804553985596, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.5918, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.6177403926849365, + "learning_rate": 2.774167343623071e-06, + "loss": 1.6353, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 2.8386614322662354, + "learning_rate": 2.772136474411048e-06, + "loss": 1.5548, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.6705095767974854, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.497, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.790919065475464, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.5945, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.7193119525909424, + "learning_rate": 2.76604386677498e-06, + "loss": 1.5627, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.795873999595642, + "eval_runtime": 49.9423, + "eval_samples_per_second": 20.023, + "eval_steps_per_second": 0.841, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 2.871537208557129, + "learning_rate": 2.764012997562957e-06, + "loss": 1.6027, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.6842691898345947, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.5327, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 2.6385788917541504, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.6181, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.6933207511901855, + "learning_rate": 2.757920389926889e-06, + "loss": 1.5561, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 2.814086437225342, + "learning_rate": 2.755889520714866e-06, + "loss": 1.5519, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.6112890243530273, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.52, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.7777466773986816, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.5885, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.782341241836548, + "learning_rate": 2.749796913078798e-06, + "loss": 1.609, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.576822519302368, + "learning_rate": 2.747766043866775e-06, + "loss": 1.5808, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.6026647090911865, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.626, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.7908697128295898, + "eval_runtime": 47.8943, + "eval_samples_per_second": 20.879, + "eval_steps_per_second": 0.877, + "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_local3_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-1600/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/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_local3_10M_seed0/runs/checkpoint-1600/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/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_local3_10M_seed0/runs/checkpoint-1600/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/latest new file mode 100644 index 0000000000000000000000000000000000000000..10b09fac99bc80ff931649e8b3378aab683b28be --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/latest @@ -0,0 +1 @@ +global_step1600 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/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_local3_10M_seed0/runs/checkpoint-1600/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/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_local3_10M_seed0/runs/checkpoint-1600/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/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_local3_10M_seed0/runs/checkpoint-1600/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..ea4bcfa8837effaa1d60090e355bc9beffbb0dc0 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/trainer_state.json @@ -0,0 +1,12513 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.7534246575342465, + "eval_steps": 10, + "global_step": 1600, + "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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.7659518718719482, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.279, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.433785915374756, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.3368, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 3.1486687660217285, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3896, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.391956090927124, + "learning_rate": 1.624087591240876e-06, + "loss": 2.297, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.6649584770202637, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.2415, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.5604891777038574, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2688, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.47900652885437, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2933, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.632723569869995, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.2143, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.8085947036743164, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2748, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.3927979469299316, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2685, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.355207920074463, + "eval_runtime": 42.3505, + "eval_samples_per_second": 23.612, + "eval_steps_per_second": 0.992, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.6170592308044434, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.3121, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.699814796447754, + "learning_rate": 1.77007299270073e-06, + "loss": 2.2853, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.3450610637664795, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3297, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.458219528198242, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2639, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.7101449966430664, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2134, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.5734996795654297, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2347, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 2.5200352668762207, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2299, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5022480487823486, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2253, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.190117359161377, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.2272, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.414524555206299, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2424, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.317634344100952, + "eval_runtime": 43.5742, + "eval_samples_per_second": 22.949, + "eval_steps_per_second": 0.964, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.3226804733276367, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1449, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.733006477355957, + "learning_rate": 1.952554744525548e-06, + "loss": 2.2027, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 2.5586233139038086, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2007, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.669914722442627, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2024, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.6148102283477783, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2496, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 2.331369638442993, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1814, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.3051886558532715, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2015, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.6890218257904053, + "learning_rate": 2.062043795620438e-06, + "loss": 2.184, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.4975883960723877, + "learning_rate": 2.08029197080292e-06, + "loss": 2.2439, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.5848891735076904, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1413, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.280966281890869, + "eval_runtime": 43.1001, + "eval_samples_per_second": 23.202, + "eval_steps_per_second": 0.974, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.2104, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.3356540203094482, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2312, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 2.3616602420806885, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.1399, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.581623077392578, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1456, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.547288656234741, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.184, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6386160850524902, + "learning_rate": 2.208029197080292e-06, + "loss": 2.1658, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.593065023422241, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1443, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.530632972717285, + "learning_rate": 2.244525547445256e-06, + "loss": 2.194, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.3553175926208496, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.1918, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 2.7536532878875732, + "learning_rate": 2.281021897810219e-06, + "loss": 2.1477, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2608063220977783, + "eval_runtime": 42.7938, + "eval_samples_per_second": 23.368, + "eval_steps_per_second": 0.981, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.160094738006592, + "learning_rate": 2.299270072992701e-06, + "loss": 2.133, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.4474735260009766, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1468, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.6654434204101562, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.1194, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.726611375808716, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1518, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 2.3560168743133545, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1537, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.664065361022949, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0888, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.3946099281311035, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.0998, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.7812652587890625, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1231, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.64595890045166, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1746, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.788524866104126, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0731, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.2340502738952637, + "eval_runtime": 42.3104, + "eval_samples_per_second": 23.635, + "eval_steps_per_second": 0.993, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 2.9249846935272217, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0564, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.5507798194885254, + "learning_rate": 2.5e-06, + "loss": 2.1714, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.1604716777801514, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0701, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.5429840087890625, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.0662, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8296737670898438, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.1378, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.9396755695343018, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1069, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.736814022064209, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1076, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 2.774909496307373, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1286, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.0861318111419678, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1499, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.624016284942627, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0989, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.215052604675293, + "eval_runtime": 43.9956, + "eval_samples_per_second": 22.73, + "eval_steps_per_second": 0.955, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.801500082015991, + "learning_rate": 2.664233576642336e-06, + "loss": 2.0795, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 2.2107114791870117, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.1035, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.030381441116333, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.0794, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 2.9719064235687256, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1496, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.688839912414551, + "learning_rate": 2.737226277372263e-06, + "loss": 2.1152, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 2.9350979328155518, + "learning_rate": 2.755474452554745e-06, + "loss": 2.0773, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.8958046436309814, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.1805, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.0781922340393066, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.1249, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.032059907913208, + "learning_rate": 2.81021897810219e-06, + "loss": 2.0632, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.798758029937744, + "learning_rate": 2.828467153284672e-06, + "loss": 2.1001, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2005105018615723, + "eval_runtime": 43.5111, + "eval_samples_per_second": 22.983, + "eval_steps_per_second": 0.965, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.7674927711486816, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.0211, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 2.6168575286865234, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0599, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 2.933353900909424, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.0169, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.81402587890625, + "learning_rate": 2.901459854014599e-06, + "loss": 2.122, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.5592572689056396, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.0957, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.738271951675415, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.0868, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.5362703800201416, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1237, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.6304945945739746, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.0825, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.040717601776123, + "learning_rate": 2.992700729927008e-06, + "loss": 2.0388, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.6226463317871094, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0163, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.186411142349243, + "eval_runtime": 44.625, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.5307068824768066, + "learning_rate": 3.029197080291971e-06, + "loss": 2.1097, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 2.9160444736480713, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.0946, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.5250024795532227, + "learning_rate": 3.065693430656935e-06, + "loss": 2.0709, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.6963064670562744, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.0937, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 2.845707416534424, + "learning_rate": 3.102189781021898e-06, + "loss": 2.0851, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.8985090255737305, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.0176, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 2.734377861022949, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.9931, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.5125555992126465, + "learning_rate": 3.156934306569343e-06, + "loss": 2.1133, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.0393502712249756, + "learning_rate": 3.175182481751825e-06, + "loss": 2.0636, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.6410164833068848, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0304, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.171868085861206, + "eval_runtime": 44.2249, + "eval_samples_per_second": 22.612, + "eval_steps_per_second": 0.95, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.7937498092651367, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0518, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7801201343536377, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.026, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7963638305664062, + "learning_rate": 3.248175182481752e-06, + "loss": 1.9642, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6330482959747314, + "learning_rate": 3.266423357664234e-06, + "loss": 2.0545, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.9771625995635986, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0445, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.6112730503082275, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0616, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.732865333557129, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0403, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.9113688468933105, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0382, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.4516773223876953, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.0416, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.781735897064209, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0348, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.160499334335327, + "eval_runtime": 44.5652, + "eval_samples_per_second": 22.439, + "eval_steps_per_second": 0.942, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.9990546703338623, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0324, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.783071279525757, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0518, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.659040689468384, + "learning_rate": 3.43065693430657e-06, + "loss": 1.9722, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 2.799798011779785, + "learning_rate": 3.4489051094890513e-06, + "loss": 2.0362, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 3.025458812713623, + "learning_rate": 3.467153284671533e-06, + "loss": 1.9863, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 2.686415195465088, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0201, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.716064929962158, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0249, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.637653350830078, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0403, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 2.7686848640441895, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0626, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.721956729888916, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.1077, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.152338981628418, + "eval_runtime": 43.5455, + "eval_samples_per_second": 22.964, + "eval_steps_per_second": 0.965, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.92571759223938, + "learning_rate": 3.576642335766424e-06, + "loss": 2.0928, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.9189202785491943, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9315, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 2.990233898162842, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9427, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.702348232269287, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9768, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6949572563171387, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.9954, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.108405113220215, + "learning_rate": 3.667883211678833e-06, + "loss": 1.9631, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.738922119140625, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9644, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.7419424057006836, + "learning_rate": 3.704379562043796e-06, + "loss": 2.0675, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9853813648223877, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0245, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.4614126682281494, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.9665, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.1342060565948486, + "eval_runtime": 46.3954, + "eval_samples_per_second": 21.554, + "eval_steps_per_second": 0.905, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.6787819862365723, + "learning_rate": 3.7591240875912412e-06, + "loss": 2.0063, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 2.563525438308716, + "learning_rate": 3.777372262773723e-06, + "loss": 1.998, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.4981701374053955, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.9738, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 2.5357978343963623, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0221, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.5365488529205322, + "learning_rate": 3.832116788321168e-06, + "loss": 1.9664, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 2.669468641281128, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9577, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 2.6950953006744385, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9786, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8249378204345703, + "learning_rate": 3.886861313868613e-06, + "loss": 2.0041, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.5924317836761475, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0338, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.1282334327697754, + "learning_rate": 3.923357664233577e-06, + "loss": 1.9698, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.1221420764923096, + "eval_runtime": 47.1127, + "eval_samples_per_second": 21.226, + "eval_steps_per_second": 0.891, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.7878193855285645, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0446, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.7137255668640137, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0575, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.286421060562134, + "learning_rate": 3.978102189781022e-06, + "loss": 1.9131, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.5466175079345703, + "learning_rate": 3.996350364963504e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.792483329772949, + "learning_rate": 4.014598540145986e-06, + "loss": 1.912, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.7533867359161377, + "learning_rate": 4.032846715328467e-06, + "loss": 1.9742, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 2.848405361175537, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9815, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.1034414768218994, + "learning_rate": 4.069343065693431e-06, + "loss": 1.9409, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.5109779834747314, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9676, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.486506938934326, + "learning_rate": 4.105839416058395e-06, + "loss": 2.0353, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1102187633514404, + "eval_runtime": 45.384, + "eval_samples_per_second": 22.034, + "eval_steps_per_second": 0.925, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.419269323348999, + "learning_rate": 4.124087591240876e-06, + "loss": 1.9622, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.4967262744903564, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.9523, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 2.7311460971832275, + "learning_rate": 4.16058394160584e-06, + "loss": 2.0234, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.241004467010498, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.907, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.821537733078003, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9761, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 2.5711426734924316, + "learning_rate": 4.215328467153285e-06, + "loss": 1.9411, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.936594009399414, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9432, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.6091108322143555, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9545, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.522674798965454, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9246, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.984649181365967, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9417, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1031947135925293, + "eval_runtime": 47.5409, + "eval_samples_per_second": 21.035, + "eval_steps_per_second": 0.883, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.8250324726104736, + "learning_rate": 4.306569343065693e-06, + "loss": 1.9061, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.7654290199279785, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9653, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.7091264724731445, + "learning_rate": 4.343065693430658e-06, + "loss": 1.9482, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 2.8019583225250244, + "learning_rate": 4.361313868613139e-06, + "loss": 1.9332, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.147368907928467, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9836, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.7050259113311768, + "learning_rate": 4.397810218978102e-06, + "loss": 1.9607, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.738743543624878, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9374, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.122777223587036, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9249, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.850745439529419, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9707, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.391998291015625, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.9703, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.0887911319732666, + "eval_runtime": 48.5149, + "eval_samples_per_second": 20.612, + "eval_steps_per_second": 0.866, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 2.9131698608398438, + "learning_rate": 4.489051094890512e-06, + "loss": 1.9989, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.1045937538146973, + "learning_rate": 4.507299270072993e-06, + "loss": 1.9718, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.7984158992767334, + "learning_rate": 4.525547445255475e-06, + "loss": 1.947, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.6223504543304443, + "learning_rate": 4.543795620437956e-06, + "loss": 1.8942, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.8221523761749268, + "learning_rate": 4.562043795620438e-06, + "loss": 2.0045, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.6657445430755615, + "learning_rate": 4.580291970802921e-06, + "loss": 2.0053, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 2.793455123901367, + "learning_rate": 4.598540145985402e-06, + "loss": 1.9577, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.551201820373535, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9256, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 2.739194631576538, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9265, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.2917332649230957, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9664, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.086106777191162, + "eval_runtime": 48.995, + "eval_samples_per_second": 20.41, + "eval_steps_per_second": 0.857, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 2.6489369869232178, + "learning_rate": 4.671532846715329e-06, + "loss": 1.9629, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 2.60017466545105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.932, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.569115400314331, + "learning_rate": 4.708029197080292e-06, + "loss": 1.9902, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 2.7225773334503174, + "learning_rate": 4.726277372262774e-06, + "loss": 1.9619, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.5064749717712402, + "learning_rate": 4.744525547445255e-06, + "loss": 1.8921, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.612931966781616, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9584, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.2524304389953613, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9675, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.8852503299713135, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9107, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 2.654776096343994, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9465, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.679086685180664, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9468, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.0842835903167725, + "eval_runtime": 51.7803, + "eval_samples_per_second": 19.312, + "eval_steps_per_second": 0.811, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.5110855102539062, + "learning_rate": 4.854014598540146e-06, + "loss": 1.88, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.1786117553710938, + "learning_rate": 4.872262773722628e-06, + "loss": 1.8994, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.388328790664673, + "learning_rate": 4.89051094890511e-06, + "loss": 1.8914, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 2.784834384918213, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.923, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.470426559448242, + "learning_rate": 4.927007299270074e-06, + "loss": 1.944, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.690580129623413, + "learning_rate": 4.945255474452555e-06, + "loss": 1.8978, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.869737386703491, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9964, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.9965591430664062, + "learning_rate": 4.981751824817518e-06, + "loss": 1.9706, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.6665658950805664, + "learning_rate": 5e-06, + "loss": 1.9313, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.9231722354888916, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9504, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.068364143371582, + "eval_runtime": 47.8334, + "eval_samples_per_second": 20.906, + "eval_steps_per_second": 0.878, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.6094462871551514, + "learning_rate": 4.995938261575955e-06, + "loss": 1.862, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.780848741531372, + "learning_rate": 4.993907392363932e-06, + "loss": 1.95, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.948662757873535, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.878, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 2.9622998237609863, + "learning_rate": 4.989845653939887e-06, + "loss": 1.8954, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 2.6401216983795166, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9021, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.029592275619507, + "learning_rate": 4.985783915515841e-06, + "loss": 1.8967, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.828392505645752, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9134, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.613926649093628, + "learning_rate": 4.981722177091796e-06, + "loss": 1.8829, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.6173207759857178, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9225, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.4657015800476074, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9142, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.059906482696533, + "eval_runtime": 49.6267, + "eval_samples_per_second": 20.15, + "eval_steps_per_second": 0.846, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7448673248291016, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9545, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.639096736907959, + "learning_rate": 4.973598700243705e-06, + "loss": 1.8664, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 2.6445884704589844, + "learning_rate": 4.971567831031681e-06, + "loss": 1.888, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.9130005836486816, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7717, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.48093318939209, + "learning_rate": 4.967506092607636e-06, + "loss": 1.8538, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.8918230533599854, + "learning_rate": 4.965475223395614e-06, + "loss": 1.886, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.511747121810913, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.8633, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.860029697418213, + "learning_rate": 4.961413484971569e-06, + "loss": 1.9063, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7304794788360596, + "learning_rate": 4.959382615759545e-06, + "loss": 1.8872, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.6569266319274902, + "learning_rate": 4.957351746547523e-06, + "loss": 1.8721, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0610575675964355, + "eval_runtime": 43.9292, + "eval_samples_per_second": 22.764, + "eval_steps_per_second": 0.956, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.1482994556427, + "learning_rate": 4.9553208773355e-06, + "loss": 2.0123, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.59568190574646, + "learning_rate": 4.953290008123478e-06, + "loss": 1.902, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.5897581577301025, + "learning_rate": 4.951259138911454e-06, + "loss": 1.8406, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.734032154083252, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.9605, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.8088371753692627, + "learning_rate": 4.947197400487409e-06, + "loss": 1.896, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.9678707122802734, + "learning_rate": 4.945166531275386e-06, + "loss": 1.8625, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.261200189590454, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8884, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 2.5811619758605957, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9311, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.6783649921417236, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9324, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.6533944606781006, + "learning_rate": 4.937043054427295e-06, + "loss": 1.8575, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.049668312072754, + "eval_runtime": 47.7347, + "eval_samples_per_second": 20.949, + "eval_steps_per_second": 0.88, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 3.0882792472839355, + "learning_rate": 4.935012185215272e-06, + "loss": 1.84, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.8259081840515137, + "learning_rate": 4.93298131600325e-06, + "loss": 1.9109, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.540908098220825, + "learning_rate": 4.930950446791227e-06, + "loss": 1.8906, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.9277396202087402, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9047, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.392449140548706, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.876, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 3.1890463829040527, + "learning_rate": 4.924857839155159e-06, + "loss": 1.8709, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.725710391998291, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8953, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.6287014484405518, + "learning_rate": 4.920796100731113e-06, + "loss": 1.8204, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.7316651344299316, + "learning_rate": 4.918765231519091e-06, + "loss": 1.882, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.5561037063598633, + "learning_rate": 4.916734362307068e-06, + "loss": 1.8331, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0412285327911377, + "eval_runtime": 50.237, + "eval_samples_per_second": 19.906, + "eval_steps_per_second": 0.836, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.974501132965088, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8414, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.4530839920043945, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8323, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.789569616317749, + "learning_rate": 4.910641754671e-06, + "loss": 1.8697, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.582704544067383, + "learning_rate": 4.908610885458977e-06, + "loss": 1.8592, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 2.776442289352417, + "learning_rate": 4.906580016246954e-06, + "loss": 1.912, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.670191764831543, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.8899, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.4534454345703125, + "learning_rate": 4.902518277822909e-06, + "loss": 1.8878, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.007469654083252, + "learning_rate": 4.900487408610886e-06, + "loss": 1.8977, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 2.856466293334961, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9062, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.817441463470459, + "learning_rate": 4.89642567018684e-06, + "loss": 1.9686, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.0344440937042236, + "eval_runtime": 47.2472, + "eval_samples_per_second": 21.165, + "eval_steps_per_second": 0.889, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.9732158184051514, + "learning_rate": 4.894394800974818e-06, + "loss": 1.8332, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 3.0413055419921875, + "learning_rate": 4.892363931762794e-06, + "loss": 1.9171, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.646749973297119, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.8403, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.1274876594543457, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8218, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.8122124671936035, + "learning_rate": 4.886271324126727e-06, + "loss": 1.9272, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.4608492851257324, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.8974, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.928083658218384, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.8646, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.6308164596557617, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8219, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7311816215515137, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8326, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.640162467956543, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8013, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0261101722717285, + "eval_runtime": 47.0605, + "eval_samples_per_second": 21.249, + "eval_steps_per_second": 0.892, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.8542518615722656, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8233, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.464303970336914, + "learning_rate": 4.872055239642567e-06, + "loss": 1.787, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8535103797912598, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8535, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.5060505867004395, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8733, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.6549222469329834, + "learning_rate": 4.865962632006499e-06, + "loss": 1.8047, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.8636224269866943, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8543, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.644256830215454, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8853, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.949490547180176, + "learning_rate": 4.859870024370431e-06, + "loss": 1.8951, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.4576282501220703, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9069, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.6140542030334473, + "learning_rate": 4.855808285946385e-06, + "loss": 1.8573, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0159800052642822, + "eval_runtime": 44.2447, + "eval_samples_per_second": 22.602, + "eval_steps_per_second": 0.949, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.525766134262085, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.844, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.5479896068573, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8572, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.628570795059204, + "learning_rate": 4.849715678310317e-06, + "loss": 1.865, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.2768666744232178, + "learning_rate": 4.847684809098294e-06, + "loss": 1.8355, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.635516881942749, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8603, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.5307435989379883, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8482, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.3688952922821045, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8894, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.6000115871429443, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.8045, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7782504558563232, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8283, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.6914777755737305, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8765, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0086047649383545, + "eval_runtime": 48.245, + "eval_samples_per_second": 20.728, + "eval_steps_per_second": 0.871, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.598365306854248, + "learning_rate": 4.833468724614135e-06, + "loss": 1.8461, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.6718530654907227, + "learning_rate": 4.831437855402113e-06, + "loss": 1.7453, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.393528699874878, + "learning_rate": 4.82940698619009e-06, + "loss": 1.8275, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 3.0622947216033936, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8658, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.431041955947876, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8601, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.939424514770508, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8649, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.8116796016693115, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8303, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.511960029602051, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8383, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 3.8372581005096436, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8111, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.511948347091675, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8448, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.9998281002044678, + "eval_runtime": 44.7624, + "eval_samples_per_second": 22.34, + "eval_steps_per_second": 0.938, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 3.5780487060546875, + "learning_rate": 4.813160032493907e-06, + "loss": 1.8161, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.81740403175354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.8633, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.8117666244506836, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8385, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.5791451930999756, + "learning_rate": 4.80706742485784e-06, + "loss": 1.896, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6816465854644775, + "learning_rate": 4.805036555645816e-06, + "loss": 1.8126, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.6419460773468018, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8753, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.666086196899414, + "learning_rate": 4.800974817221771e-06, + "loss": 1.8305, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8626585006713867, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8463, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.351825475692749, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.838, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.300717830657959, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8711, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.9904799461364746, + "eval_runtime": 43.9511, + "eval_samples_per_second": 22.753, + "eval_steps_per_second": 0.956, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.43884015083313, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8145, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.50593900680542, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8103, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.4643657207489014, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8162, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.4615302085876465, + "learning_rate": 4.786758732737612e-06, + "loss": 1.8637, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.6566433906555176, + "learning_rate": 4.784727863525589e-06, + "loss": 1.7805, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.5973432064056396, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8327, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.4996957778930664, + "learning_rate": 4.780666125101544e-06, + "loss": 1.7761, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 3.0266876220703125, + "learning_rate": 4.778635255889521e-06, + "loss": 1.8036, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.522369146347046, + "learning_rate": 4.776604386677498e-06, + "loss": 1.814, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.482727527618408, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8665, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.9888195991516113, + "eval_runtime": 44.0755, + "eval_samples_per_second": 22.688, + "eval_steps_per_second": 0.953, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 3.237544059753418, + "learning_rate": 4.772542648253453e-06, + "loss": 1.8642, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.393286943435669, + "learning_rate": 4.77051177904143e-06, + "loss": 1.7857, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.66072154045105, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8423, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.6725080013275146, + "learning_rate": 4.766450040617385e-06, + "loss": 1.815, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.7657666206359863, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8729, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.3153769969940186, + "learning_rate": 4.762388302193339e-06, + "loss": 1.8349, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5804553031921387, + "learning_rate": 4.760357432981316e-06, + "loss": 1.8155, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.7280807495117188, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8262, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.551940441131592, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8332, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.6137285232543945, + "learning_rate": 4.754264825345248e-06, + "loss": 1.7805, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.983920693397522, + "eval_runtime": 42.525, + "eval_samples_per_second": 23.516, + "eval_steps_per_second": 0.988, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.4460153579711914, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.7953, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.556032657623291, + "learning_rate": 4.750203086921203e-06, + "loss": 1.7393, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.335983991622925, + "learning_rate": 4.74817221770918e-06, + "loss": 1.8494, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.5057711601257324, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8663, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4447145462036133, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7532, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.769984483718872, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7755, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.472193956375122, + "learning_rate": 4.740048740861089e-06, + "loss": 1.8056, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.4286932945251465, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8094, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.0248074531555176, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8497, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.607700824737549, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8741, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.9779797792434692, + "eval_runtime": 43.7734, + "eval_samples_per_second": 22.845, + "eval_steps_per_second": 0.959, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 3.035045623779297, + "learning_rate": 4.731925264012998e-06, + "loss": 1.8438, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.786921977996826, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7666, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.4158425331115723, + "learning_rate": 4.727863525588953e-06, + "loss": 1.7464, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7772, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.8447, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 3.08266282081604, + "learning_rate": 4.723801787164907e-06, + "loss": 1.7812, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.791749954223633, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8047, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6457693576812744, + "learning_rate": 4.719740048740861e-06, + "loss": 1.836, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.69480037689209, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8145, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6356561183929443, + "learning_rate": 4.715678310316816e-06, + "loss": 1.7756, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.9723366498947144, + "eval_runtime": 45.5806, + "eval_samples_per_second": 21.939, + "eval_steps_per_second": 0.921, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.53474497795105, + "learning_rate": 4.713647441104794e-06, + "loss": 1.774, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.3631107807159424, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8517, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.594604730606079, + "learning_rate": 4.709585702680748e-06, + "loss": 1.8962, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.5280940532684326, + "learning_rate": 4.707554833468725e-06, + "loss": 1.7759, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.4596288204193115, + "learning_rate": 4.705523964256703e-06, + "loss": 1.7948, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.9123644828796387, + "learning_rate": 4.703493095044679e-06, + "loss": 1.7399, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.4354264736175537, + "learning_rate": 4.701462225832657e-06, + "loss": 1.8659, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.28625226020813, + "learning_rate": 4.699431356620634e-06, + "loss": 1.831, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.5445950031280518, + "learning_rate": 4.697400487408612e-06, + "loss": 1.8121, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.767825126647949, + "learning_rate": 4.695369618196588e-06, + "loss": 1.7874, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9653255939483643, + "eval_runtime": 45.2453, + "eval_samples_per_second": 22.102, + "eval_steps_per_second": 0.928, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4052765369415283, + "learning_rate": 4.693338748984566e-06, + "loss": 1.7539, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.498098373413086, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8343, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.821753740310669, + "learning_rate": 4.68927701056052e-06, + "loss": 1.8119, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.4158449172973633, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.7982, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.5229434967041016, + "learning_rate": 4.685215272136475e-06, + "loss": 1.806, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.729199171066284, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8768, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.5102570056915283, + "learning_rate": 4.681153533712429e-06, + "loss": 1.775, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4463515281677246, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7493, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.505733013153076, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8462, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.5784912109375, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7381, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9640673398971558, + "eval_runtime": 44.2714, + "eval_samples_per_second": 22.588, + "eval_steps_per_second": 0.949, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.7864291667938232, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7612, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.507809638977051, + "learning_rate": 4.670999187652316e-06, + "loss": 1.757, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.5499937534332275, + "learning_rate": 4.668968318440293e-06, + "loss": 1.8243, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.435472011566162, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7269, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.541031837463379, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.7959, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5052099227905273, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7191, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6285109519958496, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8423, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.7825605869293213, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7178, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.677319288253784, + "learning_rate": 4.656783103168157e-06, + "loss": 1.7939, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.6830129623413086, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7842, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.960136890411377, + "eval_runtime": 45.9379, + "eval_samples_per_second": 21.769, + "eval_steps_per_second": 0.914, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.5368666648864746, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8352, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.8922412395477295, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.8252, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.8332598209381104, + "learning_rate": 4.648659626320066e-06, + "loss": 1.7805, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.420780658721924, + "learning_rate": 4.646628757108043e-06, + "loss": 1.7499, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.6684062480926514, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.7857, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.8547494411468506, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.8115, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.59279203414917, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8254, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.80961275100708, + "learning_rate": 4.638505280259952e-06, + "loss": 1.7901, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.9017655849456787, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8265, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.595454216003418, + "learning_rate": 4.634443541835907e-06, + "loss": 1.7833, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9576054811477661, + "eval_runtime": 50.7265, + "eval_samples_per_second": 19.714, + "eval_steps_per_second": 0.828, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 4.260303020477295, + "learning_rate": 4.632412672623883e-06, + "loss": 1.6741, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.366147756576538, + "learning_rate": 4.630381803411861e-06, + "loss": 1.7778, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.894336462020874, + "learning_rate": 4.628350934199838e-06, + "loss": 1.8268, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.7675349712371826, + "learning_rate": 4.626320064987816e-06, + "loss": 1.7748, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.461413621902466, + "learning_rate": 4.624289195775792e-06, + "loss": 1.7599, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.670187473297119, + "learning_rate": 4.62225832656377e-06, + "loss": 1.7736, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.8119454383850098, + "learning_rate": 4.620227457351747e-06, + "loss": 1.7837, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.8020036220550537, + "learning_rate": 4.618196588139724e-06, + "loss": 1.7508, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.5977094173431396, + "learning_rate": 4.616165718927701e-06, + "loss": 1.834, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.731919765472412, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.8137, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.9465162754058838, + "eval_runtime": 48.9829, + "eval_samples_per_second": 20.415, + "eval_steps_per_second": 0.857, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.604386568069458, + "learning_rate": 4.612103980503656e-06, + "loss": 1.7854, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.601107120513916, + "learning_rate": 4.610073111291633e-06, + "loss": 1.7749, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6758370399475098, + "learning_rate": 4.60804224207961e-06, + "loss": 1.7826, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.631549119949341, + "learning_rate": 4.606011372867588e-06, + "loss": 1.7723, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.6704764366149902, + "learning_rate": 4.603980503655565e-06, + "loss": 1.7088, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 3.078573703765869, + "learning_rate": 4.601949634443542e-06, + "loss": 1.7254, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7492356300354004, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7756, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 3.0763585567474365, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7806, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.5379223823547363, + "learning_rate": 4.595857026807474e-06, + "loss": 1.7516, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8343496322631836, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8111, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.945749282836914, + "eval_runtime": 46.5453, + "eval_samples_per_second": 21.484, + "eval_steps_per_second": 0.902, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.809753179550171, + "learning_rate": 4.591795288383429e-06, + "loss": 1.7877, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.4196107387542725, + "learning_rate": 4.589764419171406e-06, + "loss": 1.7457, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.409287452697754, + "learning_rate": 4.587733549959383e-06, + "loss": 1.7338, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.414706230163574, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7185, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.591365337371826, + "learning_rate": 4.583671811535338e-06, + "loss": 1.8083, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.4648489952087402, + "learning_rate": 4.581640942323315e-06, + "loss": 1.8192, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.3804972171783447, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8077, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.283082962036133, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8294, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.61795711517334, + "learning_rate": 4.575548334687247e-06, + "loss": 1.7322, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.4035353660583496, + "learning_rate": 4.573517465475224e-06, + "loss": 1.7798, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9422866106033325, + "eval_runtime": 45.3068, + "eval_samples_per_second": 22.072, + "eval_steps_per_second": 0.927, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.7253756523132324, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7876, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.738295078277588, + "learning_rate": 4.569455727051178e-06, + "loss": 1.7933, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 3.1418697834014893, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8145, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.941603660583496, + "learning_rate": 4.565393988627132e-06, + "loss": 1.7671, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.063924551010132, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.7809, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 4.957335472106934, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7306, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.651716709136963, + "learning_rate": 4.559301380991065e-06, + "loss": 1.7428, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 4.777034282684326, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.8118, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.5549471378326416, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7294, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.474388599395752, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8338, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.9402854442596436, + "eval_runtime": 43.166, + "eval_samples_per_second": 23.166, + "eval_steps_per_second": 0.973, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.8968348503112793, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7432, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.6996726989746094, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.7837, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.4599392414093018, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7388, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.745487928390503, + "learning_rate": 4.545085296506905e-06, + "loss": 1.7439, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.6960670948028564, + "learning_rate": 4.543054427294882e-06, + "loss": 1.737, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.596876621246338, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7708, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.598267078399658, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7541, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.6069958209991455, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7756, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.490694761276245, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.7372, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.44179368019104, + "learning_rate": 4.532900081234769e-06, + "loss": 1.7652, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.934715986251831, + "eval_runtime": 44.688, + "eval_samples_per_second": 22.377, + "eval_steps_per_second": 0.94, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 3.018480062484741, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7596, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.5316643714904785, + "learning_rate": 4.528838342810723e-06, + "loss": 1.7715, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4893109798431396, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7549, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.524834156036377, + "learning_rate": 4.524776604386678e-06, + "loss": 1.8046, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7527239322662354, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7583, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.5210752487182617, + "learning_rate": 4.520714865962632e-06, + "loss": 1.7896, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.5787770748138428, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7355, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.537195920944214, + "learning_rate": 4.516653127538587e-06, + "loss": 1.7272, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.6810014247894287, + "learning_rate": 4.514622258326564e-06, + "loss": 1.753, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.5868756771087646, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7964, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.934499740600586, + "eval_runtime": 45.6556, + "eval_samples_per_second": 21.903, + "eval_steps_per_second": 0.92, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.6830856800079346, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7072, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.6574594974517822, + "learning_rate": 4.508529650690496e-06, + "loss": 1.7004, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.4415769577026367, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7667, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.825115442276001, + "learning_rate": 4.504467912266451e-06, + "loss": 1.8356, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.949010133743286, + "learning_rate": 4.502437043054428e-06, + "loss": 1.7579, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.690986394882202, + "learning_rate": 4.500406173842405e-06, + "loss": 1.7657, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 3.1015520095825195, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.7432, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.608424186706543, + "learning_rate": 4.49634443541836e-06, + "loss": 1.8112, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.5238773822784424, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7724, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.7712161540985107, + "learning_rate": 4.492282696994314e-06, + "loss": 1.7111, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9326083660125732, + "eval_runtime": 48.5359, + "eval_samples_per_second": 20.603, + "eval_steps_per_second": 0.865, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.4763033390045166, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7268, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.6310081481933594, + "learning_rate": 4.488220958570269e-06, + "loss": 1.6885, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.5950350761413574, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7775, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.769026517868042, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7737, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.4898502826690674, + "learning_rate": 4.4821283509342e-06, + "loss": 1.751, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.668081521987915, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7178, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.441615104675293, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7511, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.727665901184082, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.8639, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.7964634895324707, + "learning_rate": 4.474004874086109e-06, + "loss": 1.6945, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.6142687797546387, + "learning_rate": 4.471974004874087e-06, + "loss": 1.8028, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9281476736068726, + "eval_runtime": 48.078, + "eval_samples_per_second": 20.8, + "eval_steps_per_second": 0.874, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.6655778884887695, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.7632, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.931162118911743, + "learning_rate": 4.467912266450041e-06, + "loss": 1.7244, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.469895124435425, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7773, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.4948830604553223, + "learning_rate": 4.463850528025995e-06, + "loss": 1.788, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5961945056915283, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.756, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.452946424484253, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6968, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.9352009296417236, + "learning_rate": 4.457757920389927e-06, + "loss": 1.7414, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.7044763565063477, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7848, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.7200324535369873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7699, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.6185197830200195, + "learning_rate": 4.451665312753859e-06, + "loss": 1.7816, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9209110736846924, + "eval_runtime": 47.3321, + "eval_samples_per_second": 21.127, + "eval_steps_per_second": 0.887, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.3639564514160156, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7164, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7073519229888916, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.727, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.637080430984497, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7951, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.8013885021209717, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7333, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.512484312057495, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8048, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.3851466178894043, + "learning_rate": 4.439480097481723e-06, + "loss": 1.6667, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.8105392456054688, + "learning_rate": 4.4374492282697e-06, + "loss": 1.7622, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.555773973464966, + "learning_rate": 4.435418359057677e-06, + "loss": 1.7519, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.688596725463867, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7198, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.828646421432495, + "learning_rate": 4.431356620633632e-06, + "loss": 1.7478, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.919256567955017, + "eval_runtime": 50.5993, + "eval_samples_per_second": 19.763, + "eval_steps_per_second": 0.83, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.4523651599884033, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7344, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.4237422943115234, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6664, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.2746193408966064, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.7446, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.502819061279297, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7691, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.7032017707824707, + "learning_rate": 4.421202274573518e-06, + "loss": 1.6448, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.5879440307617188, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6908, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 3.2481982707977295, + "learning_rate": 4.417140536149473e-06, + "loss": 1.7725, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.4844820499420166, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7141, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.5596253871917725, + "learning_rate": 4.413078797725427e-06, + "loss": 1.7485, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 3.142839193344116, + "learning_rate": 4.411047928513404e-06, + "loss": 1.734, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9129430055618286, + "eval_runtime": 49.2608, + "eval_samples_per_second": 20.3, + "eval_steps_per_second": 0.853, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.5837552547454834, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7668, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.7043633460998535, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7497, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.9886765480041504, + "learning_rate": 4.404955320877336e-06, + "loss": 1.7082, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.3199501037597656, + "learning_rate": 4.402924451665313e-06, + "loss": 1.693, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.606682777404785, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7242, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.6967978477478027, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7516, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.351726531982422, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7338, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.590869188308716, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7462, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.3498220443725586, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7609, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.281002998352051, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7584, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.908841609954834, + "eval_runtime": 50.7771, + "eval_samples_per_second": 19.694, + "eval_steps_per_second": 0.827, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.468755006790161, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7654, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.3467307090759277, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7923, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.3723301887512207, + "learning_rate": 4.384646628757108e-06, + "loss": 1.74, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.4527792930603027, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6988, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.3923704624176025, + "learning_rate": 4.380584890333063e-06, + "loss": 1.6406, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.740556240081787, + "learning_rate": 4.37855402112104e-06, + "loss": 1.8032, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.3013832569122314, + "learning_rate": 4.376523151909017e-06, + "loss": 1.7259, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.4632391929626465, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.7462, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.4878177642822266, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7665, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.610898494720459, + "learning_rate": 4.370430544272949e-06, + "loss": 1.7646, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9056707620620728, + "eval_runtime": 51.8738, + "eval_samples_per_second": 19.278, + "eval_steps_per_second": 0.81, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.387721300125122, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7454, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.9966042041778564, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6789, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.664689064025879, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7486, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.4788103103637695, + "learning_rate": 4.362307067424858e-06, + "loss": 1.6981, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.4413328170776367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7279, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.6818156242370605, + "learning_rate": 4.358245329000813e-06, + "loss": 1.74, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.650348424911499, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7212, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.442805767059326, + "learning_rate": 4.354183590576767e-06, + "loss": 1.6951, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.6806585788726807, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7499, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.586792230606079, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7709, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.901674747467041, + "eval_runtime": 42.3186, + "eval_samples_per_second": 23.63, + "eval_steps_per_second": 0.992, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.547452688217163, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7458, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.5820348262786865, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.7232, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.4208321571350098, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7981, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.8791956901550293, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7372, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6537868976593018, + "learning_rate": 4.339967506092607e-06, + "loss": 1.7216, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.416097402572632, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.701, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.5565483570098877, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7558, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.5875823497772217, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7439, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5459702014923096, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.7058, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.4195566177368164, + "learning_rate": 4.329813160032495e-06, + "loss": 1.6749, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9035066366195679, + "eval_runtime": 43.7962, + "eval_samples_per_second": 22.833, + "eval_steps_per_second": 0.959, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6019818782806396, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7549, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.8627779483795166, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7257, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.582158088684082, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7138, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.4528613090515137, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7162, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 3.1956026554107666, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7556, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.3531901836395264, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.746, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.441988706588745, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7664, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.599506378173828, + "learning_rate": 4.313566206336312e-06, + "loss": 1.6955, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.4121932983398438, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7194, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.4803290367126465, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7386, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8993914127349854, + "eval_runtime": 43.0525, + "eval_samples_per_second": 23.227, + "eval_steps_per_second": 0.976, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.701594114303589, + "learning_rate": 4.307473598700244e-06, + "loss": 1.6796, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7144737243652344, + "learning_rate": 4.305442729488221e-06, + "loss": 1.763, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.395312786102295, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7522, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.3926732540130615, + "learning_rate": 4.301380991064176e-06, + "loss": 1.7705, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.6293702125549316, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7693, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.3746306896209717, + "learning_rate": 4.29731925264013e-06, + "loss": 1.72, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.5421926975250244, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6868, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.6991186141967773, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7378, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.463897705078125, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7628, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5139036178588867, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7125, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8909698724746704, + "eval_runtime": 42.5917, + "eval_samples_per_second": 23.479, + "eval_steps_per_second": 0.986, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.5220465660095215, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6975, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.432988405227661, + "learning_rate": 4.285134037367994e-06, + "loss": 1.6904, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.5894253253936768, + "learning_rate": 4.283103168155971e-06, + "loss": 1.6919, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5334458351135254, + "learning_rate": 4.281072298943948e-06, + "loss": 1.6968, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.510019302368164, + "learning_rate": 4.279041429731926e-06, + "loss": 1.6791, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.531538248062134, + "learning_rate": 4.277010560519903e-06, + "loss": 1.723, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.5621440410614014, + "learning_rate": 4.27497969130788e-06, + "loss": 1.749, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.594329833984375, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7579, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.321031093597412, + "learning_rate": 4.270917952883835e-06, + "loss": 1.6983, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.6325764656066895, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7872, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.886696219444275, + "eval_runtime": 43.4143, + "eval_samples_per_second": 23.034, + "eval_steps_per_second": 0.967, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5342679023742676, + "learning_rate": 4.266856214459789e-06, + "loss": 1.6655, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.5810720920562744, + "learning_rate": 4.264825345247766e-06, + "loss": 1.7182, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.7768406867980957, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6534, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.5680463314056396, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7243, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7162973880767822, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.7146, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.697925090789795, + "learning_rate": 4.256701868399675e-06, + "loss": 1.7296, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.4458425045013428, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7466, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6330654621124268, + "learning_rate": 4.252640129975629e-06, + "loss": 1.6949, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.350430727005005, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6574, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.505995512008667, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7111, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.8887447118759155, + "eval_runtime": 43.9304, + "eval_samples_per_second": 22.763, + "eval_steps_per_second": 0.956, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 3.036954879760742, + "learning_rate": 4.246547522339562e-06, + "loss": 1.7182, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.563606023788452, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.6566, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.516315460205078, + "learning_rate": 4.242485783915517e-06, + "loss": 1.6543, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.6289877891540527, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6996, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.7563092708587646, + "learning_rate": 4.23842404549147e-06, + "loss": 1.666, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5743000507354736, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7338, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 3.1975982189178467, + "learning_rate": 4.234362307067425e-06, + "loss": 1.7444, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.6022894382476807, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7367, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.4466192722320557, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.726, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.579015016555786, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7146, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.8904188871383667, + "eval_runtime": 43.8091, + "eval_samples_per_second": 22.826, + "eval_steps_per_second": 0.959, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.719646453857422, + "learning_rate": 4.226238830219334e-06, + "loss": 1.7209, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.4423880577087402, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7288, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5883753299713135, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7089, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.447049140930176, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7095, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.453739881515503, + "learning_rate": 4.218115353371243e-06, + "loss": 1.6747, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4540836811065674, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7578, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.5468671321868896, + "learning_rate": 4.214053614947198e-06, + "loss": 1.6556, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.443814277648926, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7307, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.4998748302459717, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6768, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.667025566101074, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7493, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.8830010890960693, + "eval_runtime": 44.7027, + "eval_samples_per_second": 22.37, + "eval_steps_per_second": 0.94, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.416248321533203, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7381, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.378025770187378, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7335, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.633173704147339, + "learning_rate": 4.201868399675061e-06, + "loss": 1.6484, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.549384117126465, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7143, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.6492598056793213, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6907, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.444086790084839, + "learning_rate": 4.195775792038993e-06, + "loss": 1.636, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.628814935684204, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7342, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.83831524848938, + "learning_rate": 4.191714053614948e-06, + "loss": 1.7028, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.467339515686035, + "learning_rate": 4.189683184402924e-06, + "loss": 1.749, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.4529521465301514, + "learning_rate": 4.187652315190902e-06, + "loss": 1.682, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.8826342821121216, + "eval_runtime": 44.7757, + "eval_samples_per_second": 22.334, + "eval_steps_per_second": 0.938, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7340762615203857, + "learning_rate": 4.185621445978879e-06, + "loss": 1.723, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.6028647422790527, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7566, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.3967695236206055, + "learning_rate": 4.181559707554833e-06, + "loss": 1.683, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5475900173187256, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6976, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.5411484241485596, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7489, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 3.458749294281006, + "learning_rate": 4.175467099918766e-06, + "loss": 1.6828, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.2977802753448486, + "learning_rate": 4.173436230706742e-06, + "loss": 1.6963, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.875307559967041, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.6857, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 4.097696304321289, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7225, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.6774847507476807, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7031, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.8783056735992432, + "eval_runtime": 44.0219, + "eval_samples_per_second": 22.716, + "eval_steps_per_second": 0.954, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.6787376403808594, + "learning_rate": 4.165312753858652e-06, + "loss": 1.6865, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.3803224563598633, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7237, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.9285683631896973, + "learning_rate": 4.161251015434607e-06, + "loss": 1.6575, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.5534448623657227, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6799, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4105820655822754, + "learning_rate": 4.157189277010561e-06, + "loss": 1.7128, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.6571805477142334, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7123, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.891935348510742, + "learning_rate": 4.153127538586516e-06, + "loss": 1.6949, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.8692519664764404, + "learning_rate": 4.151096669374492e-06, + "loss": 1.699, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.780078649520874, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7114, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.4797518253326416, + "learning_rate": 4.147034930950447e-06, + "loss": 1.6943, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.8726727962493896, + "eval_runtime": 43.5664, + "eval_samples_per_second": 22.953, + "eval_steps_per_second": 0.964, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.7010464668273926, + "learning_rate": 4.145004061738425e-06, + "loss": 1.6207, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8812713623046875, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.6608, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.6700279712677, + "learning_rate": 4.140942323314379e-06, + "loss": 1.677, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.796513557434082, + "learning_rate": 4.138911454102356e-06, + "loss": 1.7291, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.4274256229400635, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7058, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5528790950775146, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7064, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.4288926124572754, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7163, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.4939630031585693, + "learning_rate": 4.130787977254265e-06, + "loss": 1.721, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7909305095672607, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7777, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.3382463455200195, + "learning_rate": 4.12672623883022e-06, + "loss": 1.6975, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.8738009929656982, + "eval_runtime": 44.8142, + "eval_samples_per_second": 22.314, + "eval_steps_per_second": 0.937, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6876718997955322, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6851, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.450212240219116, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7209, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.4365246295928955, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.6545, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.3236255645751953, + "learning_rate": 4.118602761982129e-06, + "loss": 1.6434, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.4753000736236572, + "learning_rate": 4.116571892770106e-06, + "loss": 1.7471, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.84885835647583, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6777, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.4361016750335693, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7079, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.5868160724639893, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7288, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.6499862670898438, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7191, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.333289384841919, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5938, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.8746954202651978, + "eval_runtime": 46.8039, + "eval_samples_per_second": 21.366, + "eval_steps_per_second": 0.897, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.1299779415130615, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7585, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 3.6410014629364014, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6803, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.801745653152466, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6957, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 3.0350706577301025, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7031, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 3.5306341648101807, + "learning_rate": 4.096263200649879e-06, + "loss": 1.6598, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.591485023498535, + "learning_rate": 4.094232331437856e-06, + "loss": 1.6659, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 3.0916528701782227, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6951, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.459009885787964, + "learning_rate": 4.090170593013811e-06, + "loss": 1.6597, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.932072162628174, + "learning_rate": 4.088139723801787e-06, + "loss": 1.6698, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.773566484451294, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6824, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.8707821369171143, + "eval_runtime": 46.7886, + "eval_samples_per_second": 21.373, + "eval_steps_per_second": 0.898, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6008529663085938, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7352, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.502838134765625, + "learning_rate": 4.08204711616572e-06, + "loss": 1.6342, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 3.081169605255127, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7307, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.841200828552246, + "learning_rate": 4.077985377741674e-06, + "loss": 1.7401, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.5010390281677246, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7252, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.468873977661133, + "learning_rate": 4.073923639317629e-06, + "loss": 1.6627, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.829737663269043, + "learning_rate": 4.071892770105605e-06, + "loss": 1.6419, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.8112833499908447, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6629, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6793017387390137, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6667, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.327993869781494, + "learning_rate": 4.065800162469538e-06, + "loss": 1.71, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.874659538269043, + "eval_runtime": 49.0634, + "eval_samples_per_second": 20.382, + "eval_steps_per_second": 0.856, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.689023971557617, + "learning_rate": 4.063769293257514e-06, + "loss": 1.6881, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.612842082977295, + "learning_rate": 4.061738424045492e-06, + "loss": 1.6971, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.319627046585083, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7469, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.588916540145874, + "learning_rate": 4.057676685621446e-06, + "loss": 1.707, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.270096778869629, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6284, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.4404375553131104, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7007, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5532679557800293, + "learning_rate": 4.051584077985378e-06, + "loss": 1.72, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.5105185508728027, + "learning_rate": 4.049553208773355e-06, + "loss": 1.6772, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.577929973602295, + "learning_rate": 4.047522339561333e-06, + "loss": 1.6747, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.8271372318267822, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7951, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.8684895038604736, + "eval_runtime": 48.0288, + "eval_samples_per_second": 20.821, + "eval_steps_per_second": 0.874, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.3306479454040527, + "learning_rate": 4.043460601137287e-06, + "loss": 1.6791, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.5543673038482666, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6761, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.3913605213165283, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7579, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.3837356567382812, + "learning_rate": 4.037367993501219e-06, + "loss": 1.6636, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.6936402320861816, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6488, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.6400210857391357, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.6926, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.574533462524414, + "learning_rate": 4.031275385865151e-06, + "loss": 1.7066, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5130677223205566, + "learning_rate": 4.029244516653128e-06, + "loss": 1.6431, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.578166961669922, + "learning_rate": 4.027213647441105e-06, + "loss": 1.648, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.4736695289611816, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6822, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.8754831552505493, + "eval_runtime": 50.0971, + "eval_samples_per_second": 19.961, + "eval_steps_per_second": 0.838, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.544315814971924, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6853, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8765335083007812, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7587, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.4749417304992676, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.6536, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.493164300918579, + "learning_rate": 4.017059301380992e-06, + "loss": 1.7436, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.6274492740631104, + "learning_rate": 4.015028432168969e-06, + "loss": 1.632, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.4162979125976562, + "learning_rate": 4.012997562956945e-06, + "loss": 1.6663, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5683343410491943, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6771, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.3710429668426514, + "learning_rate": 4.0089358245329e-06, + "loss": 1.6644, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.5356292724609375, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7339, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.480884075164795, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.716, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.867702841758728, + "eval_runtime": 48.0038, + "eval_samples_per_second": 20.832, + "eval_steps_per_second": 0.875, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4768476486206055, + "learning_rate": 4.002843216896833e-06, + "loss": 1.6667, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5727357864379883, + "learning_rate": 4.000812347684809e-06, + "loss": 1.6786, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6243903636932373, + "learning_rate": 3.998781478472787e-06, + "loss": 1.6328, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.712955951690674, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7852, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.756272554397583, + "learning_rate": 3.994719740048742e-06, + "loss": 1.7006, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.919201135635376, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6624, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.471566915512085, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7122, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.46273136138916, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6621, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.5338189601898193, + "learning_rate": 3.98659626320065e-06, + "loss": 1.6869, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.3850855827331543, + "learning_rate": 3.984565393988627e-06, + "loss": 1.6925, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.8650325536727905, + "eval_runtime": 49.4167, + "eval_samples_per_second": 20.236, + "eval_steps_per_second": 0.85, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.654726028442383, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7113, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.5065038204193115, + "learning_rate": 3.980503655564582e-06, + "loss": 1.7078, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.51668381690979, + "learning_rate": 3.978472786352559e-06, + "loss": 1.6742, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4586126804351807, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7339, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.452857732772827, + "learning_rate": 3.974411047928514e-06, + "loss": 1.6993, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.608631134033203, + "learning_rate": 3.972380178716491e-06, + "loss": 1.7047, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.34022855758667, + "learning_rate": 3.970349309504468e-06, + "loss": 1.6423, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.329612970352173, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.6443, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.3301169872283936, + "learning_rate": 3.966287571080423e-06, + "loss": 1.6937, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.3010976314544678, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6896, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.863387942314148, + "eval_runtime": 47.6761, + "eval_samples_per_second": 20.975, + "eval_steps_per_second": 0.881, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.296764612197876, + "learning_rate": 3.962225832656377e-06, + "loss": 1.6098, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.437114953994751, + "learning_rate": 3.960194963444355e-06, + "loss": 1.6727, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4908952713012695, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6816, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.81123685836792, + "learning_rate": 3.956133225020309e-06, + "loss": 1.6795, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.7088773250579834, + "learning_rate": 3.954102355808286e-06, + "loss": 1.6838, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.5021824836730957, + "learning_rate": 3.952071486596264e-06, + "loss": 1.671, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.3021023273468018, + "learning_rate": 3.950040617384241e-06, + "loss": 1.651, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.866119861602783, + "learning_rate": 3.948009748172218e-06, + "loss": 1.7117, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.3471503257751465, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.6883, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4612059593200684, + "learning_rate": 3.943948009748173e-06, + "loss": 1.6898, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.8587063550949097, + "eval_runtime": 43.4462, + "eval_samples_per_second": 23.017, + "eval_steps_per_second": 0.967, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.376680374145508, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7253, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.4701383113861084, + "learning_rate": 3.939886271324127e-06, + "loss": 1.6306, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.574384927749634, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6287, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.8596720695495605, + "learning_rate": 3.935824532900082e-06, + "loss": 1.7031, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.7144267559051514, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7219, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.5807480812072754, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.6364, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 3.3094675540924072, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6526, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.390503168106079, + "learning_rate": 3.927701056051991e-06, + "loss": 1.6795, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.5525588989257812, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.6432, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.522307872772217, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6218, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.8604834079742432, + "eval_runtime": 46.6465, + "eval_samples_per_second": 21.438, + "eval_steps_per_second": 0.9, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.438201904296875, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6788, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.639164686203003, + "learning_rate": 3.9195775792039e-06, + "loss": 1.6552, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.608038902282715, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.69, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.4357974529266357, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7031, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.4964048862457275, + "learning_rate": 3.913484971567831e-06, + "loss": 1.6126, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.5115931034088135, + "learning_rate": 3.911454102355808e-06, + "loss": 1.684, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.54923152923584, + "learning_rate": 3.909423233143786e-06, + "loss": 1.6815, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.849602222442627, + "learning_rate": 3.907392363931763e-06, + "loss": 1.744, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.7209551334381104, + "learning_rate": 3.90536149471974e-06, + "loss": 1.755, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.602299451828003, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.7835, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.8590067625045776, + "eval_runtime": 49.5182, + "eval_samples_per_second": 20.195, + "eval_steps_per_second": 0.848, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.504567861557007, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6595, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.456721305847168, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6692, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.319553852081299, + "learning_rate": 3.897238017871649e-06, + "loss": 1.6664, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.45987606048584, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.65, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 3.0049448013305664, + "learning_rate": 3.893176279447604e-06, + "loss": 1.7366, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.522472858428955, + "learning_rate": 3.891145410235581e-06, + "loss": 1.7022, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.3914318084716797, + "learning_rate": 3.889114541023558e-06, + "loss": 1.6591, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.462167501449585, + "learning_rate": 3.887083671811536e-06, + "loss": 1.6158, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.6707093715667725, + "learning_rate": 3.885052802599513e-06, + "loss": 1.7453, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.493745803833008, + "learning_rate": 3.88302193338749e-06, + "loss": 1.6836, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.8536967039108276, + "eval_runtime": 50.3456, + "eval_samples_per_second": 19.863, + "eval_steps_per_second": 0.834, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.5338919162750244, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.65, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.6954877376556396, + "learning_rate": 3.878960194963445e-06, + "loss": 1.7156, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.4878089427948, + "learning_rate": 3.876929325751422e-06, + "loss": 1.6053, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.4329843521118164, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6642, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 3.118771553039551, + "learning_rate": 3.872867587327377e-06, + "loss": 1.6392, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4644217491149902, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6566, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.4643497467041016, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6654, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.7067277431488037, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.682, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.8748488426208496, + "learning_rate": 3.864744110479286e-06, + "loss": 1.6886, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.632155179977417, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7069, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8592487573623657, + "eval_runtime": 47.0773, + "eval_samples_per_second": 21.242, + "eval_steps_per_second": 0.892, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.5379488468170166, + "learning_rate": 3.86068237205524e-06, + "loss": 1.6695, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.571742057800293, + "learning_rate": 3.858651502843217e-06, + "loss": 1.6235, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.5355312824249268, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6528, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.715510606765747, + "learning_rate": 3.854589764419171e-06, + "loss": 1.7379, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.635364532470703, + "learning_rate": 3.852558895207149e-06, + "loss": 1.6958, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.4713191986083984, + "learning_rate": 3.850528025995126e-06, + "loss": 1.6471, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.584200382232666, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6548, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.4639663696289062, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6614, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.4823923110961914, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.6523, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.953594923019409, + "learning_rate": 3.842404549147035e-06, + "loss": 1.6281, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8549022674560547, + "eval_runtime": 47.2215, + "eval_samples_per_second": 21.177, + "eval_steps_per_second": 0.889, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.524352550506592, + "learning_rate": 3.840373679935013e-06, + "loss": 1.6741, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.411515235900879, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.6571, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.6827003955841064, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6927, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 3.3486077785491943, + "learning_rate": 3.834281072298944e-06, + "loss": 1.6347, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.4518957138061523, + "learning_rate": 3.832250203086921e-06, + "loss": 1.7011, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.623842239379883, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7459, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.689602851867676, + "learning_rate": 3.828188464662876e-06, + "loss": 1.6812, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.8425960540771484, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6258, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.6639413833618164, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6769, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.4269535541534424, + "learning_rate": 3.822095857026808e-06, + "loss": 1.6363, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8489640951156616, + "eval_runtime": 45.897, + "eval_samples_per_second": 21.788, + "eval_steps_per_second": 0.915, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.3430261611938477, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6563, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.322094678878784, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6588, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.6172726154327393, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.681, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.574084997177124, + "learning_rate": 3.813972380178717e-06, + "loss": 1.6977, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.437417984008789, + "learning_rate": 3.811941510966694e-06, + "loss": 1.6204, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.319276809692383, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5858, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5540425777435303, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6622, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.567634344100952, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.635, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4815967082977295, + "learning_rate": 3.803818034118603e-06, + "loss": 1.7146, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.475162982940674, + "learning_rate": 3.80178716490658e-06, + "loss": 1.6911, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.848276138305664, + "eval_runtime": 44.1011, + "eval_samples_per_second": 22.675, + "eval_steps_per_second": 0.952, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5385165214538574, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.6523, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.443129301071167, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6239, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.696885347366333, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.7318, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.8570339679718018, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7198, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.475338935852051, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7052, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.5250487327575684, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.7015, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.812121868133545, + "learning_rate": 3.787571080422421e-06, + "loss": 1.6876, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.486117362976074, + "learning_rate": 3.785540211210398e-06, + "loss": 1.6695, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.5146725177764893, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.6585, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.5808019638061523, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.7608, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8411076068878174, + "eval_runtime": 48.0798, + "eval_samples_per_second": 20.799, + "eval_steps_per_second": 0.874, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.509993076324463, + "learning_rate": 3.77944760357433e-06, + "loss": 1.7101, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.492285966873169, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6417, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.809333086013794, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6963, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.524418354034424, + "learning_rate": 3.773354995938262e-06, + "loss": 1.6957, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.5848705768585205, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7193, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.8617639541625977, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.6721, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.4655680656433105, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.586, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 3.0633299350738525, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6792, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.3882107734680176, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.6229, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.5860483646392822, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6753, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8393676280975342, + "eval_runtime": 44.3838, + "eval_samples_per_second": 22.531, + "eval_steps_per_second": 0.946, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.930290460586548, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.6955, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.8278939723968506, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.6455, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.3822925090789795, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.7163, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.6904139518737793, + "learning_rate": 3.753046303818034e-06, + "loss": 1.6942, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.535284996032715, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7237, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.669031858444214, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.6391, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.5162742137908936, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6663, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.5937418937683105, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7115, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.4892170429229736, + "learning_rate": 3.742891957757921e-06, + "loss": 1.6727, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.7901015281677246, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7179, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8420112133026123, + "eval_runtime": 42.0904, + "eval_samples_per_second": 23.758, + "eval_steps_per_second": 0.998, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.510568618774414, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.6408, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.5084173679351807, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6495, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.3693647384643555, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.6064, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.520414352416992, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5668, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.544624090194702, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.6231, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.6257777214050293, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6093, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.0384273529052734, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.7291, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.718588352203369, + "learning_rate": 3.724614134849716e-06, + "loss": 1.604, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.655592918395996, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6993, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.569904088973999, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.6595, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8419173955917358, + "eval_runtime": 42.5535, + "eval_samples_per_second": 23.5, + "eval_steps_per_second": 0.987, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.643686056137085, + "learning_rate": 3.718521527213648e-06, + "loss": 1.6483, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.7431893348693848, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5418, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 3.0191054344177246, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.6425, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.695777654647827, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.651, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.512838840484619, + "learning_rate": 3.710398050365557e-06, + "loss": 1.6181, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.9957621097564697, + "learning_rate": 3.708367181153534e-06, + "loss": 1.6454, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.5258326530456543, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.5995, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.6610865592956543, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5816, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.825767993927002, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6402, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.605870008468628, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6282, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8433754444122314, + "eval_runtime": 45.2127, + "eval_samples_per_second": 22.118, + "eval_steps_per_second": 0.929, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.7334296703338623, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5958, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.9007229804992676, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.7128, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8827497959136963, + "learning_rate": 3.694151096669375e-06, + "loss": 1.6157, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6470541954040527, + "learning_rate": 3.692120227457352e-06, + "loss": 1.643, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.5725178718566895, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.5481, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.7779407501220703, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.6241, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 3.161653995513916, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.644, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.5995497703552246, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6475, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.607166290283203, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6573, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.4532339572906494, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.6044, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.8344279527664185, + "eval_runtime": 43.2125, + "eval_samples_per_second": 23.141, + "eval_steps_per_second": 0.972, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.540632486343384, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6308, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.40977144241333, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6134, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.6345880031585693, + "learning_rate": 3.673842404549147e-06, + "loss": 1.615, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.565157413482666, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6351, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.7868876457214355, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.7186, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.7061610221862793, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.6454, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.587785482406616, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6867, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.4618277549743652, + "learning_rate": 3.663688058489034e-06, + "loss": 1.6189, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.4425477981567383, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5772, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.7514469623565674, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6368, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8366374969482422, + "eval_runtime": 44.4679, + "eval_samples_per_second": 22.488, + "eval_steps_per_second": 0.945, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.529805898666382, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.6229, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6875312328338623, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6445, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.715970993041992, + "learning_rate": 3.65353371242892e-06, + "loss": 1.5318, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.8685660362243652, + "learning_rate": 3.651502843216897e-06, + "loss": 1.6465, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.6801602840423584, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.5959, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.61037015914917, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6561, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.478137493133545, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6435, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5708508491516113, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6206, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8857522010803223, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6269, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.6102395057678223, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.648, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.836978793144226, + "eval_runtime": 44.9452, + "eval_samples_per_second": 22.249, + "eval_steps_per_second": 0.934, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7924158573150635, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6802, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.784693479537964, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.6515, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.747894525527954, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6093, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.635850191116333, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6791, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.5760908126831055, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.5721, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5258584022521973, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5644, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.919623613357544, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.6258, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.8330483436584473, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6077, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8145744800567627, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6463, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.9187726974487305, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6348, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.837236762046814, + "eval_runtime": 45.4421, + "eval_samples_per_second": 22.006, + "eval_steps_per_second": 0.924, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.6741979122161865, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6441, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.556701898574829, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5523, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.6352901458740234, + "learning_rate": 3.612916328188465e-06, + "loss": 1.6545, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.5784106254577637, + "learning_rate": 3.610885458976442e-06, + "loss": 1.6649, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.7136101722717285, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.596, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.608877182006836, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.5923, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.783069372177124, + "learning_rate": 3.604792851340374e-06, + "loss": 1.6685, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.621882438659668, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6404, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.672128915786743, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5713, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.645395040512085, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.609, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8348923921585083, + "eval_runtime": 44.8379, + "eval_samples_per_second": 22.303, + "eval_steps_per_second": 0.937, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 3.610335111618042, + "learning_rate": 3.596669374492283e-06, + "loss": 1.6693, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.7391626834869385, + "learning_rate": 3.59463850528026e-06, + "loss": 1.5847, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.658418655395508, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6552, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.6508545875549316, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6201, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 4.237835884094238, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.6725, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.672144889831543, + "learning_rate": 3.586515028432169e-06, + "loss": 1.6666, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.619480609893799, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5671, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 3.150592565536499, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6118, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.946467161178589, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6499, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.6372971534729004, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5516, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8352437019348145, + "eval_runtime": 46.1704, + "eval_samples_per_second": 21.659, + "eval_steps_per_second": 0.91, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.6211962699890137, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6414, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.567579746246338, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.6632, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.4393808841705322, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6323, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.7005021572113037, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6016, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7400808334350586, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6327, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.922241449356079, + "learning_rate": 3.566206336311942e-06, + "loss": 1.5866, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.524975299835205, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5707, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.637805461883545, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6259, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7430527210235596, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.595, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.5548908710479736, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.625, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8332648277282715, + "eval_runtime": 48.5515, + "eval_samples_per_second": 20.597, + "eval_steps_per_second": 0.865, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.5396368503570557, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.655, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.523167848587036, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6171, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.6028409004211426, + "learning_rate": 3.551990251827783e-06, + "loss": 1.5872, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.607464075088501, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6176, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.510603427886963, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.5999, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.444693088531494, + "learning_rate": 3.545897644191714e-06, + "loss": 1.5791, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6724541187286377, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6358, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.64925217628479, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5793, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.647118091583252, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6399, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8507258892059326, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6046, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8305917978286743, + "eval_runtime": 43.2738, + "eval_samples_per_second": 23.109, + "eval_steps_per_second": 0.971, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.5572781562805176, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6395, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.530216693878174, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6493, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.564739465713501, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6075, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.5772671699523926, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.64, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.828692674636841, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6584, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.700568675994873, + "learning_rate": 3.525588952071487e-06, + "loss": 1.6595, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.666743755340576, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6483, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.551187515258789, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.5946, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.553602933883667, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.5997, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.5355968475341797, + "learning_rate": 3.517465475223396e-06, + "loss": 1.571, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8321614265441895, + "eval_runtime": 46.1746, + "eval_samples_per_second": 21.657, + "eval_steps_per_second": 0.91, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.707101345062256, + "learning_rate": 3.515434606011373e-06, + "loss": 1.6389, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.8770599365234375, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.6828, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.5841145515441895, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6264, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.697502851486206, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6081, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.620100259780884, + "learning_rate": 3.507311129163282e-06, + "loss": 1.5617, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.6503913402557373, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6194, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.830296754837036, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.616, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.6960535049438477, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5523, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.59023118019104, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6378, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.582906484603882, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6368, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8257222175598145, + "eval_runtime": 47.9573, + "eval_samples_per_second": 20.852, + "eval_steps_per_second": 0.876, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.6826765537261963, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5685, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6153650283813477, + "learning_rate": 3.493095044679123e-06, + "loss": 1.6948, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.607027769088745, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5197, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.493180990219116, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6159, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.6291017532348633, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5404, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.90544056892395, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6164, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.673076629638672, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6486, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.5843210220336914, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.6127, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.55414080619812, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6255, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.496095657348633, + "learning_rate": 3.476848090982941e-06, + "loss": 1.5941, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8262794017791748, + "eval_runtime": 47.0577, + "eval_samples_per_second": 21.25, + "eval_steps_per_second": 0.893, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.597395658493042, + "learning_rate": 3.474817221770918e-06, + "loss": 1.6941, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.4487929344177246, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6417, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.63041353225708, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.611, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.481630325317383, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.575, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.7625951766967773, + "learning_rate": 3.466693744922827e-06, + "loss": 1.5988, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.663846492767334, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6326, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.594536781311035, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6004, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.5871670246124268, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6378, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.591907262802124, + "learning_rate": 3.458570268074736e-06, + "loss": 1.6093, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.456110954284668, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6014, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8266254663467407, + "eval_runtime": 47.4255, + "eval_samples_per_second": 21.086, + "eval_steps_per_second": 0.886, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.8052752017974854, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.5679, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.7458856105804443, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6239, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.4829750061035156, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.5692, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.5427238941192627, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6214, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.597419261932373, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6012, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.5192158222198486, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6469, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.4233906269073486, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5822, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.4213144779205322, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5345, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.763439655303955, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6092, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.228456497192383, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6591, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8253848552703857, + "eval_runtime": 50.1479, + "eval_samples_per_second": 19.941, + "eval_steps_per_second": 0.838, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.670700788497925, + "learning_rate": 3.434199837530463e-06, + "loss": 1.612, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.797170877456665, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5624, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8155477046966553, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6535, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.7093656063079834, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6479, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.6133127212524414, + "learning_rate": 3.426076360682372e-06, + "loss": 1.6669, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.471862554550171, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6136, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7078938484191895, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5382, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.6206846237182617, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6186, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.674201011657715, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.6805, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.7543511390686035, + "learning_rate": 3.415922014622259e-06, + "loss": 1.644, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8275290727615356, + "eval_runtime": 49.1382, + "eval_samples_per_second": 20.351, + "eval_steps_per_second": 0.855, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.934906244277954, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6502, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.6866018772125244, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6324, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.743281126022339, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6229, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.7543251514434814, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.5813, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8038830757141113, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6482, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.6368162631988525, + "learning_rate": 3.403736799350122e-06, + "loss": 1.618, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7447240352630615, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.5735, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.6714727878570557, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.686, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.7939751148223877, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6104, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.624718189239502, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6194, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8199249505996704, + "eval_runtime": 50.4903, + "eval_samples_per_second": 19.806, + "eval_steps_per_second": 0.832, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.563082695007324, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5669, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.6971375942230225, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6237, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.7873635292053223, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6241, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.570742130279541, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6304, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8116018772125244, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6068, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.6097946166992188, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.5996, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.593461275100708, + "learning_rate": 3.381397238017872e-06, + "loss": 1.5935, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.008805274963379, + "learning_rate": 3.379366368805849e-06, + "loss": 1.68, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7567968368530273, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6129, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.647456645965576, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.618, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8231184482574463, + "eval_runtime": 46.6144, + "eval_samples_per_second": 21.453, + "eval_steps_per_second": 0.901, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.842937707901001, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6118, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.794092893600464, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.5922, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.837575674057007, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6267, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.7084648609161377, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5597, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.626157283782959, + "learning_rate": 3.36515028432169e-06, + "loss": 1.5961, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.589062213897705, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6235, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.681427478790283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6568, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 3.3202719688415527, + "learning_rate": 3.359057676685622e-06, + "loss": 1.649, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.463998794555664, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5482, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.5475661754608154, + "learning_rate": 3.354995938261576e-06, + "loss": 1.586, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8188257217407227, + "eval_runtime": 51.7434, + "eval_samples_per_second": 19.326, + "eval_steps_per_second": 0.812, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.579566478729248, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5546, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6778531074523926, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.5831, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.98976731300354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6199, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.8345465660095215, + "learning_rate": 3.346872461413485e-06, + "loss": 1.5697, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.6166720390319824, + "learning_rate": 3.344841592201463e-06, + "loss": 1.5829, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.5504395961761475, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.5845, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.6515390872955322, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6154, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.589524269104004, + "learning_rate": 3.338748984565394e-06, + "loss": 1.639, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.8798511028289795, + "learning_rate": 3.336718115353372e-06, + "loss": 1.676, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.984379529953003, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6154, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.818778157234192, + "eval_runtime": 44.8621, + "eval_samples_per_second": 22.291, + "eval_steps_per_second": 0.936, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.5935728549957275, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.5895, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.7658374309539795, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6041, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.580352306365967, + "learning_rate": 3.32859463850528e-06, + "loss": 1.6127, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.689892292022705, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5624, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.708078622817993, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.5782, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.7298295497894287, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.5815, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8700411319732666, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5597, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.6671183109283447, + "learning_rate": 3.318440292445167e-06, + "loss": 1.5902, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.11922550201416, + "learning_rate": 3.316409423233144e-06, + "loss": 1.5678, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.72869610786438, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6568, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8164011240005493, + "eval_runtime": 47.4229, + "eval_samples_per_second": 21.087, + "eval_steps_per_second": 0.886, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.7037997245788574, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6191, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.6406683921813965, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6018, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.4749815464019775, + "learning_rate": 3.308285946385053e-06, + "loss": 1.679, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.5094757080078125, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5455, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.6685080528259277, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6123, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7547595500946045, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.6497, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.7807092666625977, + "learning_rate": 3.300162469536962e-06, + "loss": 1.5903, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.903127431869507, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6398, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.4666707515716553, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6194, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.9508626461029053, + "learning_rate": 3.294069861900894e-06, + "loss": 1.6774, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8119163513183594, + "eval_runtime": 51.0141, + "eval_samples_per_second": 19.602, + "eval_steps_per_second": 0.823, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.585411787033081, + "learning_rate": 3.292038992688871e-06, + "loss": 1.6086, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.7636921405792236, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.6386, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.72711181640625, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.6563, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.5698368549346924, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5501, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8764665126800537, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6229, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.653674840927124, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5343, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.896165609359741, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.6617, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.714388847351074, + "learning_rate": 3.277822908204712e-06, + "loss": 1.605, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.572202682495117, + "learning_rate": 3.275792038992689e-06, + "loss": 1.568, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.719407081604004, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6116, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.811676263809204, + "eval_runtime": 46.0392, + "eval_samples_per_second": 21.721, + "eval_steps_per_second": 0.912, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.752934694290161, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.5672, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.519256353378296, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6101, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.758317708969116, + "learning_rate": 3.267668562144598e-06, + "loss": 1.597, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.6348824501037598, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.5731, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.699477434158325, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6288, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.5104498863220215, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6024, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.680694341659546, + "learning_rate": 3.259545085296507e-06, + "loss": 1.6689, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.459789276123047, + "learning_rate": 3.257514216084484e-06, + "loss": 1.56, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.533125877380371, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.6734, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.640742301940918, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6214, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8131158351898193, + "eval_runtime": 47.2273, + "eval_samples_per_second": 21.174, + "eval_steps_per_second": 0.889, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.4593327045440674, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.5997, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.6157352924346924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6193, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.5011303424835205, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.5796, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.545781373977661, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.6128, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.5882468223571777, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6106, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.7500388622283936, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5491, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5432305335998535, + "learning_rate": 3.23923639317628e-06, + "loss": 1.5879, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.5674941539764404, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.5737, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.630491256713867, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6167, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.7517776489257812, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.6698, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8172497749328613, + "eval_runtime": 45.7554, + "eval_samples_per_second": 21.855, + "eval_steps_per_second": 0.918, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5155255794525146, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5647, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.5586776733398438, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6403, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7582011222839355, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5901, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.611339569091797, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.599, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.7254250049591064, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6167, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.5621025562286377, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5241, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849069118499756, + "learning_rate": 3.218927701056052e-06, + "loss": 1.5823, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.874762535095215, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5294, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.706822395324707, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.6529, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.7066354751586914, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5591, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.814163327217102, + "eval_runtime": 43.7518, + "eval_samples_per_second": 22.856, + "eval_steps_per_second": 0.96, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.584439516067505, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.5986, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.671645402908325, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6417, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6757681369781494, + "learning_rate": 3.206742485783916e-06, + "loss": 1.5594, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.6210391521453857, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5476, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.5632853507995605, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6165, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.64302921295166, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.631, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.4642136096954346, + "learning_rate": 3.198619008935825e-06, + "loss": 1.534, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.196588139723802e-06, + "loss": 1.598, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.5753607749938965, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6177, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.787393093109131, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.607, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.809692621231079, + "eval_runtime": 44.1276, + "eval_samples_per_second": 22.662, + "eval_steps_per_second": 0.952, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.4849026203155518, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6064, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.484736204147339, + "learning_rate": 3.188464662875711e-06, + "loss": 1.6591, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.652766466140747, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5931, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.6450769901275635, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5463, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.736511468887329, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5509, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.8212637901306152, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6244, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.7004551887512207, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6123, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.665766954421997, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.5904, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.7625415325164795, + "learning_rate": 3.174248578391552e-06, + "loss": 1.5843, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.5376832485198975, + "learning_rate": 3.172217709179529e-06, + "loss": 1.5758, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8079267740249634, + "eval_runtime": 45.139, + "eval_samples_per_second": 22.154, + "eval_steps_per_second": 0.93, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.704226016998291, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.5891, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6165294647216797, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5502, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.6848294734954834, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6134, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.778846502304077, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.611, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.647531747817993, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5531, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.605592966079712, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6011, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5820205211639404, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6121, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.7092363834381104, + "learning_rate": 3.155970755483347e-06, + "loss": 1.5928, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.7473032474517822, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6068, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.742840528488159, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6041, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8063452243804932, + "eval_runtime": 44.2112, + "eval_samples_per_second": 22.619, + "eval_steps_per_second": 0.95, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.7613816261291504, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6413, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.876913547515869, + "learning_rate": 3.147847278635256e-06, + "loss": 1.6848, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5945441722869873, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.5847, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.526456117630005, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5496, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.9870903491973877, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.5982, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6774742603302, + "learning_rate": 3.139723801787165e-06, + "loss": 1.5678, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.69058895111084, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6174, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.8981339931488037, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5529, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.522247552871704, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6061, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.709726333618164, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6267, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8052302598953247, + "eval_runtime": 42.6337, + "eval_samples_per_second": 23.456, + "eval_steps_per_second": 0.985, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.618793249130249, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6441, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5927956104278564, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6035, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.557023048400879, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6252, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5910511016845703, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.5883, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.6681272983551025, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6147, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.7024455070495605, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6317, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.6808700561523438, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.5871, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.585388422012329, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.6202, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.637092113494873, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6357, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.652611494064331, + "learning_rate": 3.111291632818847e-06, + "loss": 1.651, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.804667592048645, + "eval_runtime": 43.1932, + "eval_samples_per_second": 23.152, + "eval_steps_per_second": 0.972, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.770113945007324, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6005, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.570833206176758, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6169, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.537957191467285, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.5826, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.715669631958008, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6276, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.5450210571289062, + "learning_rate": 3.101137286758733e-06, + "loss": 1.6305, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.6358814239501953, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6367, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.7380471229553223, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.5984, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 3.030034303665161, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.6661, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.6777493953704834, + "learning_rate": 3.093013809910642e-06, + "loss": 1.6122, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8807458877563477, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.6514, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8064097166061401, + "eval_runtime": 43.3673, + "eval_samples_per_second": 23.059, + "eval_steps_per_second": 0.968, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.545091390609741, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6182, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.6389622688293457, + "learning_rate": 3.086921202274574e-06, + "loss": 1.5931, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.5784780979156494, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5579, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.67499041557312, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6128, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.558598756790161, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.5958, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6460611820220947, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5274, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.770334482192993, + "learning_rate": 3.07676685621446e-06, + "loss": 1.604, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6441917419433594, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.567, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.6142544746398926, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.5891, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.6242077350616455, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6059, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8036692142486572, + "eval_runtime": 45.6003, + "eval_samples_per_second": 21.93, + "eval_steps_per_second": 0.921, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.417332410812378, + "learning_rate": 3.068643379366369e-06, + "loss": 1.5991, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.4537513256073, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5603, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.745912551879883, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6171, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.636627674102783, + "learning_rate": 3.062550771730301e-06, + "loss": 1.5839, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.6001460552215576, + "learning_rate": 3.060519902518278e-06, + "loss": 1.5781, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.572185516357422, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5188, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.87540864944458, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.621, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.7670910358428955, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.584, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.672682523727417, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6293, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.655538558959961, + "learning_rate": 3.050365556458164e-06, + "loss": 1.5773, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8046640157699585, + "eval_runtime": 44.6243, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.582319974899292, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.5751, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.7160403728485107, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.5972, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.6800436973571777, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6311, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.8212592601776123, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6315, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.57470703125, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5414, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 3.3979809284210205, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.5732, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.606290102005005, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.5966, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.5241527557373047, + "learning_rate": 3.034118602761982e-06, + "loss": 1.588, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.7115085124969482, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5619, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.7072222232818604, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6346, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.804133415222168, + "eval_runtime": 44.5547, + "eval_samples_per_second": 22.444, + "eval_steps_per_second": 0.943, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.5874669551849365, + "learning_rate": 3.028025995125914e-06, + "loss": 1.623, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.644073247909546, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5484, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.553112506866455, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.5923, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6504719257354736, + "learning_rate": 3.021933387489846e-06, + "loss": 1.543, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.7608838081359863, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6001, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.690267562866211, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6093, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.8565986156463623, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5614, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.672799587249756, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5657, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5756146907806396, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.5664, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.8423633575439453, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.5936, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.800963044166565, + "eval_runtime": 46.5083, + "eval_samples_per_second": 21.502, + "eval_steps_per_second": 0.903, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.6263201236724854, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.5693, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 3.0006930828094482, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6084, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.7002620697021484, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.5667, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.722398519515991, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6285, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.738555908203125, + "learning_rate": 2.999593826157596e-06, + "loss": 1.5849, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.726271390914917, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.5951, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.6749048233032227, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.5807, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.774726629257202, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6355, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6740102767944336, + "learning_rate": 2.991470349309505e-06, + "loss": 1.5871, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.7571794986724854, + "learning_rate": 2.989439480097482e-06, + "loss": 1.578, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.7960126399993896, + "eval_runtime": 44.4777, + "eval_samples_per_second": 22.483, + "eval_steps_per_second": 0.944, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.67576265335083, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.5917, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.8302817344665527, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6313, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.874002456665039, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6212, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6992602348327637, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6011, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.726292371749878, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.5981, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.988934278488159, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6272, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.6781861782073975, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.5742, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.818401336669922, + "learning_rate": 2.9731925264013e-06, + "loss": 1.5713, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.6900908946990967, + "learning_rate": 2.971161657189277e-06, + "loss": 1.5754, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.9017207622528076, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.5746, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.7994096279144287, + "eval_runtime": 46.997, + "eval_samples_per_second": 21.278, + "eval_steps_per_second": 0.894, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.0305914878845215, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.5733, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.695298910140991, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6379, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.60552716255188, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5488, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.664668321609497, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6256, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.700955390930176, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6228, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.637998104095459, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.5565, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.664620876312256, + "learning_rate": 2.954914703493095e-06, + "loss": 1.5805, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.8120782375335693, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.5615, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.641434907913208, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.56, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.845808982849121, + "learning_rate": 2.948822095857027e-06, + "loss": 1.5954, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8004399538040161, + "eval_runtime": 47.1938, + "eval_samples_per_second": 21.189, + "eval_steps_per_second": 0.89, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.713960886001587, + "learning_rate": 2.946791226645004e-06, + "loss": 1.5628, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.8967084884643555, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.5445, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 3.3338470458984375, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.5735, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.864295721054077, + "learning_rate": 2.940698619008936e-06, + "loss": 1.5892, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.704193353652954, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.5916, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.6473755836486816, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.6059, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.610670804977417, + "learning_rate": 2.934606011372868e-06, + "loss": 1.6012, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.6023919582366943, + "learning_rate": 2.932575142160845e-06, + "loss": 1.5973, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8217804431915283, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.5794, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.8418939113616943, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.5942, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.8003290891647339, + "eval_runtime": 44.3696, + "eval_samples_per_second": 22.538, + "eval_steps_per_second": 0.947, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.6487412452697754, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.6297, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 2.6183269023895264, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.5916, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.6346395015716553, + "learning_rate": 2.922420796100731e-06, + "loss": 1.5209, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.631850242614746, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.6268, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.7087175846099854, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.5348, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.70473575592041, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.5943, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 3.0979015827178955, + "learning_rate": 2.91429731925264e-06, + "loss": 1.6308, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.6136019229888916, + "learning_rate": 2.912266450040618e-06, + "loss": 1.5955, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.6788594722747803, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.5909, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 3.681739330291748, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.5794, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.7974433898925781, + "eval_runtime": 50.5413, + "eval_samples_per_second": 19.786, + "eval_steps_per_second": 0.831, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.543687582015991, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.5455, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 3.0568184852600098, + "learning_rate": 2.904142973192527e-06, + "loss": 1.6445, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.557318925857544, + "learning_rate": 2.902112103980504e-06, + "loss": 1.5933, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.4658801555633545, + "learning_rate": 2.900081234768481e-06, + "loss": 1.5899, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.610834836959839, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.5557, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.643775224685669, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.5523, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.8988914489746094, + "learning_rate": 2.893988627132413e-06, + "loss": 1.5787, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.686321973800659, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.5205, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.6282148361206055, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.5942, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.581450939178467, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.5779, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.7954648733139038, + "eval_runtime": 45.0426, + "eval_samples_per_second": 22.201, + "eval_steps_per_second": 0.932, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.6617720127105713, + "learning_rate": 2.885865150284322e-06, + "loss": 1.6168, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.643260955810547, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.5764, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.5922341346740723, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.5385, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.7868714332580566, + "learning_rate": 2.879772542648254e-06, + "loss": 1.5905, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.6874098777770996, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.516, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.9882795810699463, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.6497, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.574932336807251, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.6217, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 2.625720739364624, + "learning_rate": 2.871649065800163e-06, + "loss": 1.6436, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 2.6322028636932373, + "learning_rate": 2.86961819658814e-06, + "loss": 1.5743, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.753891944885254, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.594, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.7947734594345093, + "eval_runtime": 47.0036, + "eval_samples_per_second": 21.275, + "eval_steps_per_second": 0.894, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.6878788471221924, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.5801, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.7409934997558594, + "learning_rate": 2.863525588952072e-06, + "loss": 1.6251, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.517815351486206, + "learning_rate": 2.861494719740049e-06, + "loss": 1.5055, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.7027196884155273, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.5536, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.8066604137420654, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.5819, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.7119011878967285, + "learning_rate": 2.855402112103981e-06, + "loss": 1.5776, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.6952199935913086, + "learning_rate": 2.853371242891958e-06, + "loss": 1.5682, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.7582387924194336, + "learning_rate": 2.851340373679935e-06, + "loss": 1.5863, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.8032495975494385, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.6204, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.6014182567596436, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.5816, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.7986392974853516, + "eval_runtime": 45.7419, + "eval_samples_per_second": 21.862, + "eval_steps_per_second": 0.918, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.611240863800049, + "learning_rate": 2.845247766043867e-06, + "loss": 1.5563, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.604238510131836, + "learning_rate": 2.843216896831844e-06, + "loss": 1.5143, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.531214475631714, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.5579, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.6469998359680176, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.6049, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 3.1636979579925537, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.6072, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.6377527713775635, + "learning_rate": 2.835093419983753e-06, + "loss": 1.574, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.6896755695343018, + "learning_rate": 2.833062550771731e-06, + "loss": 1.5498, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.6172642707824707, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.5413, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.7607314586639404, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.5816, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.5872104167938232, + "learning_rate": 2.826969943135662e-06, + "loss": 1.5868, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.7989377975463867, + "eval_runtime": 46.6112, + "eval_samples_per_second": 21.454, + "eval_steps_per_second": 0.901, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.517178773880005, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.5875, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.531451940536499, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.5583, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.6367409229278564, + "learning_rate": 2.820877335499594e-06, + "loss": 1.5605, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.496851682662964, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.5227, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.583904981613159, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.5434, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.7455286979675293, + "learning_rate": 2.814784727863526e-06, + "loss": 1.5732, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.6893272399902344, + "learning_rate": 2.812753858651503e-06, + "loss": 1.5807, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 2.6912310123443604, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.5484, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 2.6918725967407227, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.535, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 2.808300256729126, + "learning_rate": 2.806661251015435e-06, + "loss": 1.6278, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.8001086711883545, + "eval_runtime": 50.3556, + "eval_samples_per_second": 19.859, + "eval_steps_per_second": 0.834, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.702766180038452, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.6223, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 2.890214681625366, + "learning_rate": 2.802599512591389e-06, + "loss": 1.68, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.6849842071533203, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.5836, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.678279161453247, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.529, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 2.782949686050415, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.5814, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.8183817863464355, + "learning_rate": 2.794476035743298e-06, + "loss": 1.6401, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.917726516723633, + "learning_rate": 2.792445166531276e-06, + "loss": 1.5936, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 2.8861331939697266, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.6213, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.7103989124298096, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.5537, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.876448392868042, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.5464, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.7939362525939941, + "eval_runtime": 47.8308, + "eval_samples_per_second": 20.907, + "eval_steps_per_second": 0.878, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.613130569458008, + "learning_rate": 2.784321689683185e-06, + "loss": 1.5427, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.553765058517456, + "learning_rate": 2.782290820471162e-06, + "loss": 1.5496, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.7586376667022705, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.6221, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.6437292098999023, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.6075, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.6939804553985596, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.5918, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.6177403926849365, + "learning_rate": 2.774167343623071e-06, + "loss": 1.6353, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 2.8386614322662354, + "learning_rate": 2.772136474411048e-06, + "loss": 1.5548, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.6705095767974854, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.497, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.790919065475464, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.5945, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.7193119525909424, + "learning_rate": 2.76604386677498e-06, + "loss": 1.5627, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.795873999595642, + "eval_runtime": 49.9423, + "eval_samples_per_second": 20.023, + "eval_steps_per_second": 0.841, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 2.871537208557129, + "learning_rate": 2.764012997562957e-06, + "loss": 1.6027, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.6842691898345947, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.5327, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 2.6385788917541504, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.6181, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.6933207511901855, + "learning_rate": 2.757920389926889e-06, + "loss": 1.5561, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 2.814086437225342, + "learning_rate": 2.755889520714866e-06, + "loss": 1.5519, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.6112890243530273, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.52, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.7777466773986816, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.5885, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.782341241836548, + "learning_rate": 2.749796913078798e-06, + "loss": 1.609, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.576822519302368, + "learning_rate": 2.747766043866775e-06, + "loss": 1.5808, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.6026647090911865, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.626, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.7908697128295898, + "eval_runtime": 47.8943, + "eval_samples_per_second": 20.879, + "eval_steps_per_second": 0.877, + "step": 1400 + }, + { + "epoch": 1.5353424657534247, + "grad_norm": 2.645002603530884, + "learning_rate": 2.7437043054427297e-06, + "loss": 1.6664, + "step": 1401 + }, + { + "epoch": 1.5364383561643835, + "grad_norm": 2.676917552947998, + "learning_rate": 2.741673436230707e-06, + "loss": 1.572, + "step": 1402 + }, + { + "epoch": 1.5375342465753423, + "grad_norm": 2.6958320140838623, + "learning_rate": 2.739642567018684e-06, + "loss": 1.5712, + "step": 1403 + }, + { + "epoch": 1.5386301369863014, + "grad_norm": 2.8492650985717773, + "learning_rate": 2.7376116978066615e-06, + "loss": 1.5796, + "step": 1404 + }, + { + "epoch": 1.5397260273972604, + "grad_norm": 2.7714121341705322, + "learning_rate": 2.7355808285946388e-06, + "loss": 1.568, + "step": 1405 + }, + { + "epoch": 1.5408219178082192, + "grad_norm": 2.6548564434051514, + "learning_rate": 2.733549959382616e-06, + "loss": 1.5754, + "step": 1406 + }, + { + "epoch": 1.541917808219178, + "grad_norm": 2.6842844486236572, + "learning_rate": 2.7315190901705933e-06, + "loss": 1.6325, + "step": 1407 + }, + { + "epoch": 1.543013698630137, + "grad_norm": 2.700601816177368, + "learning_rate": 2.7294882209585706e-06, + "loss": 1.5986, + "step": 1408 + }, + { + "epoch": 1.5441095890410959, + "grad_norm": 2.5768840312957764, + "learning_rate": 2.727457351746548e-06, + "loss": 1.5543, + "step": 1409 + }, + { + "epoch": 1.5452054794520547, + "grad_norm": 2.633925199508667, + "learning_rate": 2.725426482534525e-06, + "loss": 1.5324, + "step": 1410 + }, + { + "epoch": 1.5452054794520547, + "eval_loss": 1.793561577796936, + "eval_runtime": 44.8327, + "eval_samples_per_second": 22.305, + "eval_steps_per_second": 0.937, + "step": 1410 + }, + { + "epoch": 1.5463013698630137, + "grad_norm": 2.5587072372436523, + "learning_rate": 2.723395613322502e-06, + "loss": 1.5725, + "step": 1411 + }, + { + "epoch": 1.5473972602739727, + "grad_norm": 2.8127360343933105, + "learning_rate": 2.7213647441104797e-06, + "loss": 1.5697, + "step": 1412 + }, + { + "epoch": 1.5484931506849315, + "grad_norm": 2.649578332901001, + "learning_rate": 2.7193338748984565e-06, + "loss": 1.5611, + "step": 1413 + }, + { + "epoch": 1.5495890410958904, + "grad_norm": 2.792600393295288, + "learning_rate": 2.7173030056864342e-06, + "loss": 1.7033, + "step": 1414 + }, + { + "epoch": 1.5506849315068494, + "grad_norm": 2.945256233215332, + "learning_rate": 2.715272136474411e-06, + "loss": 1.5325, + "step": 1415 + }, + { + "epoch": 1.5517808219178082, + "grad_norm": 2.815673589706421, + "learning_rate": 2.713241267262389e-06, + "loss": 1.5764, + "step": 1416 + }, + { + "epoch": 1.552876712328767, + "grad_norm": 2.5527024269104004, + "learning_rate": 2.7112103980503656e-06, + "loss": 1.5647, + "step": 1417 + }, + { + "epoch": 1.553972602739726, + "grad_norm": 2.725724697113037, + "learning_rate": 2.7091795288383433e-06, + "loss": 1.5979, + "step": 1418 + }, + { + "epoch": 1.555068493150685, + "grad_norm": 2.7697906494140625, + "learning_rate": 2.70714865962632e-06, + "loss": 1.5521, + "step": 1419 + }, + { + "epoch": 1.5561643835616439, + "grad_norm": 2.5871148109436035, + "learning_rate": 2.705117790414298e-06, + "loss": 1.5812, + "step": 1420 + }, + { + "epoch": 1.5561643835616439, + "eval_loss": 1.798141360282898, + "eval_runtime": 51.1729, + "eval_samples_per_second": 19.542, + "eval_steps_per_second": 0.821, + "step": 1420 + }, + { + "epoch": 1.5572602739726027, + "grad_norm": 2.5430808067321777, + "learning_rate": 2.7030869212022747e-06, + "loss": 1.6085, + "step": 1421 + }, + { + "epoch": 1.5583561643835617, + "grad_norm": 2.878905773162842, + "learning_rate": 2.7010560519902516e-06, + "loss": 1.5544, + "step": 1422 + }, + { + "epoch": 1.5594520547945205, + "grad_norm": 2.7553086280822754, + "learning_rate": 2.6990251827782293e-06, + "loss": 1.5712, + "step": 1423 + }, + { + "epoch": 1.5605479452054793, + "grad_norm": 2.636136054992676, + "learning_rate": 2.696994313566206e-06, + "loss": 1.5574, + "step": 1424 + }, + { + "epoch": 1.5616438356164384, + "grad_norm": 2.8743226528167725, + "learning_rate": 2.694963444354184e-06, + "loss": 1.6279, + "step": 1425 + }, + { + "epoch": 1.5627397260273974, + "grad_norm": 2.7313430309295654, + "learning_rate": 2.6929325751421607e-06, + "loss": 1.5717, + "step": 1426 + }, + { + "epoch": 1.563835616438356, + "grad_norm": 2.6373813152313232, + "learning_rate": 2.6909017059301384e-06, + "loss": 1.5935, + "step": 1427 + }, + { + "epoch": 1.564931506849315, + "grad_norm": 2.7440435886383057, + "learning_rate": 2.6888708367181153e-06, + "loss": 1.5402, + "step": 1428 + }, + { + "epoch": 1.566027397260274, + "grad_norm": 2.5901665687561035, + "learning_rate": 2.686839967506093e-06, + "loss": 1.5946, + "step": 1429 + }, + { + "epoch": 1.5671232876712329, + "grad_norm": 2.680370807647705, + "learning_rate": 2.68480909829407e-06, + "loss": 1.5068, + "step": 1430 + }, + { + "epoch": 1.5671232876712329, + "eval_loss": 1.7998759746551514, + "eval_runtime": 49.5336, + "eval_samples_per_second": 20.188, + "eval_steps_per_second": 0.848, + "step": 1430 + }, + { + "epoch": 1.5682191780821917, + "grad_norm": 2.6154282093048096, + "learning_rate": 2.6827782290820475e-06, + "loss": 1.4984, + "step": 1431 + }, + { + "epoch": 1.5693150684931507, + "grad_norm": 2.981612205505371, + "learning_rate": 2.6807473598700244e-06, + "loss": 1.6224, + "step": 1432 + }, + { + "epoch": 1.5704109589041098, + "grad_norm": 3.0227465629577637, + "learning_rate": 2.678716490658002e-06, + "loss": 1.5942, + "step": 1433 + }, + { + "epoch": 1.5715068493150683, + "grad_norm": 2.803877115249634, + "learning_rate": 2.676685621445979e-06, + "loss": 1.5821, + "step": 1434 + }, + { + "epoch": 1.5726027397260274, + "grad_norm": 2.7260076999664307, + "learning_rate": 2.674654752233956e-06, + "loss": 1.5788, + "step": 1435 + }, + { + "epoch": 1.5736986301369864, + "grad_norm": 2.672656297683716, + "learning_rate": 2.6726238830219335e-06, + "loss": 1.5798, + "step": 1436 + }, + { + "epoch": 1.5747945205479452, + "grad_norm": 2.7088775634765625, + "learning_rate": 2.6705930138099107e-06, + "loss": 1.6142, + "step": 1437 + }, + { + "epoch": 1.575890410958904, + "grad_norm": 2.5411477088928223, + "learning_rate": 2.668562144597888e-06, + "loss": 1.4999, + "step": 1438 + }, + { + "epoch": 1.576986301369863, + "grad_norm": 2.5658116340637207, + "learning_rate": 2.6665312753858653e-06, + "loss": 1.5898, + "step": 1439 + }, + { + "epoch": 1.5780821917808219, + "grad_norm": 2.626997232437134, + "learning_rate": 2.6645004061738426e-06, + "loss": 1.5795, + "step": 1440 + }, + { + "epoch": 1.5780821917808219, + "eval_loss": 1.7931181192398071, + "eval_runtime": 49.7844, + "eval_samples_per_second": 20.087, + "eval_steps_per_second": 0.844, + "step": 1440 + }, + { + "epoch": 1.5791780821917807, + "grad_norm": 2.737011194229126, + "learning_rate": 2.66246953696182e-06, + "loss": 1.5275, + "step": 1441 + }, + { + "epoch": 1.5802739726027397, + "grad_norm": 2.6428184509277344, + "learning_rate": 2.660438667749797e-06, + "loss": 1.5429, + "step": 1442 + }, + { + "epoch": 1.5813698630136988, + "grad_norm": 2.6428184509277344, + "learning_rate": 2.660438667749797e-06, + "loss": 1.5771, + "step": 1443 + }, + { + "epoch": 1.5824657534246576, + "grad_norm": 3.038092613220215, + "learning_rate": 2.6584077985377744e-06, + "loss": 1.579, + "step": 1444 + }, + { + "epoch": 1.5835616438356164, + "grad_norm": 2.4511592388153076, + "learning_rate": 2.6563769293257517e-06, + "loss": 1.5409, + "step": 1445 + }, + { + "epoch": 1.5846575342465754, + "grad_norm": 2.576601505279541, + "learning_rate": 2.654346060113729e-06, + "loss": 1.5411, + "step": 1446 + }, + { + "epoch": 1.5857534246575342, + "grad_norm": 2.54603910446167, + "learning_rate": 2.6523151909017062e-06, + "loss": 1.5455, + "step": 1447 + }, + { + "epoch": 1.586849315068493, + "grad_norm": 2.6850075721740723, + "learning_rate": 2.6502843216896835e-06, + "loss": 1.6355, + "step": 1448 + }, + { + "epoch": 1.587945205479452, + "grad_norm": 2.625603675842285, + "learning_rate": 2.6482534524776603e-06, + "loss": 1.5954, + "step": 1449 + }, + { + "epoch": 1.589041095890411, + "grad_norm": 2.755312919616699, + "learning_rate": 2.646222583265638e-06, + "loss": 1.5867, + "step": 1450 + }, + { + "epoch": 1.589041095890411, + "eval_loss": 1.7949016094207764, + "eval_runtime": 48.7453, + "eval_samples_per_second": 20.515, + "eval_steps_per_second": 0.862, + "step": 1450 + }, + { + "epoch": 1.59013698630137, + "grad_norm": 2.6282010078430176, + "learning_rate": 2.644191714053615e-06, + "loss": 1.5539, + "step": 1451 + }, + { + "epoch": 1.5912328767123287, + "grad_norm": 2.5556142330169678, + "learning_rate": 2.6421608448415926e-06, + "loss": 1.5324, + "step": 1452 + }, + { + "epoch": 1.5923287671232877, + "grad_norm": 2.9253478050231934, + "learning_rate": 2.6401299756295694e-06, + "loss": 1.58, + "step": 1453 + }, + { + "epoch": 1.5934246575342466, + "grad_norm": 2.8228557109832764, + "learning_rate": 2.638099106417547e-06, + "loss": 1.5671, + "step": 1454 + }, + { + "epoch": 1.5945205479452054, + "grad_norm": 2.7850286960601807, + "learning_rate": 2.636068237205524e-06, + "loss": 1.6302, + "step": 1455 + }, + { + "epoch": 1.5956164383561644, + "grad_norm": 2.5926599502563477, + "learning_rate": 2.6340373679935017e-06, + "loss": 1.4949, + "step": 1456 + }, + { + "epoch": 1.5967123287671234, + "grad_norm": 2.7452011108398438, + "learning_rate": 2.6320064987814785e-06, + "loss": 1.5828, + "step": 1457 + }, + { + "epoch": 1.5978082191780822, + "grad_norm": 2.626157283782959, + "learning_rate": 2.6299756295694562e-06, + "loss": 1.6194, + "step": 1458 + }, + { + "epoch": 1.598904109589041, + "grad_norm": 2.7487716674804688, + "learning_rate": 2.627944760357433e-06, + "loss": 1.635, + "step": 1459 + }, + { + "epoch": 1.6, + "grad_norm": 2.738173723220825, + "learning_rate": 2.6259138911454104e-06, + "loss": 1.615, + "step": 1460 + }, + { + "epoch": 1.6, + "eval_loss": 1.789223313331604, + "eval_runtime": 45.8318, + "eval_samples_per_second": 21.819, + "eval_steps_per_second": 0.916, + "step": 1460 + }, + { + "epoch": 1.601095890410959, + "grad_norm": 2.535860061645508, + "learning_rate": 2.6238830219333877e-06, + "loss": 1.5307, + "step": 1461 + }, + { + "epoch": 1.6021917808219177, + "grad_norm": 2.745260000228882, + "learning_rate": 2.621852152721365e-06, + "loss": 1.5883, + "step": 1462 + }, + { + "epoch": 1.6032876712328767, + "grad_norm": 2.7904720306396484, + "learning_rate": 2.619821283509342e-06, + "loss": 1.5309, + "step": 1463 + }, + { + "epoch": 1.6043835616438358, + "grad_norm": 2.70133900642395, + "learning_rate": 2.6177904142973195e-06, + "loss": 1.5906, + "step": 1464 + }, + { + "epoch": 1.6054794520547944, + "grad_norm": 2.61118745803833, + "learning_rate": 2.6157595450852968e-06, + "loss": 1.6168, + "step": 1465 + }, + { + "epoch": 1.6065753424657534, + "grad_norm": 2.841352939605713, + "learning_rate": 2.613728675873274e-06, + "loss": 1.6228, + "step": 1466 + }, + { + "epoch": 1.6076712328767124, + "grad_norm": 2.5176472663879395, + "learning_rate": 2.6116978066612513e-06, + "loss": 1.5735, + "step": 1467 + }, + { + "epoch": 1.6087671232876712, + "grad_norm": 2.6179540157318115, + "learning_rate": 2.6096669374492286e-06, + "loss": 1.5506, + "step": 1468 + }, + { + "epoch": 1.60986301369863, + "grad_norm": 2.59338116645813, + "learning_rate": 2.607636068237206e-06, + "loss": 1.5964, + "step": 1469 + }, + { + "epoch": 1.610958904109589, + "grad_norm": 2.569547414779663, + "learning_rate": 2.605605199025183e-06, + "loss": 1.5716, + "step": 1470 + }, + { + "epoch": 1.610958904109589, + "eval_loss": 1.7863329648971558, + "eval_runtime": 43.4508, + "eval_samples_per_second": 23.015, + "eval_steps_per_second": 0.967, + "step": 1470 + }, + { + "epoch": 1.612054794520548, + "grad_norm": 2.603998899459839, + "learning_rate": 2.6035743298131604e-06, + "loss": 1.6011, + "step": 1471 + }, + { + "epoch": 1.6131506849315067, + "grad_norm": 2.7348268032073975, + "learning_rate": 2.6015434606011377e-06, + "loss": 1.6082, + "step": 1472 + }, + { + "epoch": 1.6142465753424657, + "grad_norm": 2.5288398265838623, + "learning_rate": 2.5995125913891145e-06, + "loss": 1.5756, + "step": 1473 + }, + { + "epoch": 1.6153424657534248, + "grad_norm": 2.914170742034912, + "learning_rate": 2.5974817221770922e-06, + "loss": 1.5961, + "step": 1474 + }, + { + "epoch": 1.6164383561643836, + "grad_norm": 2.7135908603668213, + "learning_rate": 2.595450852965069e-06, + "loss": 1.6216, + "step": 1475 + }, + { + "epoch": 1.6175342465753424, + "grad_norm": 2.5617594718933105, + "learning_rate": 2.593419983753047e-06, + "loss": 1.5639, + "step": 1476 + }, + { + "epoch": 1.6186301369863014, + "grad_norm": 2.931432008743286, + "learning_rate": 2.5913891145410236e-06, + "loss": 1.6058, + "step": 1477 + }, + { + "epoch": 1.6197260273972602, + "grad_norm": 2.577707052230835, + "learning_rate": 2.5893582453290013e-06, + "loss": 1.5524, + "step": 1478 + }, + { + "epoch": 1.620821917808219, + "grad_norm": 2.5421078205108643, + "learning_rate": 2.587327376116978e-06, + "loss": 1.5715, + "step": 1479 + }, + { + "epoch": 1.621917808219178, + "grad_norm": 2.6675302982330322, + "learning_rate": 2.585296506904956e-06, + "loss": 1.5514, + "step": 1480 + }, + { + "epoch": 1.621917808219178, + "eval_loss": 1.7871757745742798, + "eval_runtime": 43.254, + "eval_samples_per_second": 23.119, + "eval_steps_per_second": 0.971, + "step": 1480 + }, + { + "epoch": 1.6230136986301371, + "grad_norm": 2.6179966926574707, + "learning_rate": 2.5832656376929327e-06, + "loss": 1.5026, + "step": 1481 + }, + { + "epoch": 1.624109589041096, + "grad_norm": 2.8699069023132324, + "learning_rate": 2.5812347684809104e-06, + "loss": 1.6569, + "step": 1482 + }, + { + "epoch": 1.6252054794520547, + "grad_norm": 2.5897462368011475, + "learning_rate": 2.5792038992688873e-06, + "loss": 1.4955, + "step": 1483 + }, + { + "epoch": 1.6263013698630138, + "grad_norm": 2.618124485015869, + "learning_rate": 2.577173030056865e-06, + "loss": 1.5972, + "step": 1484 + }, + { + "epoch": 1.6273972602739726, + "grad_norm": 2.6729276180267334, + "learning_rate": 2.575142160844842e-06, + "loss": 1.5994, + "step": 1485 + }, + { + "epoch": 1.6284931506849314, + "grad_norm": 2.52236008644104, + "learning_rate": 2.5731112916328187e-06, + "loss": 1.536, + "step": 1486 + }, + { + "epoch": 1.6295890410958904, + "grad_norm": 2.5696699619293213, + "learning_rate": 2.5710804224207964e-06, + "loss": 1.5055, + "step": 1487 + }, + { + "epoch": 1.6306849315068495, + "grad_norm": 2.6239490509033203, + "learning_rate": 2.5690495532087733e-06, + "loss": 1.579, + "step": 1488 + }, + { + "epoch": 1.6317808219178083, + "grad_norm": 2.6407887935638428, + "learning_rate": 2.567018683996751e-06, + "loss": 1.541, + "step": 1489 + }, + { + "epoch": 1.632876712328767, + "grad_norm": 2.747039556503296, + "learning_rate": 2.564987814784728e-06, + "loss": 1.5592, + "step": 1490 + }, + { + "epoch": 1.632876712328767, + "eval_loss": 1.7917380332946777, + "eval_runtime": 43.2553, + "eval_samples_per_second": 23.119, + "eval_steps_per_second": 0.971, + "step": 1490 + }, + { + "epoch": 1.633972602739726, + "grad_norm": 2.574068069458008, + "learning_rate": 2.5629569455727055e-06, + "loss": 1.5694, + "step": 1491 + }, + { + "epoch": 1.635068493150685, + "grad_norm": 2.7107486724853516, + "learning_rate": 2.5609260763606824e-06, + "loss": 1.6631, + "step": 1492 + }, + { + "epoch": 1.6361643835616437, + "grad_norm": 2.603778839111328, + "learning_rate": 2.55889520714866e-06, + "loss": 1.5579, + "step": 1493 + }, + { + "epoch": 1.6372602739726028, + "grad_norm": 2.770524740219116, + "learning_rate": 2.556864337936637e-06, + "loss": 1.5619, + "step": 1494 + }, + { + "epoch": 1.6383561643835618, + "grad_norm": 2.6948940753936768, + "learning_rate": 2.5548334687246146e-06, + "loss": 1.6224, + "step": 1495 + }, + { + "epoch": 1.6394520547945204, + "grad_norm": 2.6838672161102295, + "learning_rate": 2.5528025995125915e-06, + "loss": 1.5727, + "step": 1496 + }, + { + "epoch": 1.6405479452054794, + "grad_norm": 2.6357474327087402, + "learning_rate": 2.550771730300569e-06, + "loss": 1.5985, + "step": 1497 + }, + { + "epoch": 1.6416438356164385, + "grad_norm": 2.8199994564056396, + "learning_rate": 2.548740861088546e-06, + "loss": 1.5065, + "step": 1498 + }, + { + "epoch": 1.6427397260273973, + "grad_norm": 2.595996141433716, + "learning_rate": 2.5467099918765233e-06, + "loss": 1.4976, + "step": 1499 + }, + { + "epoch": 1.643835616438356, + "grad_norm": 2.9895756244659424, + "learning_rate": 2.5446791226645006e-06, + "loss": 1.6085, + "step": 1500 + }, + { + "epoch": 1.643835616438356, + "eval_loss": 1.7889667749404907, + "eval_runtime": 42.5759, + "eval_samples_per_second": 23.487, + "eval_steps_per_second": 0.986, + "step": 1500 + }, + { + "epoch": 1.644931506849315, + "grad_norm": 2.686858654022217, + "learning_rate": 2.542648253452478e-06, + "loss": 1.5811, + "step": 1501 + }, + { + "epoch": 1.6460273972602741, + "grad_norm": 2.674805164337158, + "learning_rate": 2.540617384240455e-06, + "loss": 1.5365, + "step": 1502 + }, + { + "epoch": 1.6471232876712327, + "grad_norm": 2.6717612743377686, + "learning_rate": 2.5385865150284324e-06, + "loss": 1.558, + "step": 1503 + }, + { + "epoch": 1.6482191780821918, + "grad_norm": 2.8866565227508545, + "learning_rate": 2.5365556458164097e-06, + "loss": 1.5649, + "step": 1504 + }, + { + "epoch": 1.6493150684931508, + "grad_norm": 2.622196912765503, + "learning_rate": 2.534524776604387e-06, + "loss": 1.4744, + "step": 1505 + }, + { + "epoch": 1.6504109589041096, + "grad_norm": 2.6443939208984375, + "learning_rate": 2.5324939073923642e-06, + "loss": 1.5541, + "step": 1506 + }, + { + "epoch": 1.6515068493150684, + "grad_norm": 2.7976269721984863, + "learning_rate": 2.5304630381803415e-06, + "loss": 1.5654, + "step": 1507 + }, + { + "epoch": 1.6526027397260274, + "grad_norm": 2.8624062538146973, + "learning_rate": 2.5284321689683188e-06, + "loss": 1.5672, + "step": 1508 + }, + { + "epoch": 1.6536986301369863, + "grad_norm": 2.75897479057312, + "learning_rate": 2.526401299756296e-06, + "loss": 1.5536, + "step": 1509 + }, + { + "epoch": 1.654794520547945, + "grad_norm": 2.671563148498535, + "learning_rate": 2.524370430544273e-06, + "loss": 1.5327, + "step": 1510 + }, + { + "epoch": 1.654794520547945, + "eval_loss": 1.789809226989746, + "eval_runtime": 42.1012, + "eval_samples_per_second": 23.752, + "eval_steps_per_second": 0.998, + "step": 1510 + }, + { + "epoch": 1.655890410958904, + "grad_norm": 3.337616205215454, + "learning_rate": 2.5223395613322506e-06, + "loss": 1.5794, + "step": 1511 + }, + { + "epoch": 1.6569863013698631, + "grad_norm": 2.7513420581817627, + "learning_rate": 2.5203086921202274e-06, + "loss": 1.5561, + "step": 1512 + }, + { + "epoch": 1.658082191780822, + "grad_norm": 2.9580318927764893, + "learning_rate": 2.518277822908205e-06, + "loss": 1.572, + "step": 1513 + }, + { + "epoch": 1.6591780821917808, + "grad_norm": 3.01985764503479, + "learning_rate": 2.516246953696182e-06, + "loss": 1.585, + "step": 1514 + }, + { + "epoch": 1.6602739726027398, + "grad_norm": 2.762683629989624, + "learning_rate": 2.5142160844841597e-06, + "loss": 1.5817, + "step": 1515 + }, + { + "epoch": 1.6613698630136986, + "grad_norm": 2.6364688873291016, + "learning_rate": 2.5121852152721365e-06, + "loss": 1.5848, + "step": 1516 + }, + { + "epoch": 1.6624657534246574, + "grad_norm": 2.559342861175537, + "learning_rate": 2.5101543460601142e-06, + "loss": 1.512, + "step": 1517 + }, + { + "epoch": 1.6635616438356164, + "grad_norm": 2.657026529312134, + "learning_rate": 2.508123476848091e-06, + "loss": 1.6042, + "step": 1518 + }, + { + "epoch": 1.6646575342465755, + "grad_norm": 2.721721649169922, + "learning_rate": 2.506092607636069e-06, + "loss": 1.5585, + "step": 1519 + }, + { + "epoch": 1.6657534246575343, + "grad_norm": 2.5659472942352295, + "learning_rate": 2.5040617384240457e-06, + "loss": 1.5456, + "step": 1520 + }, + { + "epoch": 1.6657534246575343, + "eval_loss": 1.7858935594558716, + "eval_runtime": 43.2977, + "eval_samples_per_second": 23.096, + "eval_steps_per_second": 0.97, + "step": 1520 + }, + { + "epoch": 1.666849315068493, + "grad_norm": 2.689267158508301, + "learning_rate": 2.5020308692120234e-06, + "loss": 1.616, + "step": 1521 + }, + { + "epoch": 1.6679452054794521, + "grad_norm": 2.7258920669555664, + "learning_rate": 2.5e-06, + "loss": 1.5499, + "step": 1522 + }, + { + "epoch": 1.669041095890411, + "grad_norm": 2.9586708545684814, + "learning_rate": 2.4979691307879775e-06, + "loss": 1.5476, + "step": 1523 + }, + { + "epoch": 1.6701369863013698, + "grad_norm": 2.8209455013275146, + "learning_rate": 2.4959382615759548e-06, + "loss": 1.6178, + "step": 1524 + }, + { + "epoch": 1.6712328767123288, + "grad_norm": 2.6040000915527344, + "learning_rate": 2.493907392363932e-06, + "loss": 1.5612, + "step": 1525 + }, + { + "epoch": 1.6723287671232878, + "grad_norm": 2.551750659942627, + "learning_rate": 2.4918765231519093e-06, + "loss": 1.5724, + "step": 1526 + }, + { + "epoch": 1.6734246575342466, + "grad_norm": 2.681349992752075, + "learning_rate": 2.4898456539398866e-06, + "loss": 1.6296, + "step": 1527 + }, + { + "epoch": 1.6745205479452054, + "grad_norm": 2.674236536026001, + "learning_rate": 2.487814784727864e-06, + "loss": 1.5584, + "step": 1528 + }, + { + "epoch": 1.6756164383561645, + "grad_norm": 2.648228406906128, + "learning_rate": 2.4857839155158407e-06, + "loss": 1.5591, + "step": 1529 + }, + { + "epoch": 1.6767123287671233, + "grad_norm": 2.62040114402771, + "learning_rate": 2.483753046303818e-06, + "loss": 1.5627, + "step": 1530 + }, + { + "epoch": 1.6767123287671233, + "eval_loss": 1.7849520444869995, + "eval_runtime": 43.6712, + "eval_samples_per_second": 22.898, + "eval_steps_per_second": 0.962, + "step": 1530 + }, + { + "epoch": 1.677808219178082, + "grad_norm": 2.7421514987945557, + "learning_rate": 2.4817221770917953e-06, + "loss": 1.5925, + "step": 1531 + }, + { + "epoch": 1.6789041095890411, + "grad_norm": 3.0015368461608887, + "learning_rate": 2.4796913078797725e-06, + "loss": 1.5954, + "step": 1532 + }, + { + "epoch": 1.6800000000000002, + "grad_norm": 2.5985584259033203, + "learning_rate": 2.47766043866775e-06, + "loss": 1.5948, + "step": 1533 + }, + { + "epoch": 1.6810958904109587, + "grad_norm": 2.55576229095459, + "learning_rate": 2.475629569455727e-06, + "loss": 1.5397, + "step": 1534 + }, + { + "epoch": 1.6821917808219178, + "grad_norm": 2.4863014221191406, + "learning_rate": 2.4735987002437044e-06, + "loss": 1.5366, + "step": 1535 + }, + { + "epoch": 1.6832876712328768, + "grad_norm": 2.5521163940429688, + "learning_rate": 2.4715678310316816e-06, + "loss": 1.5527, + "step": 1536 + }, + { + "epoch": 1.6843835616438356, + "grad_norm": 2.6156206130981445, + "learning_rate": 2.469536961819659e-06, + "loss": 1.5657, + "step": 1537 + }, + { + "epoch": 1.6854794520547944, + "grad_norm": 3.497575044631958, + "learning_rate": 2.467506092607636e-06, + "loss": 1.5461, + "step": 1538 + }, + { + "epoch": 1.6865753424657535, + "grad_norm": 2.9800186157226562, + "learning_rate": 2.4654752233956135e-06, + "loss": 1.6193, + "step": 1539 + }, + { + "epoch": 1.6876712328767123, + "grad_norm": 2.5514957904815674, + "learning_rate": 2.4634443541835907e-06, + "loss": 1.5715, + "step": 1540 + }, + { + "epoch": 1.6876712328767123, + "eval_loss": 1.7836337089538574, + "eval_runtime": 45.9798, + "eval_samples_per_second": 21.749, + "eval_steps_per_second": 0.913, + "step": 1540 + }, + { + "epoch": 1.688767123287671, + "grad_norm": 2.6997995376586914, + "learning_rate": 2.461413484971568e-06, + "loss": 1.6398, + "step": 1541 + }, + { + "epoch": 1.6898630136986301, + "grad_norm": 2.8419151306152344, + "learning_rate": 2.4593826157595453e-06, + "loss": 1.5212, + "step": 1542 + }, + { + "epoch": 1.6909589041095892, + "grad_norm": 3.348747968673706, + "learning_rate": 2.4573517465475226e-06, + "loss": 1.5462, + "step": 1543 + }, + { + "epoch": 1.692054794520548, + "grad_norm": 2.4968204498291016, + "learning_rate": 2.4553208773355e-06, + "loss": 1.5652, + "step": 1544 + }, + { + "epoch": 1.6931506849315068, + "grad_norm": 2.534217357635498, + "learning_rate": 2.453290008123477e-06, + "loss": 1.5908, + "step": 1545 + }, + { + "epoch": 1.6942465753424658, + "grad_norm": 2.585479974746704, + "learning_rate": 2.4512591389114544e-06, + "loss": 1.5888, + "step": 1546 + }, + { + "epoch": 1.6953424657534246, + "grad_norm": 2.7593765258789062, + "learning_rate": 2.4492282696994317e-06, + "loss": 1.5538, + "step": 1547 + }, + { + "epoch": 1.6964383561643834, + "grad_norm": 2.7336785793304443, + "learning_rate": 2.447197400487409e-06, + "loss": 1.5795, + "step": 1548 + }, + { + "epoch": 1.6975342465753425, + "grad_norm": 3.5008962154388428, + "learning_rate": 2.4451665312753862e-06, + "loss": 1.5996, + "step": 1549 + }, + { + "epoch": 1.6986301369863015, + "grad_norm": 2.5645105838775635, + "learning_rate": 2.4431356620633635e-06, + "loss": 1.5865, + "step": 1550 + }, + { + "epoch": 1.6986301369863015, + "eval_loss": 1.7836376428604126, + "eval_runtime": 45.7525, + "eval_samples_per_second": 21.857, + "eval_steps_per_second": 0.918, + "step": 1550 + }, + { + "epoch": 1.6997260273972603, + "grad_norm": 2.62166690826416, + "learning_rate": 2.4411047928513408e-06, + "loss": 1.5788, + "step": 1551 + }, + { + "epoch": 1.7008219178082191, + "grad_norm": 2.7162153720855713, + "learning_rate": 2.439073923639318e-06, + "loss": 1.5215, + "step": 1552 + }, + { + "epoch": 1.7019178082191782, + "grad_norm": 2.7269327640533447, + "learning_rate": 2.4370430544272953e-06, + "loss": 1.5109, + "step": 1553 + }, + { + "epoch": 1.703013698630137, + "grad_norm": 2.9356398582458496, + "learning_rate": 2.435012185215272e-06, + "loss": 1.5491, + "step": 1554 + }, + { + "epoch": 1.7041095890410958, + "grad_norm": 2.738612174987793, + "learning_rate": 2.4329813160032495e-06, + "loss": 1.5703, + "step": 1555 + }, + { + "epoch": 1.7052054794520548, + "grad_norm": 2.671029567718506, + "learning_rate": 2.4309504467912267e-06, + "loss": 1.5546, + "step": 1556 + }, + { + "epoch": 1.7063013698630138, + "grad_norm": 2.821136713027954, + "learning_rate": 2.428919577579204e-06, + "loss": 1.6421, + "step": 1557 + }, + { + "epoch": 1.7073972602739727, + "grad_norm": 2.562530517578125, + "learning_rate": 2.4268887083671813e-06, + "loss": 1.5436, + "step": 1558 + }, + { + "epoch": 1.7084931506849315, + "grad_norm": 2.560969114303589, + "learning_rate": 2.4248578391551586e-06, + "loss": 1.5375, + "step": 1559 + }, + { + "epoch": 1.7095890410958905, + "grad_norm": 2.6686229705810547, + "learning_rate": 2.422826969943136e-06, + "loss": 1.5184, + "step": 1560 + }, + { + "epoch": 1.7095890410958905, + "eval_loss": 1.7841863632202148, + "eval_runtime": 44.8027, + "eval_samples_per_second": 22.32, + "eval_steps_per_second": 0.937, + "step": 1560 + }, + { + "epoch": 1.7106849315068493, + "grad_norm": 2.5086288452148438, + "learning_rate": 2.420796100731113e-06, + "loss": 1.5809, + "step": 1561 + }, + { + "epoch": 1.7117808219178081, + "grad_norm": 2.536912679672241, + "learning_rate": 2.4187652315190904e-06, + "loss": 1.5822, + "step": 1562 + }, + { + "epoch": 1.7128767123287671, + "grad_norm": 2.660067081451416, + "learning_rate": 2.4167343623070677e-06, + "loss": 1.5535, + "step": 1563 + }, + { + "epoch": 1.7139726027397262, + "grad_norm": 2.8859071731567383, + "learning_rate": 2.414703493095045e-06, + "loss": 1.5636, + "step": 1564 + }, + { + "epoch": 1.7150684931506848, + "grad_norm": 2.773730993270874, + "learning_rate": 2.4126726238830222e-06, + "loss": 1.5466, + "step": 1565 + }, + { + "epoch": 1.7161643835616438, + "grad_norm": 2.7574939727783203, + "learning_rate": 2.410641754670999e-06, + "loss": 1.5978, + "step": 1566 + }, + { + "epoch": 1.7172602739726028, + "grad_norm": 2.7123234272003174, + "learning_rate": 2.4086108854589763e-06, + "loss": 1.551, + "step": 1567 + }, + { + "epoch": 1.7183561643835616, + "grad_norm": 2.726815938949585, + "learning_rate": 2.4065800162469536e-06, + "loss": 1.5627, + "step": 1568 + }, + { + "epoch": 1.7194520547945205, + "grad_norm": 2.7364342212677, + "learning_rate": 2.404549147034931e-06, + "loss": 1.5967, + "step": 1569 + }, + { + "epoch": 1.7205479452054795, + "grad_norm": 2.6859424114227295, + "learning_rate": 2.402518277822908e-06, + "loss": 1.5347, + "step": 1570 + }, + { + "epoch": 1.7205479452054795, + "eval_loss": 1.7833315134048462, + "eval_runtime": 45.1817, + "eval_samples_per_second": 22.133, + "eval_steps_per_second": 0.93, + "step": 1570 + }, + { + "epoch": 1.7216438356164385, + "grad_norm": 2.7983601093292236, + "learning_rate": 2.4004874086108854e-06, + "loss": 1.5828, + "step": 1571 + }, + { + "epoch": 1.7227397260273971, + "grad_norm": 2.62542462348938, + "learning_rate": 2.3984565393988627e-06, + "loss": 1.5417, + "step": 1572 + }, + { + "epoch": 1.7238356164383561, + "grad_norm": 2.8035542964935303, + "learning_rate": 2.39642567018684e-06, + "loss": 1.5539, + "step": 1573 + }, + { + "epoch": 1.7249315068493152, + "grad_norm": 2.8192529678344727, + "learning_rate": 2.3943948009748173e-06, + "loss": 1.5824, + "step": 1574 + }, + { + "epoch": 1.726027397260274, + "grad_norm": 2.63417387008667, + "learning_rate": 2.3923639317627945e-06, + "loss": 1.5296, + "step": 1575 + }, + { + "epoch": 1.7271232876712328, + "grad_norm": 2.596536874771118, + "learning_rate": 2.390333062550772e-06, + "loss": 1.4545, + "step": 1576 + }, + { + "epoch": 1.7282191780821918, + "grad_norm": 2.8503546714782715, + "learning_rate": 2.388302193338749e-06, + "loss": 1.6365, + "step": 1577 + }, + { + "epoch": 1.7293150684931506, + "grad_norm": 2.6589956283569336, + "learning_rate": 2.3862713241267264e-06, + "loss": 1.566, + "step": 1578 + }, + { + "epoch": 1.7304109589041095, + "grad_norm": 2.728832721710205, + "learning_rate": 2.3842404549147037e-06, + "loss": 1.5242, + "step": 1579 + }, + { + "epoch": 1.7315068493150685, + "grad_norm": 2.7927703857421875, + "learning_rate": 2.382209585702681e-06, + "loss": 1.6126, + "step": 1580 + }, + { + "epoch": 1.7315068493150685, + "eval_loss": 1.78191077709198, + "eval_runtime": 46.2407, + "eval_samples_per_second": 21.626, + "eval_steps_per_second": 0.908, + "step": 1580 + }, + { + "epoch": 1.7326027397260275, + "grad_norm": 2.800701379776001, + "learning_rate": 2.380178716490658e-06, + "loss": 1.5566, + "step": 1581 + }, + { + "epoch": 1.7336986301369863, + "grad_norm": 2.746800661087036, + "learning_rate": 2.3781478472786355e-06, + "loss": 1.5979, + "step": 1582 + }, + { + "epoch": 1.7347945205479451, + "grad_norm": 2.7532665729522705, + "learning_rate": 2.3761169780666128e-06, + "loss": 1.6007, + "step": 1583 + }, + { + "epoch": 1.7358904109589042, + "grad_norm": 2.7455027103424072, + "learning_rate": 2.37408610885459e-06, + "loss": 1.5478, + "step": 1584 + }, + { + "epoch": 1.736986301369863, + "grad_norm": 2.693115711212158, + "learning_rate": 2.3720552396425673e-06, + "loss": 1.6249, + "step": 1585 + }, + { + "epoch": 1.7380821917808218, + "grad_norm": 2.6274261474609375, + "learning_rate": 2.3700243704305446e-06, + "loss": 1.574, + "step": 1586 + }, + { + "epoch": 1.7391780821917808, + "grad_norm": 2.6467201709747314, + "learning_rate": 2.367993501218522e-06, + "loss": 1.5556, + "step": 1587 + }, + { + "epoch": 1.7402739726027399, + "grad_norm": 2.7470309734344482, + "learning_rate": 2.365962632006499e-06, + "loss": 1.5932, + "step": 1588 + }, + { + "epoch": 1.7413698630136987, + "grad_norm": 2.694892406463623, + "learning_rate": 2.3639317627944764e-06, + "loss": 1.5889, + "step": 1589 + }, + { + "epoch": 1.7424657534246575, + "grad_norm": 2.614192008972168, + "learning_rate": 2.3619008935824537e-06, + "loss": 1.5858, + "step": 1590 + }, + { + "epoch": 1.7424657534246575, + "eval_loss": 1.7806529998779297, + "eval_runtime": 47.8714, + "eval_samples_per_second": 20.889, + "eval_steps_per_second": 0.877, + "step": 1590 + }, + { + "epoch": 1.7435616438356165, + "grad_norm": 2.5324199199676514, + "learning_rate": 2.3598700243704305e-06, + "loss": 1.5872, + "step": 1591 + }, + { + "epoch": 1.7446575342465753, + "grad_norm": 2.6094815731048584, + "learning_rate": 2.357839155158408e-06, + "loss": 1.5729, + "step": 1592 + }, + { + "epoch": 1.7457534246575341, + "grad_norm": 2.702822208404541, + "learning_rate": 2.355808285946385e-06, + "loss": 1.5945, + "step": 1593 + }, + { + "epoch": 1.7468493150684932, + "grad_norm": 2.7433812618255615, + "learning_rate": 2.3537774167343624e-06, + "loss": 1.5772, + "step": 1594 + }, + { + "epoch": 1.7479452054794522, + "grad_norm": 2.6817004680633545, + "learning_rate": 2.3517465475223396e-06, + "loss": 1.5803, + "step": 1595 + }, + { + "epoch": 1.749041095890411, + "grad_norm": 2.8029403686523438, + "learning_rate": 2.349715678310317e-06, + "loss": 1.5449, + "step": 1596 + }, + { + "epoch": 1.7501369863013698, + "grad_norm": 2.82515287399292, + "learning_rate": 2.347684809098294e-06, + "loss": 1.5967, + "step": 1597 + }, + { + "epoch": 1.7512328767123289, + "grad_norm": 2.7368521690368652, + "learning_rate": 2.3456539398862715e-06, + "loss": 1.5325, + "step": 1598 + }, + { + "epoch": 1.7523287671232877, + "grad_norm": 2.6093783378601074, + "learning_rate": 2.3436230706742487e-06, + "loss": 1.5384, + "step": 1599 + }, + { + "epoch": 1.7534246575342465, + "grad_norm": 2.6561474800109863, + "learning_rate": 2.341592201462226e-06, + "loss": 1.5497, + "step": 1600 + }, + { + "epoch": 1.7534246575342465, + "eval_loss": 1.7796835899353027, + "eval_runtime": 44.7534, + "eval_samples_per_second": 22.345, + "eval_steps_per_second": 0.938, + "step": 1600 + } + ], + "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": 4.987695054979072e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-1600/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-2100/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/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_local3_10M_seed0/runs/checkpoint-2100/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/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_local3_10M_seed0/runs/checkpoint-2100/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/latest new file mode 100644 index 0000000000000000000000000000000000000000..b05c33b7aff64e27ba9c8b2e7dc2454bea58ee7f --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/latest @@ -0,0 +1 @@ +global_step2100 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/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_local3_10M_seed0/runs/checkpoint-2100/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/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_local3_10M_seed0/runs/checkpoint-2100/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/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_local3_10M_seed0/runs/checkpoint-2100/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..c1628b2316689eb60b1ef96c6a138c46a1353ef0 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/trainer_state.json @@ -0,0 +1,16413 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.3013698630136985, + "eval_steps": 10, + "global_step": 2100, + "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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.7659518718719482, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.279, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.433785915374756, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.3368, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 3.1486687660217285, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3896, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.391956090927124, + "learning_rate": 1.624087591240876e-06, + "loss": 2.297, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.6649584770202637, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.2415, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.5604891777038574, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2688, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.47900652885437, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2933, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.632723569869995, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.2143, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.8085947036743164, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2748, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.3927979469299316, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2685, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.355207920074463, + "eval_runtime": 42.3505, + "eval_samples_per_second": 23.612, + "eval_steps_per_second": 0.992, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.6170592308044434, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.3121, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.699814796447754, + "learning_rate": 1.77007299270073e-06, + "loss": 2.2853, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.3450610637664795, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3297, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.458219528198242, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2639, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.7101449966430664, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2134, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.5734996795654297, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2347, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 2.5200352668762207, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2299, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5022480487823486, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2253, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.190117359161377, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.2272, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.414524555206299, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2424, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.317634344100952, + "eval_runtime": 43.5742, + "eval_samples_per_second": 22.949, + "eval_steps_per_second": 0.964, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.3226804733276367, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1449, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.733006477355957, + "learning_rate": 1.952554744525548e-06, + "loss": 2.2027, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 2.5586233139038086, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2007, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.669914722442627, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2024, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.6148102283477783, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2496, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 2.331369638442993, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1814, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.3051886558532715, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2015, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.6890218257904053, + "learning_rate": 2.062043795620438e-06, + "loss": 2.184, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.4975883960723877, + "learning_rate": 2.08029197080292e-06, + "loss": 2.2439, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.5848891735076904, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1413, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.280966281890869, + "eval_runtime": 43.1001, + "eval_samples_per_second": 23.202, + "eval_steps_per_second": 0.974, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.2104, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.3356540203094482, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2312, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 2.3616602420806885, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.1399, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.581623077392578, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1456, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.547288656234741, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.184, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6386160850524902, + "learning_rate": 2.208029197080292e-06, + "loss": 2.1658, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.593065023422241, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1443, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.530632972717285, + "learning_rate": 2.244525547445256e-06, + "loss": 2.194, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.3553175926208496, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.1918, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 2.7536532878875732, + "learning_rate": 2.281021897810219e-06, + "loss": 2.1477, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2608063220977783, + "eval_runtime": 42.7938, + "eval_samples_per_second": 23.368, + "eval_steps_per_second": 0.981, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.160094738006592, + "learning_rate": 2.299270072992701e-06, + "loss": 2.133, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.4474735260009766, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1468, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.6654434204101562, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.1194, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.726611375808716, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1518, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 2.3560168743133545, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1537, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.664065361022949, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0888, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.3946099281311035, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.0998, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.7812652587890625, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1231, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.64595890045166, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1746, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.788524866104126, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0731, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.2340502738952637, + "eval_runtime": 42.3104, + "eval_samples_per_second": 23.635, + "eval_steps_per_second": 0.993, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 2.9249846935272217, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0564, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.5507798194885254, + "learning_rate": 2.5e-06, + "loss": 2.1714, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.1604716777801514, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0701, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.5429840087890625, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.0662, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8296737670898438, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.1378, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.9396755695343018, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1069, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.736814022064209, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1076, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 2.774909496307373, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1286, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.0861318111419678, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1499, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.624016284942627, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0989, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.215052604675293, + "eval_runtime": 43.9956, + "eval_samples_per_second": 22.73, + "eval_steps_per_second": 0.955, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.801500082015991, + "learning_rate": 2.664233576642336e-06, + "loss": 2.0795, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 2.2107114791870117, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.1035, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.030381441116333, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.0794, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 2.9719064235687256, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1496, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.688839912414551, + "learning_rate": 2.737226277372263e-06, + "loss": 2.1152, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 2.9350979328155518, + "learning_rate": 2.755474452554745e-06, + "loss": 2.0773, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.8958046436309814, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.1805, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.0781922340393066, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.1249, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.032059907913208, + "learning_rate": 2.81021897810219e-06, + "loss": 2.0632, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.798758029937744, + "learning_rate": 2.828467153284672e-06, + "loss": 2.1001, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2005105018615723, + "eval_runtime": 43.5111, + "eval_samples_per_second": 22.983, + "eval_steps_per_second": 0.965, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.7674927711486816, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.0211, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 2.6168575286865234, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0599, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 2.933353900909424, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.0169, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.81402587890625, + "learning_rate": 2.901459854014599e-06, + "loss": 2.122, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.5592572689056396, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.0957, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.738271951675415, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.0868, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.5362703800201416, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1237, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.6304945945739746, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.0825, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.040717601776123, + "learning_rate": 2.992700729927008e-06, + "loss": 2.0388, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.6226463317871094, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0163, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.186411142349243, + "eval_runtime": 44.625, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.5307068824768066, + "learning_rate": 3.029197080291971e-06, + "loss": 2.1097, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 2.9160444736480713, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.0946, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.5250024795532227, + "learning_rate": 3.065693430656935e-06, + "loss": 2.0709, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.6963064670562744, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.0937, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 2.845707416534424, + "learning_rate": 3.102189781021898e-06, + "loss": 2.0851, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.8985090255737305, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.0176, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 2.734377861022949, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.9931, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.5125555992126465, + "learning_rate": 3.156934306569343e-06, + "loss": 2.1133, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.0393502712249756, + "learning_rate": 3.175182481751825e-06, + "loss": 2.0636, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.6410164833068848, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0304, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.171868085861206, + "eval_runtime": 44.2249, + "eval_samples_per_second": 22.612, + "eval_steps_per_second": 0.95, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.7937498092651367, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0518, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7801201343536377, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.026, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7963638305664062, + "learning_rate": 3.248175182481752e-06, + "loss": 1.9642, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6330482959747314, + "learning_rate": 3.266423357664234e-06, + "loss": 2.0545, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.9771625995635986, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0445, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.6112730503082275, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0616, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.732865333557129, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0403, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.9113688468933105, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0382, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.4516773223876953, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.0416, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.781735897064209, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0348, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.160499334335327, + "eval_runtime": 44.5652, + "eval_samples_per_second": 22.439, + "eval_steps_per_second": 0.942, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.9990546703338623, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0324, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.783071279525757, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0518, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.659040689468384, + "learning_rate": 3.43065693430657e-06, + "loss": 1.9722, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 2.799798011779785, + "learning_rate": 3.4489051094890513e-06, + "loss": 2.0362, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 3.025458812713623, + "learning_rate": 3.467153284671533e-06, + "loss": 1.9863, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 2.686415195465088, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0201, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.716064929962158, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0249, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.637653350830078, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0403, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 2.7686848640441895, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0626, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.721956729888916, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.1077, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.152338981628418, + "eval_runtime": 43.5455, + "eval_samples_per_second": 22.964, + "eval_steps_per_second": 0.965, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.92571759223938, + "learning_rate": 3.576642335766424e-06, + "loss": 2.0928, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.9189202785491943, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9315, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 2.990233898162842, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9427, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.702348232269287, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9768, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6949572563171387, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.9954, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.108405113220215, + "learning_rate": 3.667883211678833e-06, + "loss": 1.9631, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.738922119140625, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9644, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.7419424057006836, + "learning_rate": 3.704379562043796e-06, + "loss": 2.0675, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9853813648223877, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0245, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.4614126682281494, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.9665, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.1342060565948486, + "eval_runtime": 46.3954, + "eval_samples_per_second": 21.554, + "eval_steps_per_second": 0.905, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.6787819862365723, + "learning_rate": 3.7591240875912412e-06, + "loss": 2.0063, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 2.563525438308716, + "learning_rate": 3.777372262773723e-06, + "loss": 1.998, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.4981701374053955, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.9738, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 2.5357978343963623, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0221, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.5365488529205322, + "learning_rate": 3.832116788321168e-06, + "loss": 1.9664, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 2.669468641281128, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9577, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 2.6950953006744385, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9786, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8249378204345703, + "learning_rate": 3.886861313868613e-06, + "loss": 2.0041, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.5924317836761475, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0338, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.1282334327697754, + "learning_rate": 3.923357664233577e-06, + "loss": 1.9698, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.1221420764923096, + "eval_runtime": 47.1127, + "eval_samples_per_second": 21.226, + "eval_steps_per_second": 0.891, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.7878193855285645, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0446, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.7137255668640137, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0575, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.286421060562134, + "learning_rate": 3.978102189781022e-06, + "loss": 1.9131, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.5466175079345703, + "learning_rate": 3.996350364963504e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.792483329772949, + "learning_rate": 4.014598540145986e-06, + "loss": 1.912, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.7533867359161377, + "learning_rate": 4.032846715328467e-06, + "loss": 1.9742, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 2.848405361175537, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9815, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.1034414768218994, + "learning_rate": 4.069343065693431e-06, + "loss": 1.9409, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.5109779834747314, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9676, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.486506938934326, + "learning_rate": 4.105839416058395e-06, + "loss": 2.0353, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1102187633514404, + "eval_runtime": 45.384, + "eval_samples_per_second": 22.034, + "eval_steps_per_second": 0.925, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.419269323348999, + "learning_rate": 4.124087591240876e-06, + "loss": 1.9622, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.4967262744903564, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.9523, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 2.7311460971832275, + "learning_rate": 4.16058394160584e-06, + "loss": 2.0234, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.241004467010498, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.907, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.821537733078003, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9761, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 2.5711426734924316, + "learning_rate": 4.215328467153285e-06, + "loss": 1.9411, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.936594009399414, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9432, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.6091108322143555, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9545, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.522674798965454, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9246, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.984649181365967, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9417, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1031947135925293, + "eval_runtime": 47.5409, + "eval_samples_per_second": 21.035, + "eval_steps_per_second": 0.883, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.8250324726104736, + "learning_rate": 4.306569343065693e-06, + "loss": 1.9061, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.7654290199279785, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9653, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.7091264724731445, + "learning_rate": 4.343065693430658e-06, + "loss": 1.9482, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 2.8019583225250244, + "learning_rate": 4.361313868613139e-06, + "loss": 1.9332, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.147368907928467, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9836, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.7050259113311768, + "learning_rate": 4.397810218978102e-06, + "loss": 1.9607, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.738743543624878, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9374, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.122777223587036, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9249, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.850745439529419, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9707, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.391998291015625, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.9703, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.0887911319732666, + "eval_runtime": 48.5149, + "eval_samples_per_second": 20.612, + "eval_steps_per_second": 0.866, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 2.9131698608398438, + "learning_rate": 4.489051094890512e-06, + "loss": 1.9989, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.1045937538146973, + "learning_rate": 4.507299270072993e-06, + "loss": 1.9718, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.7984158992767334, + "learning_rate": 4.525547445255475e-06, + "loss": 1.947, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.6223504543304443, + "learning_rate": 4.543795620437956e-06, + "loss": 1.8942, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.8221523761749268, + "learning_rate": 4.562043795620438e-06, + "loss": 2.0045, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.6657445430755615, + "learning_rate": 4.580291970802921e-06, + "loss": 2.0053, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 2.793455123901367, + "learning_rate": 4.598540145985402e-06, + "loss": 1.9577, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.551201820373535, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9256, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 2.739194631576538, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9265, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.2917332649230957, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9664, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.086106777191162, + "eval_runtime": 48.995, + "eval_samples_per_second": 20.41, + "eval_steps_per_second": 0.857, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 2.6489369869232178, + "learning_rate": 4.671532846715329e-06, + "loss": 1.9629, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 2.60017466545105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.932, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.569115400314331, + "learning_rate": 4.708029197080292e-06, + "loss": 1.9902, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 2.7225773334503174, + "learning_rate": 4.726277372262774e-06, + "loss": 1.9619, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.5064749717712402, + "learning_rate": 4.744525547445255e-06, + "loss": 1.8921, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.612931966781616, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9584, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.2524304389953613, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9675, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.8852503299713135, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9107, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 2.654776096343994, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9465, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.679086685180664, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9468, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.0842835903167725, + "eval_runtime": 51.7803, + "eval_samples_per_second": 19.312, + "eval_steps_per_second": 0.811, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.5110855102539062, + "learning_rate": 4.854014598540146e-06, + "loss": 1.88, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.1786117553710938, + "learning_rate": 4.872262773722628e-06, + "loss": 1.8994, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.388328790664673, + "learning_rate": 4.89051094890511e-06, + "loss": 1.8914, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 2.784834384918213, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.923, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.470426559448242, + "learning_rate": 4.927007299270074e-06, + "loss": 1.944, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.690580129623413, + "learning_rate": 4.945255474452555e-06, + "loss": 1.8978, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.869737386703491, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9964, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.9965591430664062, + "learning_rate": 4.981751824817518e-06, + "loss": 1.9706, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.6665658950805664, + "learning_rate": 5e-06, + "loss": 1.9313, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.9231722354888916, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9504, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.068364143371582, + "eval_runtime": 47.8334, + "eval_samples_per_second": 20.906, + "eval_steps_per_second": 0.878, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.6094462871551514, + "learning_rate": 4.995938261575955e-06, + "loss": 1.862, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.780848741531372, + "learning_rate": 4.993907392363932e-06, + "loss": 1.95, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.948662757873535, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.878, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 2.9622998237609863, + "learning_rate": 4.989845653939887e-06, + "loss": 1.8954, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 2.6401216983795166, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9021, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.029592275619507, + "learning_rate": 4.985783915515841e-06, + "loss": 1.8967, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.828392505645752, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9134, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.613926649093628, + "learning_rate": 4.981722177091796e-06, + "loss": 1.8829, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.6173207759857178, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9225, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.4657015800476074, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9142, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.059906482696533, + "eval_runtime": 49.6267, + "eval_samples_per_second": 20.15, + "eval_steps_per_second": 0.846, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7448673248291016, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9545, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.639096736907959, + "learning_rate": 4.973598700243705e-06, + "loss": 1.8664, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 2.6445884704589844, + "learning_rate": 4.971567831031681e-06, + "loss": 1.888, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.9130005836486816, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7717, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.48093318939209, + "learning_rate": 4.967506092607636e-06, + "loss": 1.8538, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.8918230533599854, + "learning_rate": 4.965475223395614e-06, + "loss": 1.886, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.511747121810913, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.8633, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.860029697418213, + "learning_rate": 4.961413484971569e-06, + "loss": 1.9063, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7304794788360596, + "learning_rate": 4.959382615759545e-06, + "loss": 1.8872, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.6569266319274902, + "learning_rate": 4.957351746547523e-06, + "loss": 1.8721, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0610575675964355, + "eval_runtime": 43.9292, + "eval_samples_per_second": 22.764, + "eval_steps_per_second": 0.956, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.1482994556427, + "learning_rate": 4.9553208773355e-06, + "loss": 2.0123, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.59568190574646, + "learning_rate": 4.953290008123478e-06, + "loss": 1.902, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.5897581577301025, + "learning_rate": 4.951259138911454e-06, + "loss": 1.8406, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.734032154083252, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.9605, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.8088371753692627, + "learning_rate": 4.947197400487409e-06, + "loss": 1.896, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.9678707122802734, + "learning_rate": 4.945166531275386e-06, + "loss": 1.8625, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.261200189590454, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8884, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 2.5811619758605957, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9311, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.6783649921417236, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9324, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.6533944606781006, + "learning_rate": 4.937043054427295e-06, + "loss": 1.8575, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.049668312072754, + "eval_runtime": 47.7347, + "eval_samples_per_second": 20.949, + "eval_steps_per_second": 0.88, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 3.0882792472839355, + "learning_rate": 4.935012185215272e-06, + "loss": 1.84, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.8259081840515137, + "learning_rate": 4.93298131600325e-06, + "loss": 1.9109, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.540908098220825, + "learning_rate": 4.930950446791227e-06, + "loss": 1.8906, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.9277396202087402, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9047, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.392449140548706, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.876, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 3.1890463829040527, + "learning_rate": 4.924857839155159e-06, + "loss": 1.8709, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.725710391998291, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8953, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.6287014484405518, + "learning_rate": 4.920796100731113e-06, + "loss": 1.8204, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.7316651344299316, + "learning_rate": 4.918765231519091e-06, + "loss": 1.882, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.5561037063598633, + "learning_rate": 4.916734362307068e-06, + "loss": 1.8331, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0412285327911377, + "eval_runtime": 50.237, + "eval_samples_per_second": 19.906, + "eval_steps_per_second": 0.836, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.974501132965088, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8414, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.4530839920043945, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8323, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.789569616317749, + "learning_rate": 4.910641754671e-06, + "loss": 1.8697, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.582704544067383, + "learning_rate": 4.908610885458977e-06, + "loss": 1.8592, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 2.776442289352417, + "learning_rate": 4.906580016246954e-06, + "loss": 1.912, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.670191764831543, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.8899, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.4534454345703125, + "learning_rate": 4.902518277822909e-06, + "loss": 1.8878, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.007469654083252, + "learning_rate": 4.900487408610886e-06, + "loss": 1.8977, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 2.856466293334961, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9062, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.817441463470459, + "learning_rate": 4.89642567018684e-06, + "loss": 1.9686, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.0344440937042236, + "eval_runtime": 47.2472, + "eval_samples_per_second": 21.165, + "eval_steps_per_second": 0.889, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.9732158184051514, + "learning_rate": 4.894394800974818e-06, + "loss": 1.8332, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 3.0413055419921875, + "learning_rate": 4.892363931762794e-06, + "loss": 1.9171, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.646749973297119, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.8403, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.1274876594543457, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8218, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.8122124671936035, + "learning_rate": 4.886271324126727e-06, + "loss": 1.9272, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.4608492851257324, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.8974, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.928083658218384, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.8646, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.6308164596557617, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8219, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7311816215515137, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8326, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.640162467956543, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8013, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0261101722717285, + "eval_runtime": 47.0605, + "eval_samples_per_second": 21.249, + "eval_steps_per_second": 0.892, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.8542518615722656, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8233, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.464303970336914, + "learning_rate": 4.872055239642567e-06, + "loss": 1.787, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8535103797912598, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8535, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.5060505867004395, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8733, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.6549222469329834, + "learning_rate": 4.865962632006499e-06, + "loss": 1.8047, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.8636224269866943, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8543, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.644256830215454, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8853, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.949490547180176, + "learning_rate": 4.859870024370431e-06, + "loss": 1.8951, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.4576282501220703, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9069, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.6140542030334473, + "learning_rate": 4.855808285946385e-06, + "loss": 1.8573, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0159800052642822, + "eval_runtime": 44.2447, + "eval_samples_per_second": 22.602, + "eval_steps_per_second": 0.949, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.525766134262085, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.844, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.5479896068573, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8572, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.628570795059204, + "learning_rate": 4.849715678310317e-06, + "loss": 1.865, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.2768666744232178, + "learning_rate": 4.847684809098294e-06, + "loss": 1.8355, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.635516881942749, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8603, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.5307435989379883, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8482, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.3688952922821045, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8894, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.6000115871429443, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.8045, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7782504558563232, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8283, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.6914777755737305, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8765, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0086047649383545, + "eval_runtime": 48.245, + "eval_samples_per_second": 20.728, + "eval_steps_per_second": 0.871, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.598365306854248, + "learning_rate": 4.833468724614135e-06, + "loss": 1.8461, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.6718530654907227, + "learning_rate": 4.831437855402113e-06, + "loss": 1.7453, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.393528699874878, + "learning_rate": 4.82940698619009e-06, + "loss": 1.8275, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 3.0622947216033936, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8658, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.431041955947876, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8601, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.939424514770508, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8649, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.8116796016693115, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8303, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.511960029602051, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8383, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 3.8372581005096436, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8111, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.511948347091675, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8448, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.9998281002044678, + "eval_runtime": 44.7624, + "eval_samples_per_second": 22.34, + "eval_steps_per_second": 0.938, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 3.5780487060546875, + "learning_rate": 4.813160032493907e-06, + "loss": 1.8161, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.81740403175354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.8633, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.8117666244506836, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8385, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.5791451930999756, + "learning_rate": 4.80706742485784e-06, + "loss": 1.896, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6816465854644775, + "learning_rate": 4.805036555645816e-06, + "loss": 1.8126, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.6419460773468018, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8753, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.666086196899414, + "learning_rate": 4.800974817221771e-06, + "loss": 1.8305, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8626585006713867, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8463, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.351825475692749, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.838, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.300717830657959, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8711, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.9904799461364746, + "eval_runtime": 43.9511, + "eval_samples_per_second": 22.753, + "eval_steps_per_second": 0.956, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.43884015083313, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8145, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.50593900680542, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8103, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.4643657207489014, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8162, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.4615302085876465, + "learning_rate": 4.786758732737612e-06, + "loss": 1.8637, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.6566433906555176, + "learning_rate": 4.784727863525589e-06, + "loss": 1.7805, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.5973432064056396, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8327, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.4996957778930664, + "learning_rate": 4.780666125101544e-06, + "loss": 1.7761, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 3.0266876220703125, + "learning_rate": 4.778635255889521e-06, + "loss": 1.8036, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.522369146347046, + "learning_rate": 4.776604386677498e-06, + "loss": 1.814, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.482727527618408, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8665, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.9888195991516113, + "eval_runtime": 44.0755, + "eval_samples_per_second": 22.688, + "eval_steps_per_second": 0.953, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 3.237544059753418, + "learning_rate": 4.772542648253453e-06, + "loss": 1.8642, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.393286943435669, + "learning_rate": 4.77051177904143e-06, + "loss": 1.7857, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.66072154045105, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8423, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.6725080013275146, + "learning_rate": 4.766450040617385e-06, + "loss": 1.815, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.7657666206359863, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8729, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.3153769969940186, + "learning_rate": 4.762388302193339e-06, + "loss": 1.8349, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5804553031921387, + "learning_rate": 4.760357432981316e-06, + "loss": 1.8155, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.7280807495117188, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8262, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.551940441131592, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8332, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.6137285232543945, + "learning_rate": 4.754264825345248e-06, + "loss": 1.7805, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.983920693397522, + "eval_runtime": 42.525, + "eval_samples_per_second": 23.516, + "eval_steps_per_second": 0.988, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.4460153579711914, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.7953, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.556032657623291, + "learning_rate": 4.750203086921203e-06, + "loss": 1.7393, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.335983991622925, + "learning_rate": 4.74817221770918e-06, + "loss": 1.8494, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.5057711601257324, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8663, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4447145462036133, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7532, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.769984483718872, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7755, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.472193956375122, + "learning_rate": 4.740048740861089e-06, + "loss": 1.8056, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.4286932945251465, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8094, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.0248074531555176, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8497, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.607700824737549, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8741, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.9779797792434692, + "eval_runtime": 43.7734, + "eval_samples_per_second": 22.845, + "eval_steps_per_second": 0.959, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 3.035045623779297, + "learning_rate": 4.731925264012998e-06, + "loss": 1.8438, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.786921977996826, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7666, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.4158425331115723, + "learning_rate": 4.727863525588953e-06, + "loss": 1.7464, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7772, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.8447, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 3.08266282081604, + "learning_rate": 4.723801787164907e-06, + "loss": 1.7812, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.791749954223633, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8047, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6457693576812744, + "learning_rate": 4.719740048740861e-06, + "loss": 1.836, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.69480037689209, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8145, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6356561183929443, + "learning_rate": 4.715678310316816e-06, + "loss": 1.7756, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.9723366498947144, + "eval_runtime": 45.5806, + "eval_samples_per_second": 21.939, + "eval_steps_per_second": 0.921, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.53474497795105, + "learning_rate": 4.713647441104794e-06, + "loss": 1.774, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.3631107807159424, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8517, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.594604730606079, + "learning_rate": 4.709585702680748e-06, + "loss": 1.8962, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.5280940532684326, + "learning_rate": 4.707554833468725e-06, + "loss": 1.7759, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.4596288204193115, + "learning_rate": 4.705523964256703e-06, + "loss": 1.7948, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.9123644828796387, + "learning_rate": 4.703493095044679e-06, + "loss": 1.7399, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.4354264736175537, + "learning_rate": 4.701462225832657e-06, + "loss": 1.8659, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.28625226020813, + "learning_rate": 4.699431356620634e-06, + "loss": 1.831, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.5445950031280518, + "learning_rate": 4.697400487408612e-06, + "loss": 1.8121, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.767825126647949, + "learning_rate": 4.695369618196588e-06, + "loss": 1.7874, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9653255939483643, + "eval_runtime": 45.2453, + "eval_samples_per_second": 22.102, + "eval_steps_per_second": 0.928, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4052765369415283, + "learning_rate": 4.693338748984566e-06, + "loss": 1.7539, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.498098373413086, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8343, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.821753740310669, + "learning_rate": 4.68927701056052e-06, + "loss": 1.8119, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.4158449172973633, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.7982, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.5229434967041016, + "learning_rate": 4.685215272136475e-06, + "loss": 1.806, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.729199171066284, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8768, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.5102570056915283, + "learning_rate": 4.681153533712429e-06, + "loss": 1.775, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4463515281677246, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7493, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.505733013153076, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8462, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.5784912109375, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7381, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9640673398971558, + "eval_runtime": 44.2714, + "eval_samples_per_second": 22.588, + "eval_steps_per_second": 0.949, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.7864291667938232, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7612, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.507809638977051, + "learning_rate": 4.670999187652316e-06, + "loss": 1.757, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.5499937534332275, + "learning_rate": 4.668968318440293e-06, + "loss": 1.8243, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.435472011566162, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7269, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.541031837463379, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.7959, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5052099227905273, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7191, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6285109519958496, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8423, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.7825605869293213, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7178, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.677319288253784, + "learning_rate": 4.656783103168157e-06, + "loss": 1.7939, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.6830129623413086, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7842, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.960136890411377, + "eval_runtime": 45.9379, + "eval_samples_per_second": 21.769, + "eval_steps_per_second": 0.914, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.5368666648864746, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8352, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.8922412395477295, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.8252, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.8332598209381104, + "learning_rate": 4.648659626320066e-06, + "loss": 1.7805, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.420780658721924, + "learning_rate": 4.646628757108043e-06, + "loss": 1.7499, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.6684062480926514, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.7857, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.8547494411468506, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.8115, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.59279203414917, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8254, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.80961275100708, + "learning_rate": 4.638505280259952e-06, + "loss": 1.7901, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.9017655849456787, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8265, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.595454216003418, + "learning_rate": 4.634443541835907e-06, + "loss": 1.7833, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9576054811477661, + "eval_runtime": 50.7265, + "eval_samples_per_second": 19.714, + "eval_steps_per_second": 0.828, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 4.260303020477295, + "learning_rate": 4.632412672623883e-06, + "loss": 1.6741, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.366147756576538, + "learning_rate": 4.630381803411861e-06, + "loss": 1.7778, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.894336462020874, + "learning_rate": 4.628350934199838e-06, + "loss": 1.8268, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.7675349712371826, + "learning_rate": 4.626320064987816e-06, + "loss": 1.7748, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.461413621902466, + "learning_rate": 4.624289195775792e-06, + "loss": 1.7599, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.670187473297119, + "learning_rate": 4.62225832656377e-06, + "loss": 1.7736, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.8119454383850098, + "learning_rate": 4.620227457351747e-06, + "loss": 1.7837, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.8020036220550537, + "learning_rate": 4.618196588139724e-06, + "loss": 1.7508, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.5977094173431396, + "learning_rate": 4.616165718927701e-06, + "loss": 1.834, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.731919765472412, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.8137, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.9465162754058838, + "eval_runtime": 48.9829, + "eval_samples_per_second": 20.415, + "eval_steps_per_second": 0.857, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.604386568069458, + "learning_rate": 4.612103980503656e-06, + "loss": 1.7854, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.601107120513916, + "learning_rate": 4.610073111291633e-06, + "loss": 1.7749, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6758370399475098, + "learning_rate": 4.60804224207961e-06, + "loss": 1.7826, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.631549119949341, + "learning_rate": 4.606011372867588e-06, + "loss": 1.7723, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.6704764366149902, + "learning_rate": 4.603980503655565e-06, + "loss": 1.7088, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 3.078573703765869, + "learning_rate": 4.601949634443542e-06, + "loss": 1.7254, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7492356300354004, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7756, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 3.0763585567474365, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7806, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.5379223823547363, + "learning_rate": 4.595857026807474e-06, + "loss": 1.7516, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8343496322631836, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8111, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.945749282836914, + "eval_runtime": 46.5453, + "eval_samples_per_second": 21.484, + "eval_steps_per_second": 0.902, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.809753179550171, + "learning_rate": 4.591795288383429e-06, + "loss": 1.7877, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.4196107387542725, + "learning_rate": 4.589764419171406e-06, + "loss": 1.7457, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.409287452697754, + "learning_rate": 4.587733549959383e-06, + "loss": 1.7338, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.414706230163574, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7185, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.591365337371826, + "learning_rate": 4.583671811535338e-06, + "loss": 1.8083, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.4648489952087402, + "learning_rate": 4.581640942323315e-06, + "loss": 1.8192, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.3804972171783447, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8077, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.283082962036133, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8294, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.61795711517334, + "learning_rate": 4.575548334687247e-06, + "loss": 1.7322, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.4035353660583496, + "learning_rate": 4.573517465475224e-06, + "loss": 1.7798, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9422866106033325, + "eval_runtime": 45.3068, + "eval_samples_per_second": 22.072, + "eval_steps_per_second": 0.927, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.7253756523132324, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7876, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.738295078277588, + "learning_rate": 4.569455727051178e-06, + "loss": 1.7933, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 3.1418697834014893, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8145, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.941603660583496, + "learning_rate": 4.565393988627132e-06, + "loss": 1.7671, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.063924551010132, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.7809, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 4.957335472106934, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7306, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.651716709136963, + "learning_rate": 4.559301380991065e-06, + "loss": 1.7428, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 4.777034282684326, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.8118, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.5549471378326416, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7294, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.474388599395752, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8338, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.9402854442596436, + "eval_runtime": 43.166, + "eval_samples_per_second": 23.166, + "eval_steps_per_second": 0.973, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.8968348503112793, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7432, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.6996726989746094, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.7837, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.4599392414093018, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7388, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.745487928390503, + "learning_rate": 4.545085296506905e-06, + "loss": 1.7439, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.6960670948028564, + "learning_rate": 4.543054427294882e-06, + "loss": 1.737, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.596876621246338, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7708, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.598267078399658, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7541, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.6069958209991455, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7756, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.490694761276245, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.7372, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.44179368019104, + "learning_rate": 4.532900081234769e-06, + "loss": 1.7652, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.934715986251831, + "eval_runtime": 44.688, + "eval_samples_per_second": 22.377, + "eval_steps_per_second": 0.94, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 3.018480062484741, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7596, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.5316643714904785, + "learning_rate": 4.528838342810723e-06, + "loss": 1.7715, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4893109798431396, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7549, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.524834156036377, + "learning_rate": 4.524776604386678e-06, + "loss": 1.8046, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7527239322662354, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7583, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.5210752487182617, + "learning_rate": 4.520714865962632e-06, + "loss": 1.7896, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.5787770748138428, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7355, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.537195920944214, + "learning_rate": 4.516653127538587e-06, + "loss": 1.7272, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.6810014247894287, + "learning_rate": 4.514622258326564e-06, + "loss": 1.753, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.5868756771087646, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7964, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.934499740600586, + "eval_runtime": 45.6556, + "eval_samples_per_second": 21.903, + "eval_steps_per_second": 0.92, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.6830856800079346, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7072, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.6574594974517822, + "learning_rate": 4.508529650690496e-06, + "loss": 1.7004, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.4415769577026367, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7667, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.825115442276001, + "learning_rate": 4.504467912266451e-06, + "loss": 1.8356, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.949010133743286, + "learning_rate": 4.502437043054428e-06, + "loss": 1.7579, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.690986394882202, + "learning_rate": 4.500406173842405e-06, + "loss": 1.7657, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 3.1015520095825195, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.7432, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.608424186706543, + "learning_rate": 4.49634443541836e-06, + "loss": 1.8112, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.5238773822784424, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7724, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.7712161540985107, + "learning_rate": 4.492282696994314e-06, + "loss": 1.7111, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9326083660125732, + "eval_runtime": 48.5359, + "eval_samples_per_second": 20.603, + "eval_steps_per_second": 0.865, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.4763033390045166, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7268, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.6310081481933594, + "learning_rate": 4.488220958570269e-06, + "loss": 1.6885, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.5950350761413574, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7775, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.769026517868042, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7737, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.4898502826690674, + "learning_rate": 4.4821283509342e-06, + "loss": 1.751, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.668081521987915, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7178, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.441615104675293, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7511, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.727665901184082, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.8639, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.7964634895324707, + "learning_rate": 4.474004874086109e-06, + "loss": 1.6945, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.6142687797546387, + "learning_rate": 4.471974004874087e-06, + "loss": 1.8028, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9281476736068726, + "eval_runtime": 48.078, + "eval_samples_per_second": 20.8, + "eval_steps_per_second": 0.874, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.6655778884887695, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.7632, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.931162118911743, + "learning_rate": 4.467912266450041e-06, + "loss": 1.7244, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.469895124435425, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7773, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.4948830604553223, + "learning_rate": 4.463850528025995e-06, + "loss": 1.788, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5961945056915283, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.756, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.452946424484253, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6968, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.9352009296417236, + "learning_rate": 4.457757920389927e-06, + "loss": 1.7414, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.7044763565063477, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7848, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.7200324535369873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7699, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.6185197830200195, + "learning_rate": 4.451665312753859e-06, + "loss": 1.7816, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9209110736846924, + "eval_runtime": 47.3321, + "eval_samples_per_second": 21.127, + "eval_steps_per_second": 0.887, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.3639564514160156, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7164, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7073519229888916, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.727, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.637080430984497, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7951, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.8013885021209717, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7333, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.512484312057495, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8048, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.3851466178894043, + "learning_rate": 4.439480097481723e-06, + "loss": 1.6667, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.8105392456054688, + "learning_rate": 4.4374492282697e-06, + "loss": 1.7622, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.555773973464966, + "learning_rate": 4.435418359057677e-06, + "loss": 1.7519, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.688596725463867, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7198, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.828646421432495, + "learning_rate": 4.431356620633632e-06, + "loss": 1.7478, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.919256567955017, + "eval_runtime": 50.5993, + "eval_samples_per_second": 19.763, + "eval_steps_per_second": 0.83, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.4523651599884033, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7344, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.4237422943115234, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6664, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.2746193408966064, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.7446, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.502819061279297, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7691, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.7032017707824707, + "learning_rate": 4.421202274573518e-06, + "loss": 1.6448, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.5879440307617188, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6908, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 3.2481982707977295, + "learning_rate": 4.417140536149473e-06, + "loss": 1.7725, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.4844820499420166, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7141, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.5596253871917725, + "learning_rate": 4.413078797725427e-06, + "loss": 1.7485, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 3.142839193344116, + "learning_rate": 4.411047928513404e-06, + "loss": 1.734, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9129430055618286, + "eval_runtime": 49.2608, + "eval_samples_per_second": 20.3, + "eval_steps_per_second": 0.853, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.5837552547454834, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7668, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.7043633460998535, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7497, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.9886765480041504, + "learning_rate": 4.404955320877336e-06, + "loss": 1.7082, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.3199501037597656, + "learning_rate": 4.402924451665313e-06, + "loss": 1.693, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.606682777404785, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7242, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.6967978477478027, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7516, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.351726531982422, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7338, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.590869188308716, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7462, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.3498220443725586, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7609, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.281002998352051, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7584, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.908841609954834, + "eval_runtime": 50.7771, + "eval_samples_per_second": 19.694, + "eval_steps_per_second": 0.827, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.468755006790161, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7654, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.3467307090759277, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7923, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.3723301887512207, + "learning_rate": 4.384646628757108e-06, + "loss": 1.74, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.4527792930603027, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6988, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.3923704624176025, + "learning_rate": 4.380584890333063e-06, + "loss": 1.6406, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.740556240081787, + "learning_rate": 4.37855402112104e-06, + "loss": 1.8032, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.3013832569122314, + "learning_rate": 4.376523151909017e-06, + "loss": 1.7259, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.4632391929626465, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.7462, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.4878177642822266, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7665, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.610898494720459, + "learning_rate": 4.370430544272949e-06, + "loss": 1.7646, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9056707620620728, + "eval_runtime": 51.8738, + "eval_samples_per_second": 19.278, + "eval_steps_per_second": 0.81, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.387721300125122, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7454, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.9966042041778564, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6789, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.664689064025879, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7486, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.4788103103637695, + "learning_rate": 4.362307067424858e-06, + "loss": 1.6981, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.4413328170776367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7279, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.6818156242370605, + "learning_rate": 4.358245329000813e-06, + "loss": 1.74, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.650348424911499, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7212, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.442805767059326, + "learning_rate": 4.354183590576767e-06, + "loss": 1.6951, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.6806585788726807, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7499, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.586792230606079, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7709, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.901674747467041, + "eval_runtime": 42.3186, + "eval_samples_per_second": 23.63, + "eval_steps_per_second": 0.992, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.547452688217163, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7458, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.5820348262786865, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.7232, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.4208321571350098, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7981, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.8791956901550293, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7372, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6537868976593018, + "learning_rate": 4.339967506092607e-06, + "loss": 1.7216, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.416097402572632, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.701, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.5565483570098877, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7558, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.5875823497772217, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7439, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5459702014923096, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.7058, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.4195566177368164, + "learning_rate": 4.329813160032495e-06, + "loss": 1.6749, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9035066366195679, + "eval_runtime": 43.7962, + "eval_samples_per_second": 22.833, + "eval_steps_per_second": 0.959, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6019818782806396, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7549, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.8627779483795166, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7257, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.582158088684082, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7138, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.4528613090515137, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7162, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 3.1956026554107666, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7556, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.3531901836395264, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.746, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.441988706588745, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7664, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.599506378173828, + "learning_rate": 4.313566206336312e-06, + "loss": 1.6955, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.4121932983398438, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7194, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.4803290367126465, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7386, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8993914127349854, + "eval_runtime": 43.0525, + "eval_samples_per_second": 23.227, + "eval_steps_per_second": 0.976, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.701594114303589, + "learning_rate": 4.307473598700244e-06, + "loss": 1.6796, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7144737243652344, + "learning_rate": 4.305442729488221e-06, + "loss": 1.763, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.395312786102295, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7522, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.3926732540130615, + "learning_rate": 4.301380991064176e-06, + "loss": 1.7705, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.6293702125549316, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7693, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.3746306896209717, + "learning_rate": 4.29731925264013e-06, + "loss": 1.72, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.5421926975250244, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6868, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.6991186141967773, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7378, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.463897705078125, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7628, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5139036178588867, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7125, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8909698724746704, + "eval_runtime": 42.5917, + "eval_samples_per_second": 23.479, + "eval_steps_per_second": 0.986, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.5220465660095215, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6975, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.432988405227661, + "learning_rate": 4.285134037367994e-06, + "loss": 1.6904, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.5894253253936768, + "learning_rate": 4.283103168155971e-06, + "loss": 1.6919, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5334458351135254, + "learning_rate": 4.281072298943948e-06, + "loss": 1.6968, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.510019302368164, + "learning_rate": 4.279041429731926e-06, + "loss": 1.6791, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.531538248062134, + "learning_rate": 4.277010560519903e-06, + "loss": 1.723, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.5621440410614014, + "learning_rate": 4.27497969130788e-06, + "loss": 1.749, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.594329833984375, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7579, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.321031093597412, + "learning_rate": 4.270917952883835e-06, + "loss": 1.6983, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.6325764656066895, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7872, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.886696219444275, + "eval_runtime": 43.4143, + "eval_samples_per_second": 23.034, + "eval_steps_per_second": 0.967, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5342679023742676, + "learning_rate": 4.266856214459789e-06, + "loss": 1.6655, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.5810720920562744, + "learning_rate": 4.264825345247766e-06, + "loss": 1.7182, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.7768406867980957, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6534, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.5680463314056396, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7243, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7162973880767822, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.7146, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.697925090789795, + "learning_rate": 4.256701868399675e-06, + "loss": 1.7296, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.4458425045013428, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7466, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6330654621124268, + "learning_rate": 4.252640129975629e-06, + "loss": 1.6949, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.350430727005005, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6574, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.505995512008667, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7111, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.8887447118759155, + "eval_runtime": 43.9304, + "eval_samples_per_second": 22.763, + "eval_steps_per_second": 0.956, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 3.036954879760742, + "learning_rate": 4.246547522339562e-06, + "loss": 1.7182, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.563606023788452, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.6566, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.516315460205078, + "learning_rate": 4.242485783915517e-06, + "loss": 1.6543, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.6289877891540527, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6996, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.7563092708587646, + "learning_rate": 4.23842404549147e-06, + "loss": 1.666, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5743000507354736, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7338, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 3.1975982189178467, + "learning_rate": 4.234362307067425e-06, + "loss": 1.7444, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.6022894382476807, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7367, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.4466192722320557, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.726, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.579015016555786, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7146, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.8904188871383667, + "eval_runtime": 43.8091, + "eval_samples_per_second": 22.826, + "eval_steps_per_second": 0.959, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.719646453857422, + "learning_rate": 4.226238830219334e-06, + "loss": 1.7209, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.4423880577087402, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7288, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5883753299713135, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7089, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.447049140930176, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7095, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.453739881515503, + "learning_rate": 4.218115353371243e-06, + "loss": 1.6747, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4540836811065674, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7578, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.5468671321868896, + "learning_rate": 4.214053614947198e-06, + "loss": 1.6556, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.443814277648926, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7307, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.4998748302459717, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6768, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.667025566101074, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7493, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.8830010890960693, + "eval_runtime": 44.7027, + "eval_samples_per_second": 22.37, + "eval_steps_per_second": 0.94, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.416248321533203, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7381, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.378025770187378, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7335, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.633173704147339, + "learning_rate": 4.201868399675061e-06, + "loss": 1.6484, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.549384117126465, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7143, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.6492598056793213, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6907, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.444086790084839, + "learning_rate": 4.195775792038993e-06, + "loss": 1.636, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.628814935684204, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7342, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.83831524848938, + "learning_rate": 4.191714053614948e-06, + "loss": 1.7028, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.467339515686035, + "learning_rate": 4.189683184402924e-06, + "loss": 1.749, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.4529521465301514, + "learning_rate": 4.187652315190902e-06, + "loss": 1.682, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.8826342821121216, + "eval_runtime": 44.7757, + "eval_samples_per_second": 22.334, + "eval_steps_per_second": 0.938, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7340762615203857, + "learning_rate": 4.185621445978879e-06, + "loss": 1.723, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.6028647422790527, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7566, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.3967695236206055, + "learning_rate": 4.181559707554833e-06, + "loss": 1.683, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5475900173187256, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6976, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.5411484241485596, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7489, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 3.458749294281006, + "learning_rate": 4.175467099918766e-06, + "loss": 1.6828, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.2977802753448486, + "learning_rate": 4.173436230706742e-06, + "loss": 1.6963, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.875307559967041, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.6857, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 4.097696304321289, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7225, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.6774847507476807, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7031, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.8783056735992432, + "eval_runtime": 44.0219, + "eval_samples_per_second": 22.716, + "eval_steps_per_second": 0.954, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.6787376403808594, + "learning_rate": 4.165312753858652e-06, + "loss": 1.6865, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.3803224563598633, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7237, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.9285683631896973, + "learning_rate": 4.161251015434607e-06, + "loss": 1.6575, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.5534448623657227, + "learning_rate": 4.159220146222583e-06, + "loss": 1.6799, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4105820655822754, + "learning_rate": 4.157189277010561e-06, + "loss": 1.7128, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.6571805477142334, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7123, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.891935348510742, + "learning_rate": 4.153127538586516e-06, + "loss": 1.6949, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.8692519664764404, + "learning_rate": 4.151096669374492e-06, + "loss": 1.699, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.780078649520874, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7114, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.4797518253326416, + "learning_rate": 4.147034930950447e-06, + "loss": 1.6943, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.8726727962493896, + "eval_runtime": 43.5664, + "eval_samples_per_second": 22.953, + "eval_steps_per_second": 0.964, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.7010464668273926, + "learning_rate": 4.145004061738425e-06, + "loss": 1.6207, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.8812713623046875, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.6608, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.6700279712677, + "learning_rate": 4.140942323314379e-06, + "loss": 1.677, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.796513557434082, + "learning_rate": 4.138911454102356e-06, + "loss": 1.7291, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.4274256229400635, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7058, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.5528790950775146, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7064, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.4288926124572754, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7163, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.4939630031585693, + "learning_rate": 4.130787977254265e-06, + "loss": 1.721, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7909305095672607, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7777, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.3382463455200195, + "learning_rate": 4.12672623883022e-06, + "loss": 1.6975, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.8738009929656982, + "eval_runtime": 44.8142, + "eval_samples_per_second": 22.314, + "eval_steps_per_second": 0.937, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6876718997955322, + "learning_rate": 4.124695369618197e-06, + "loss": 1.6851, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.450212240219116, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7209, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.4365246295928955, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.6545, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.3236255645751953, + "learning_rate": 4.118602761982129e-06, + "loss": 1.6434, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.4753000736236572, + "learning_rate": 4.116571892770106e-06, + "loss": 1.7471, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.84885835647583, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6777, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.4361016750335693, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7079, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.5868160724639893, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7288, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.6499862670898438, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7191, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.333289384841919, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.5938, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.8746954202651978, + "eval_runtime": 46.8039, + "eval_samples_per_second": 21.366, + "eval_steps_per_second": 0.897, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 3.1299779415130615, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7585, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 3.6410014629364014, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6803, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.801745653152466, + "learning_rate": 4.100324939073924e-06, + "loss": 1.6957, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 3.0350706577301025, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7031, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 3.5306341648101807, + "learning_rate": 4.096263200649879e-06, + "loss": 1.6598, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.591485023498535, + "learning_rate": 4.094232331437856e-06, + "loss": 1.6659, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 3.0916528701782227, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6951, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.459009885787964, + "learning_rate": 4.090170593013811e-06, + "loss": 1.6597, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.932072162628174, + "learning_rate": 4.088139723801787e-06, + "loss": 1.6698, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.773566484451294, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6824, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.8707821369171143, + "eval_runtime": 46.7886, + "eval_samples_per_second": 21.373, + "eval_steps_per_second": 0.898, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.6008529663085938, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7352, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.502838134765625, + "learning_rate": 4.08204711616572e-06, + "loss": 1.6342, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 3.081169605255127, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7307, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.841200828552246, + "learning_rate": 4.077985377741674e-06, + "loss": 1.7401, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.5010390281677246, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7252, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.468873977661133, + "learning_rate": 4.073923639317629e-06, + "loss": 1.6627, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.829737663269043, + "learning_rate": 4.071892770105605e-06, + "loss": 1.6419, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.8112833499908447, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6629, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6793017387390137, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6667, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.327993869781494, + "learning_rate": 4.065800162469538e-06, + "loss": 1.71, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.874659538269043, + "eval_runtime": 49.0634, + "eval_samples_per_second": 20.382, + "eval_steps_per_second": 0.856, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.689023971557617, + "learning_rate": 4.063769293257514e-06, + "loss": 1.6881, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.612842082977295, + "learning_rate": 4.061738424045492e-06, + "loss": 1.6971, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.319627046585083, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7469, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.588916540145874, + "learning_rate": 4.057676685621446e-06, + "loss": 1.707, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.270096778869629, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.6284, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.4404375553131104, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7007, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5532679557800293, + "learning_rate": 4.051584077985378e-06, + "loss": 1.72, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.5105185508728027, + "learning_rate": 4.049553208773355e-06, + "loss": 1.6772, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.577929973602295, + "learning_rate": 4.047522339561333e-06, + "loss": 1.6747, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.8271372318267822, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7951, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.8684895038604736, + "eval_runtime": 48.0288, + "eval_samples_per_second": 20.821, + "eval_steps_per_second": 0.874, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.3306479454040527, + "learning_rate": 4.043460601137287e-06, + "loss": 1.6791, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.5543673038482666, + "learning_rate": 4.041429731925264e-06, + "loss": 1.6761, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.3913605213165283, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7579, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.3837356567382812, + "learning_rate": 4.037367993501219e-06, + "loss": 1.6636, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.6936402320861816, + "learning_rate": 4.035337124289196e-06, + "loss": 1.6488, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.6400210857391357, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.6926, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.574533462524414, + "learning_rate": 4.031275385865151e-06, + "loss": 1.7066, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.5130677223205566, + "learning_rate": 4.029244516653128e-06, + "loss": 1.6431, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.578166961669922, + "learning_rate": 4.027213647441105e-06, + "loss": 1.648, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.4736695289611816, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6822, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.8754831552505493, + "eval_runtime": 50.0971, + "eval_samples_per_second": 19.961, + "eval_steps_per_second": 0.838, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.544315814971924, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6853, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.8765335083007812, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7587, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.4749417304992676, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.6536, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.493164300918579, + "learning_rate": 4.017059301380992e-06, + "loss": 1.7436, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.6274492740631104, + "learning_rate": 4.015028432168969e-06, + "loss": 1.632, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.4162979125976562, + "learning_rate": 4.012997562956945e-06, + "loss": 1.6663, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.5683343410491943, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6771, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.3710429668426514, + "learning_rate": 4.0089358245329e-06, + "loss": 1.6644, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.5356292724609375, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7339, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.480884075164795, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.716, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.867702841758728, + "eval_runtime": 48.0038, + "eval_samples_per_second": 20.832, + "eval_steps_per_second": 0.875, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4768476486206055, + "learning_rate": 4.002843216896833e-06, + "loss": 1.6667, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5727357864379883, + "learning_rate": 4.000812347684809e-06, + "loss": 1.6786, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.6243903636932373, + "learning_rate": 3.998781478472787e-06, + "loss": 1.6328, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.712955951690674, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7852, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.756272554397583, + "learning_rate": 3.994719740048742e-06, + "loss": 1.7006, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.919201135635376, + "learning_rate": 3.992688870836718e-06, + "loss": 1.6624, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.471566915512085, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7122, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.46273136138916, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6621, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.5338189601898193, + "learning_rate": 3.98659626320065e-06, + "loss": 1.6869, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.3850855827331543, + "learning_rate": 3.984565393988627e-06, + "loss": 1.6925, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.8650325536727905, + "eval_runtime": 49.4167, + "eval_samples_per_second": 20.236, + "eval_steps_per_second": 0.85, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.654726028442383, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7113, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.5065038204193115, + "learning_rate": 3.980503655564582e-06, + "loss": 1.7078, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.51668381690979, + "learning_rate": 3.978472786352559e-06, + "loss": 1.6742, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4586126804351807, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7339, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.452857732772827, + "learning_rate": 3.974411047928514e-06, + "loss": 1.6993, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.608631134033203, + "learning_rate": 3.972380178716491e-06, + "loss": 1.7047, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.34022855758667, + "learning_rate": 3.970349309504468e-06, + "loss": 1.6423, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.329612970352173, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.6443, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.3301169872283936, + "learning_rate": 3.966287571080423e-06, + "loss": 1.6937, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.3010976314544678, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6896, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.863387942314148, + "eval_runtime": 47.6761, + "eval_samples_per_second": 20.975, + "eval_steps_per_second": 0.881, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.296764612197876, + "learning_rate": 3.962225832656377e-06, + "loss": 1.6098, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.437114953994751, + "learning_rate": 3.960194963444355e-06, + "loss": 1.6727, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4908952713012695, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6816, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.81123685836792, + "learning_rate": 3.956133225020309e-06, + "loss": 1.6795, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.7088773250579834, + "learning_rate": 3.954102355808286e-06, + "loss": 1.6838, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.5021824836730957, + "learning_rate": 3.952071486596264e-06, + "loss": 1.671, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.3021023273468018, + "learning_rate": 3.950040617384241e-06, + "loss": 1.651, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.866119861602783, + "learning_rate": 3.948009748172218e-06, + "loss": 1.7117, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.3471503257751465, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.6883, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4612059593200684, + "learning_rate": 3.943948009748173e-06, + "loss": 1.6898, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.8587063550949097, + "eval_runtime": 43.4462, + "eval_samples_per_second": 23.017, + "eval_steps_per_second": 0.967, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.376680374145508, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7253, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.4701383113861084, + "learning_rate": 3.939886271324127e-06, + "loss": 1.6306, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.574384927749634, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6287, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.8596720695495605, + "learning_rate": 3.935824532900082e-06, + "loss": 1.7031, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.7144267559051514, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7219, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.5807480812072754, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.6364, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 3.3094675540924072, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6526, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.390503168106079, + "learning_rate": 3.927701056051991e-06, + "loss": 1.6795, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.5525588989257812, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.6432, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.522307872772217, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6218, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.8604834079742432, + "eval_runtime": 46.6465, + "eval_samples_per_second": 21.438, + "eval_steps_per_second": 0.9, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.438201904296875, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6788, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.639164686203003, + "learning_rate": 3.9195775792039e-06, + "loss": 1.6552, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.608038902282715, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.69, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.4357974529266357, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7031, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.4964048862457275, + "learning_rate": 3.913484971567831e-06, + "loss": 1.6126, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.5115931034088135, + "learning_rate": 3.911454102355808e-06, + "loss": 1.684, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.54923152923584, + "learning_rate": 3.909423233143786e-06, + "loss": 1.6815, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.849602222442627, + "learning_rate": 3.907392363931763e-06, + "loss": 1.744, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.7209551334381104, + "learning_rate": 3.90536149471974e-06, + "loss": 1.755, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.602299451828003, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.7835, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.8590067625045776, + "eval_runtime": 49.5182, + "eval_samples_per_second": 20.195, + "eval_steps_per_second": 0.848, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.504567861557007, + "learning_rate": 3.901299756295695e-06, + "loss": 1.6595, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.456721305847168, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6692, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.319553852081299, + "learning_rate": 3.897238017871649e-06, + "loss": 1.6664, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.45987606048584, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.65, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 3.0049448013305664, + "learning_rate": 3.893176279447604e-06, + "loss": 1.7366, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.522472858428955, + "learning_rate": 3.891145410235581e-06, + "loss": 1.7022, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.3914318084716797, + "learning_rate": 3.889114541023558e-06, + "loss": 1.6591, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.462167501449585, + "learning_rate": 3.887083671811536e-06, + "loss": 1.6158, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.6707093715667725, + "learning_rate": 3.885052802599513e-06, + "loss": 1.7453, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.493745803833008, + "learning_rate": 3.88302193338749e-06, + "loss": 1.6836, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.8536967039108276, + "eval_runtime": 50.3456, + "eval_samples_per_second": 19.863, + "eval_steps_per_second": 0.834, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.5338919162750244, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.65, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.6954877376556396, + "learning_rate": 3.878960194963445e-06, + "loss": 1.7156, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.4878089427948, + "learning_rate": 3.876929325751422e-06, + "loss": 1.6053, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.4329843521118164, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6642, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 3.118771553039551, + "learning_rate": 3.872867587327377e-06, + "loss": 1.6392, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4644217491149902, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6566, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.4643497467041016, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6654, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.7067277431488037, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.682, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.8748488426208496, + "learning_rate": 3.864744110479286e-06, + "loss": 1.6886, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.632155179977417, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7069, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8592487573623657, + "eval_runtime": 47.0773, + "eval_samples_per_second": 21.242, + "eval_steps_per_second": 0.892, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.5379488468170166, + "learning_rate": 3.86068237205524e-06, + "loss": 1.6695, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.571742057800293, + "learning_rate": 3.858651502843217e-06, + "loss": 1.6235, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.5355312824249268, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6528, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.715510606765747, + "learning_rate": 3.854589764419171e-06, + "loss": 1.7379, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.635364532470703, + "learning_rate": 3.852558895207149e-06, + "loss": 1.6958, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.4713191986083984, + "learning_rate": 3.850528025995126e-06, + "loss": 1.6471, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.584200382232666, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6548, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.4639663696289062, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6614, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.4823923110961914, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.6523, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.953594923019409, + "learning_rate": 3.842404549147035e-06, + "loss": 1.6281, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8549022674560547, + "eval_runtime": 47.2215, + "eval_samples_per_second": 21.177, + "eval_steps_per_second": 0.889, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.524352550506592, + "learning_rate": 3.840373679935013e-06, + "loss": 1.6741, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.411515235900879, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.6571, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.6827003955841064, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6927, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 3.3486077785491943, + "learning_rate": 3.834281072298944e-06, + "loss": 1.6347, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.4518957138061523, + "learning_rate": 3.832250203086921e-06, + "loss": 1.7011, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.623842239379883, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7459, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.689602851867676, + "learning_rate": 3.828188464662876e-06, + "loss": 1.6812, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.8425960540771484, + "learning_rate": 3.826157595450853e-06, + "loss": 1.6258, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.6639413833618164, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6769, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.4269535541534424, + "learning_rate": 3.822095857026808e-06, + "loss": 1.6363, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8489640951156616, + "eval_runtime": 45.897, + "eval_samples_per_second": 21.788, + "eval_steps_per_second": 0.915, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.3430261611938477, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6563, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.322094678878784, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6588, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.6172726154327393, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.681, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.574084997177124, + "learning_rate": 3.813972380178717e-06, + "loss": 1.6977, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.437417984008789, + "learning_rate": 3.811941510966694e-06, + "loss": 1.6204, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.319276809692383, + "learning_rate": 3.809910641754671e-06, + "loss": 1.5858, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5540425777435303, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6622, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.567634344100952, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.635, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4815967082977295, + "learning_rate": 3.803818034118603e-06, + "loss": 1.7146, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.475162982940674, + "learning_rate": 3.80178716490658e-06, + "loss": 1.6911, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.848276138305664, + "eval_runtime": 44.1011, + "eval_samples_per_second": 22.675, + "eval_steps_per_second": 0.952, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5385165214538574, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.6523, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.443129301071167, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6239, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.696885347366333, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.7318, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.8570339679718018, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7198, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.475338935852051, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7052, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.5250487327575684, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.7015, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.812121868133545, + "learning_rate": 3.787571080422421e-06, + "loss": 1.6876, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.486117362976074, + "learning_rate": 3.785540211210398e-06, + "loss": 1.6695, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.5146725177764893, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.6585, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.5808019638061523, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.7608, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8411076068878174, + "eval_runtime": 48.0798, + "eval_samples_per_second": 20.799, + "eval_steps_per_second": 0.874, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.509993076324463, + "learning_rate": 3.77944760357433e-06, + "loss": 1.7101, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.492285966873169, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.6417, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.809333086013794, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6963, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.524418354034424, + "learning_rate": 3.773354995938262e-06, + "loss": 1.6957, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.5848705768585205, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7193, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.8617639541625977, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.6721, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.4655680656433105, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.586, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 3.0633299350738525, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6792, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.3882107734680176, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.6229, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.5860483646392822, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6753, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8393676280975342, + "eval_runtime": 44.3838, + "eval_samples_per_second": 22.531, + "eval_steps_per_second": 0.946, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.930290460586548, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.6955, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.8278939723968506, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.6455, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.3822925090789795, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.7163, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.6904139518737793, + "learning_rate": 3.753046303818034e-06, + "loss": 1.6942, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.535284996032715, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7237, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.669031858444214, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.6391, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.5162742137908936, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6663, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.5937418937683105, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7115, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.4892170429229736, + "learning_rate": 3.742891957757921e-06, + "loss": 1.6727, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.7901015281677246, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7179, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8420112133026123, + "eval_runtime": 42.0904, + "eval_samples_per_second": 23.758, + "eval_steps_per_second": 0.998, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.510568618774414, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.6408, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.5084173679351807, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6495, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.3693647384643555, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.6064, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.520414352416992, + "learning_rate": 3.732737611697807e-06, + "loss": 1.5668, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.544624090194702, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.6231, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.6257777214050293, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6093, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 3.0384273529052734, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.7291, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.718588352203369, + "learning_rate": 3.724614134849716e-06, + "loss": 1.604, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.655592918395996, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6993, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.569904088973999, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.6595, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8419173955917358, + "eval_runtime": 42.5535, + "eval_samples_per_second": 23.5, + "eval_steps_per_second": 0.987, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.643686056137085, + "learning_rate": 3.718521527213648e-06, + "loss": 1.6483, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.7431893348693848, + "learning_rate": 3.716490658001625e-06, + "loss": 1.5418, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 3.0191054344177246, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.6425, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.695777654647827, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.651, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.512838840484619, + "learning_rate": 3.710398050365557e-06, + "loss": 1.6181, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.9957621097564697, + "learning_rate": 3.708367181153534e-06, + "loss": 1.6454, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.5258326530456543, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.5995, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.6610865592956543, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.5816, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.825767993927002, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6402, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.605870008468628, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6282, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8433754444122314, + "eval_runtime": 45.2127, + "eval_samples_per_second": 22.118, + "eval_steps_per_second": 0.929, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.7334296703338623, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.5958, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.9007229804992676, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.7128, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.8827497959136963, + "learning_rate": 3.694151096669375e-06, + "loss": 1.6157, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6470541954040527, + "learning_rate": 3.692120227457352e-06, + "loss": 1.643, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.5725178718566895, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.5481, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.7779407501220703, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.6241, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 3.161653995513916, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.644, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.5995497703552246, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6475, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.607166290283203, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6573, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.4532339572906494, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.6044, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.8344279527664185, + "eval_runtime": 43.2125, + "eval_samples_per_second": 23.141, + "eval_steps_per_second": 0.972, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.540632486343384, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6308, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.40977144241333, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6134, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.6345880031585693, + "learning_rate": 3.673842404549147e-06, + "loss": 1.615, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.565157413482666, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6351, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.7868876457214355, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.7186, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.7061610221862793, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.6454, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.587785482406616, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6867, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.4618277549743652, + "learning_rate": 3.663688058489034e-06, + "loss": 1.6189, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.4425477981567383, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.5772, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.7514469623565674, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6368, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8366374969482422, + "eval_runtime": 44.4679, + "eval_samples_per_second": 22.488, + "eval_steps_per_second": 0.945, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.529805898666382, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.6229, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6875312328338623, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6445, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.715970993041992, + "learning_rate": 3.65353371242892e-06, + "loss": 1.5318, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.8685660362243652, + "learning_rate": 3.651502843216897e-06, + "loss": 1.6465, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.6801602840423584, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.5959, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.61037015914917, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6561, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.478137493133545, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6435, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5708508491516113, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6206, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.8857522010803223, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6269, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.6102395057678223, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.648, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.836978793144226, + "eval_runtime": 44.9452, + "eval_samples_per_second": 22.249, + "eval_steps_per_second": 0.934, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.7924158573150635, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6802, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.784693479537964, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.6515, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.747894525527954, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6093, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.635850191116333, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6791, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.5760908126831055, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.5721, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5258584022521973, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.5644, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.919623613357544, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.6258, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.8330483436584473, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6077, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.8145744800567627, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6463, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.9187726974487305, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6348, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.837236762046814, + "eval_runtime": 45.4421, + "eval_samples_per_second": 22.006, + "eval_steps_per_second": 0.924, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.6741979122161865, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6441, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.556701898574829, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5523, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.6352901458740234, + "learning_rate": 3.612916328188465e-06, + "loss": 1.6545, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.5784106254577637, + "learning_rate": 3.610885458976442e-06, + "loss": 1.6649, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.7136101722717285, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.596, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.608877182006836, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.5923, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.783069372177124, + "learning_rate": 3.604792851340374e-06, + "loss": 1.6685, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.621882438659668, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6404, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.672128915786743, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.5713, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.645395040512085, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.609, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8348923921585083, + "eval_runtime": 44.8379, + "eval_samples_per_second": 22.303, + "eval_steps_per_second": 0.937, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 3.610335111618042, + "learning_rate": 3.596669374492283e-06, + "loss": 1.6693, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.7391626834869385, + "learning_rate": 3.59463850528026e-06, + "loss": 1.5847, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.658418655395508, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6552, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.6508545875549316, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6201, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 4.237835884094238, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.6725, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.672144889831543, + "learning_rate": 3.586515028432169e-06, + "loss": 1.6666, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.619480609893799, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.5671, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 3.150592565536499, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6118, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.946467161178589, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6499, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.6372971534729004, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5516, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8352437019348145, + "eval_runtime": 46.1704, + "eval_samples_per_second": 21.659, + "eval_steps_per_second": 0.91, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.6211962699890137, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6414, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.567579746246338, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.6632, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.4393808841705322, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6323, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.7005021572113037, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6016, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7400808334350586, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6327, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.922241449356079, + "learning_rate": 3.566206336311942e-06, + "loss": 1.5866, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.524975299835205, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5707, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.637805461883545, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6259, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.7430527210235596, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.595, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.5548908710479736, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.625, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8332648277282715, + "eval_runtime": 48.5515, + "eval_samples_per_second": 20.597, + "eval_steps_per_second": 0.865, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.5396368503570557, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.655, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.523167848587036, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6171, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.6028409004211426, + "learning_rate": 3.551990251827783e-06, + "loss": 1.5872, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.607464075088501, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6176, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.510603427886963, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.5999, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.444693088531494, + "learning_rate": 3.545897644191714e-06, + "loss": 1.5791, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6724541187286377, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6358, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.64925217628479, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.5793, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.647118091583252, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6399, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.8507258892059326, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6046, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8305917978286743, + "eval_runtime": 43.2738, + "eval_samples_per_second": 23.109, + "eval_steps_per_second": 0.971, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.5572781562805176, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6395, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.530216693878174, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6493, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.564739465713501, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6075, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.5772671699523926, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.64, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.828692674636841, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6584, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.700568675994873, + "learning_rate": 3.525588952071487e-06, + "loss": 1.6595, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.666743755340576, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6483, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.551187515258789, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.5946, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.553602933883667, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.5997, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.5355968475341797, + "learning_rate": 3.517465475223396e-06, + "loss": 1.571, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8321614265441895, + "eval_runtime": 46.1746, + "eval_samples_per_second": 21.657, + "eval_steps_per_second": 0.91, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.707101345062256, + "learning_rate": 3.515434606011373e-06, + "loss": 1.6389, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.8770599365234375, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.6828, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.5841145515441895, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6264, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.697502851486206, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6081, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.620100259780884, + "learning_rate": 3.507311129163282e-06, + "loss": 1.5617, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.6503913402557373, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6194, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.830296754837036, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.616, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.6960535049438477, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5523, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.59023118019104, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6378, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.582906484603882, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6368, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8257222175598145, + "eval_runtime": 47.9573, + "eval_samples_per_second": 20.852, + "eval_steps_per_second": 0.876, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.6826765537261963, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.5685, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6153650283813477, + "learning_rate": 3.493095044679123e-06, + "loss": 1.6948, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.607027769088745, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5197, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.493180990219116, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6159, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.6291017532348633, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5404, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.90544056892395, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6164, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.673076629638672, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6486, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.5843210220336914, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.6127, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.55414080619812, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6255, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.496095657348633, + "learning_rate": 3.476848090982941e-06, + "loss": 1.5941, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8262794017791748, + "eval_runtime": 47.0577, + "eval_samples_per_second": 21.25, + "eval_steps_per_second": 0.893, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.597395658493042, + "learning_rate": 3.474817221770918e-06, + "loss": 1.6941, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.4487929344177246, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6417, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.63041353225708, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.611, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.481630325317383, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.575, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.7625951766967773, + "learning_rate": 3.466693744922827e-06, + "loss": 1.5988, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.663846492767334, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6326, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.594536781311035, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6004, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.5871670246124268, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6378, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.591907262802124, + "learning_rate": 3.458570268074736e-06, + "loss": 1.6093, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.456110954284668, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6014, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8266254663467407, + "eval_runtime": 47.4255, + "eval_samples_per_second": 21.086, + "eval_steps_per_second": 0.886, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.8052752017974854, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.5679, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.7458856105804443, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6239, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.4829750061035156, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.5692, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.5427238941192627, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6214, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.597419261932373, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6012, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.5192158222198486, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6469, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.4233906269073486, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.5822, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.4213144779205322, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5345, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.763439655303955, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6092, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 3.228456497192383, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6591, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8253848552703857, + "eval_runtime": 50.1479, + "eval_samples_per_second": 19.941, + "eval_steps_per_second": 0.838, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.670700788497925, + "learning_rate": 3.434199837530463e-06, + "loss": 1.612, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.797170877456665, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5624, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8155477046966553, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6535, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.7093656063079834, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6479, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.6133127212524414, + "learning_rate": 3.426076360682372e-06, + "loss": 1.6669, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.471862554550171, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6136, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.7078938484191895, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5382, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.6206846237182617, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6186, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.674201011657715, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.6805, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.7543511390686035, + "learning_rate": 3.415922014622259e-06, + "loss": 1.644, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8275290727615356, + "eval_runtime": 49.1382, + "eval_samples_per_second": 20.351, + "eval_steps_per_second": 0.855, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.934906244277954, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6502, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.6866018772125244, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6324, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.743281126022339, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6229, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.7543251514434814, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.5813, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.8038830757141113, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6482, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.6368162631988525, + "learning_rate": 3.403736799350122e-06, + "loss": 1.618, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7447240352630615, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.5735, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.6714727878570557, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.686, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.7939751148223877, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6104, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.624718189239502, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6194, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8199249505996704, + "eval_runtime": 50.4903, + "eval_samples_per_second": 19.806, + "eval_steps_per_second": 0.832, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.563082695007324, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5669, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.6971375942230225, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6237, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.7873635292053223, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6241, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.570742130279541, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6304, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.8116018772125244, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6068, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.6097946166992188, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.5996, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.593461275100708, + "learning_rate": 3.381397238017872e-06, + "loss": 1.5935, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 3.008805274963379, + "learning_rate": 3.379366368805849e-06, + "loss": 1.68, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7567968368530273, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6129, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.647456645965576, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.618, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8231184482574463, + "eval_runtime": 46.6144, + "eval_samples_per_second": 21.453, + "eval_steps_per_second": 0.901, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.842937707901001, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6118, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.794092893600464, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.5922, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.837575674057007, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6267, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.7084648609161377, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5597, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.626157283782959, + "learning_rate": 3.36515028432169e-06, + "loss": 1.5961, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.589062213897705, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6235, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.681427478790283, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6568, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 3.3202719688415527, + "learning_rate": 3.359057676685622e-06, + "loss": 1.649, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.463998794555664, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5482, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.5475661754608154, + "learning_rate": 3.354995938261576e-06, + "loss": 1.586, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8188257217407227, + "eval_runtime": 51.7434, + "eval_samples_per_second": 19.326, + "eval_steps_per_second": 0.812, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.579566478729248, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5546, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6778531074523926, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.5831, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.98976731300354, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6199, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.8345465660095215, + "learning_rate": 3.346872461413485e-06, + "loss": 1.5697, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.6166720390319824, + "learning_rate": 3.344841592201463e-06, + "loss": 1.5829, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.5504395961761475, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.5845, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.6515390872955322, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6154, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.589524269104004, + "learning_rate": 3.338748984565394e-06, + "loss": 1.639, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.8798511028289795, + "learning_rate": 3.336718115353372e-06, + "loss": 1.676, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.984379529953003, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6154, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.818778157234192, + "eval_runtime": 44.8621, + "eval_samples_per_second": 22.291, + "eval_steps_per_second": 0.936, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.5935728549957275, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.5895, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.7658374309539795, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6041, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.580352306365967, + "learning_rate": 3.32859463850528e-06, + "loss": 1.6127, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.689892292022705, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5624, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.708078622817993, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.5782, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.7298295497894287, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.5815, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.8700411319732666, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5597, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.6671183109283447, + "learning_rate": 3.318440292445167e-06, + "loss": 1.5902, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 3.11922550201416, + "learning_rate": 3.316409423233144e-06, + "loss": 1.5678, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.72869610786438, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6568, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8164011240005493, + "eval_runtime": 47.4229, + "eval_samples_per_second": 21.087, + "eval_steps_per_second": 0.886, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.7037997245788574, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6191, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.6406683921813965, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6018, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.4749815464019775, + "learning_rate": 3.308285946385053e-06, + "loss": 1.679, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.5094757080078125, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5455, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.6685080528259277, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6123, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7547595500946045, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.6497, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.7807092666625977, + "learning_rate": 3.300162469536962e-06, + "loss": 1.5903, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.903127431869507, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6398, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.4666707515716553, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6194, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.9508626461029053, + "learning_rate": 3.294069861900894e-06, + "loss": 1.6774, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8119163513183594, + "eval_runtime": 51.0141, + "eval_samples_per_second": 19.602, + "eval_steps_per_second": 0.823, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.585411787033081, + "learning_rate": 3.292038992688871e-06, + "loss": 1.6086, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.7636921405792236, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.6386, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.72711181640625, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.6563, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.5698368549346924, + "learning_rate": 3.285946385052803e-06, + "loss": 1.5501, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.8764665126800537, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6229, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.653674840927124, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5343, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.896165609359741, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.6617, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.714388847351074, + "learning_rate": 3.277822908204712e-06, + "loss": 1.605, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.572202682495117, + "learning_rate": 3.275792038992689e-06, + "loss": 1.568, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.719407081604004, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6116, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.811676263809204, + "eval_runtime": 46.0392, + "eval_samples_per_second": 21.721, + "eval_steps_per_second": 0.912, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.752934694290161, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.5672, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.519256353378296, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6101, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.758317708969116, + "learning_rate": 3.267668562144598e-06, + "loss": 1.597, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.6348824501037598, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.5731, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.699477434158325, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6288, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.5104498863220215, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6024, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.680694341659546, + "learning_rate": 3.259545085296507e-06, + "loss": 1.6689, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.459789276123047, + "learning_rate": 3.257514216084484e-06, + "loss": 1.56, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.533125877380371, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.6734, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.640742301940918, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6214, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8131158351898193, + "eval_runtime": 47.2273, + "eval_samples_per_second": 21.174, + "eval_steps_per_second": 0.889, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.4593327045440674, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.5997, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.6157352924346924, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6193, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.5011303424835205, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.5796, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.545781373977661, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.6128, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.5882468223571777, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6106, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.7500388622283936, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5491, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5432305335998535, + "learning_rate": 3.23923639317628e-06, + "loss": 1.5879, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.5674941539764404, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.5737, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.630491256713867, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6167, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.7517776489257812, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.6698, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8172497749328613, + "eval_runtime": 45.7554, + "eval_samples_per_second": 21.855, + "eval_steps_per_second": 0.918, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5155255794525146, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.5647, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.5586776733398438, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6403, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.7582011222839355, + "learning_rate": 3.227051177904143e-06, + "loss": 1.5901, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.611339569091797, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.599, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.7254250049591064, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6167, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.5621025562286377, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5241, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849069118499756, + "learning_rate": 3.218927701056052e-06, + "loss": 1.5823, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.874762535095215, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5294, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.706822395324707, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.6529, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.7066354751586914, + "learning_rate": 3.212835093419984e-06, + "loss": 1.5591, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.814163327217102, + "eval_runtime": 43.7518, + "eval_samples_per_second": 22.856, + "eval_steps_per_second": 0.96, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.584439516067505, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.5986, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.671645402908325, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6417, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6757681369781494, + "learning_rate": 3.206742485783916e-06, + "loss": 1.5594, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.6210391521453857, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5476, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.5632853507995605, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6165, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.64302921295166, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.631, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.4642136096954346, + "learning_rate": 3.198619008935825e-06, + "loss": 1.534, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6485342979431152, + "learning_rate": 3.196588139723802e-06, + "loss": 1.598, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.5753607749938965, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6177, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.787393093109131, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.607, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.809692621231079, + "eval_runtime": 44.1276, + "eval_samples_per_second": 22.662, + "eval_steps_per_second": 0.952, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.4849026203155518, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6064, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.484736204147339, + "learning_rate": 3.188464662875711e-06, + "loss": 1.6591, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.652766466140747, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.5931, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.6450769901275635, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5463, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.736511468887329, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5509, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.8212637901306152, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6244, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.7004551887512207, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6123, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.665766954421997, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.5904, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.7625415325164795, + "learning_rate": 3.174248578391552e-06, + "loss": 1.5843, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.5376832485198975, + "learning_rate": 3.172217709179529e-06, + "loss": 1.5758, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8079267740249634, + "eval_runtime": 45.139, + "eval_samples_per_second": 22.154, + "eval_steps_per_second": 0.93, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.704226016998291, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.5891, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6165294647216797, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5502, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.6848294734954834, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6134, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.778846502304077, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.611, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.647531747817993, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5531, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.605592966079712, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6011, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5820205211639404, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6121, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.7092363834381104, + "learning_rate": 3.155970755483347e-06, + "loss": 1.5928, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.7473032474517822, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6068, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.742840528488159, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6041, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8063452243804932, + "eval_runtime": 44.2112, + "eval_samples_per_second": 22.619, + "eval_steps_per_second": 0.95, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.7613816261291504, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6413, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.876913547515869, + "learning_rate": 3.147847278635256e-06, + "loss": 1.6848, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5945441722869873, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.5847, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.526456117630005, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5496, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.9870903491973877, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.5982, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6774742603302, + "learning_rate": 3.139723801787165e-06, + "loss": 1.5678, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.69058895111084, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6174, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.8981339931488037, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5529, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.522247552871704, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6061, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.709726333618164, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6267, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8052302598953247, + "eval_runtime": 42.6337, + "eval_samples_per_second": 23.456, + "eval_steps_per_second": 0.985, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.618793249130249, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6441, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5927956104278564, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6035, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.557023048400879, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6252, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5910511016845703, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.5883, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.6681272983551025, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6147, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.7024455070495605, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6317, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.6808700561523438, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.5871, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.585388422012329, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.6202, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.637092113494873, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6357, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.652611494064331, + "learning_rate": 3.111291632818847e-06, + "loss": 1.651, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.804667592048645, + "eval_runtime": 43.1932, + "eval_samples_per_second": 23.152, + "eval_steps_per_second": 0.972, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.770113945007324, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6005, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.570833206176758, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6169, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.537957191467285, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.5826, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.715669631958008, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6276, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.5450210571289062, + "learning_rate": 3.101137286758733e-06, + "loss": 1.6305, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.6358814239501953, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6367, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.7380471229553223, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.5984, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 3.030034303665161, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.6661, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.6777493953704834, + "learning_rate": 3.093013809910642e-06, + "loss": 1.6122, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8807458877563477, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.6514, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8064097166061401, + "eval_runtime": 43.3673, + "eval_samples_per_second": 23.059, + "eval_steps_per_second": 0.968, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.545091390609741, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6182, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.6389622688293457, + "learning_rate": 3.086921202274574e-06, + "loss": 1.5931, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.5784780979156494, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5579, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.67499041557312, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6128, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.558598756790161, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.5958, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6460611820220947, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5274, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.770334482192993, + "learning_rate": 3.07676685621446e-06, + "loss": 1.604, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6441917419433594, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.567, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.6142544746398926, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.5891, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.6242077350616455, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6059, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8036692142486572, + "eval_runtime": 45.6003, + "eval_samples_per_second": 21.93, + "eval_steps_per_second": 0.921, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.417332410812378, + "learning_rate": 3.068643379366369e-06, + "loss": 1.5991, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.4537513256073, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5603, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.745912551879883, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6171, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.636627674102783, + "learning_rate": 3.062550771730301e-06, + "loss": 1.5839, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.6001460552215576, + "learning_rate": 3.060519902518278e-06, + "loss": 1.5781, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.572185516357422, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5188, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.87540864944458, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.621, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.7670910358428955, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.584, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.672682523727417, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6293, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.655538558959961, + "learning_rate": 3.050365556458164e-06, + "loss": 1.5773, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8046640157699585, + "eval_runtime": 44.6243, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.582319974899292, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.5751, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.7160403728485107, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.5972, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.6800436973571777, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6311, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.8212592601776123, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6315, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.57470703125, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5414, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 3.3979809284210205, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.5732, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.606290102005005, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.5966, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.5241527557373047, + "learning_rate": 3.034118602761982e-06, + "loss": 1.588, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.7115085124969482, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5619, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.7072222232818604, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6346, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.804133415222168, + "eval_runtime": 44.5547, + "eval_samples_per_second": 22.444, + "eval_steps_per_second": 0.943, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.5874669551849365, + "learning_rate": 3.028025995125914e-06, + "loss": 1.623, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.644073247909546, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5484, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.553112506866455, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.5923, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6504719257354736, + "learning_rate": 3.021933387489846e-06, + "loss": 1.543, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.7608838081359863, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6001, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.690267562866211, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6093, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.8565986156463623, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5614, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.672799587249756, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5657, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5756146907806396, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.5664, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.8423633575439453, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.5936, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.800963044166565, + "eval_runtime": 46.5083, + "eval_samples_per_second": 21.502, + "eval_steps_per_second": 0.903, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.6263201236724854, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.5693, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 3.0006930828094482, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6084, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.7002620697021484, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.5667, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.722398519515991, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6285, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.738555908203125, + "learning_rate": 2.999593826157596e-06, + "loss": 1.5849, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.726271390914917, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.5951, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.6749048233032227, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.5807, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.774726629257202, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6355, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6740102767944336, + "learning_rate": 2.991470349309505e-06, + "loss": 1.5871, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.7571794986724854, + "learning_rate": 2.989439480097482e-06, + "loss": 1.578, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.7960126399993896, + "eval_runtime": 44.4777, + "eval_samples_per_second": 22.483, + "eval_steps_per_second": 0.944, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.67576265335083, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.5917, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.8302817344665527, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6313, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.874002456665039, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6212, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6992602348327637, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6011, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.726292371749878, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.5981, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.988934278488159, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6272, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.6781861782073975, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.5742, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.818401336669922, + "learning_rate": 2.9731925264013e-06, + "loss": 1.5713, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.6900908946990967, + "learning_rate": 2.971161657189277e-06, + "loss": 1.5754, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.9017207622528076, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.5746, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.7994096279144287, + "eval_runtime": 46.997, + "eval_samples_per_second": 21.278, + "eval_steps_per_second": 0.894, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.0305914878845215, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.5733, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.695298910140991, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6379, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.60552716255188, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5488, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.664668321609497, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6256, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.700955390930176, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6228, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.637998104095459, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.5565, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.664620876312256, + "learning_rate": 2.954914703493095e-06, + "loss": 1.5805, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.8120782375335693, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.5615, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.641434907913208, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.56, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.845808982849121, + "learning_rate": 2.948822095857027e-06, + "loss": 1.5954, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8004399538040161, + "eval_runtime": 47.1938, + "eval_samples_per_second": 21.189, + "eval_steps_per_second": 0.89, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.713960886001587, + "learning_rate": 2.946791226645004e-06, + "loss": 1.5628, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.8967084884643555, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.5445, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 3.3338470458984375, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.5735, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.864295721054077, + "learning_rate": 2.940698619008936e-06, + "loss": 1.5892, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.704193353652954, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.5916, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.6473755836486816, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.6059, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.610670804977417, + "learning_rate": 2.934606011372868e-06, + "loss": 1.6012, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.6023919582366943, + "learning_rate": 2.932575142160845e-06, + "loss": 1.5973, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8217804431915283, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.5794, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.8418939113616943, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.5942, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.8003290891647339, + "eval_runtime": 44.3696, + "eval_samples_per_second": 22.538, + "eval_steps_per_second": 0.947, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.6487412452697754, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.6297, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 2.6183269023895264, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.5916, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.6346395015716553, + "learning_rate": 2.922420796100731e-06, + "loss": 1.5209, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.631850242614746, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.6268, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.7087175846099854, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.5348, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.70473575592041, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.5943, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 3.0979015827178955, + "learning_rate": 2.91429731925264e-06, + "loss": 1.6308, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.6136019229888916, + "learning_rate": 2.912266450040618e-06, + "loss": 1.5955, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.6788594722747803, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.5909, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 3.681739330291748, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.5794, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.7974433898925781, + "eval_runtime": 50.5413, + "eval_samples_per_second": 19.786, + "eval_steps_per_second": 0.831, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.543687582015991, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.5455, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 3.0568184852600098, + "learning_rate": 2.904142973192527e-06, + "loss": 1.6445, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.557318925857544, + "learning_rate": 2.902112103980504e-06, + "loss": 1.5933, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.4658801555633545, + "learning_rate": 2.900081234768481e-06, + "loss": 1.5899, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.610834836959839, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.5557, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.643775224685669, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.5523, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.8988914489746094, + "learning_rate": 2.893988627132413e-06, + "loss": 1.5787, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.686321973800659, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.5205, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.6282148361206055, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.5942, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.581450939178467, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.5779, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.7954648733139038, + "eval_runtime": 45.0426, + "eval_samples_per_second": 22.201, + "eval_steps_per_second": 0.932, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.6617720127105713, + "learning_rate": 2.885865150284322e-06, + "loss": 1.6168, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.643260955810547, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.5764, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.5922341346740723, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.5385, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.7868714332580566, + "learning_rate": 2.879772542648254e-06, + "loss": 1.5905, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.6874098777770996, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.516, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.9882795810699463, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.6497, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.574932336807251, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.6217, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 2.625720739364624, + "learning_rate": 2.871649065800163e-06, + "loss": 1.6436, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 2.6322028636932373, + "learning_rate": 2.86961819658814e-06, + "loss": 1.5743, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.753891944885254, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.594, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.7947734594345093, + "eval_runtime": 47.0036, + "eval_samples_per_second": 21.275, + "eval_steps_per_second": 0.894, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.6878788471221924, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.5801, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.7409934997558594, + "learning_rate": 2.863525588952072e-06, + "loss": 1.6251, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.517815351486206, + "learning_rate": 2.861494719740049e-06, + "loss": 1.5055, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.7027196884155273, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.5536, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.8066604137420654, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.5819, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.7119011878967285, + "learning_rate": 2.855402112103981e-06, + "loss": 1.5776, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.6952199935913086, + "learning_rate": 2.853371242891958e-06, + "loss": 1.5682, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.7582387924194336, + "learning_rate": 2.851340373679935e-06, + "loss": 1.5863, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.8032495975494385, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.6204, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.6014182567596436, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.5816, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.7986392974853516, + "eval_runtime": 45.7419, + "eval_samples_per_second": 21.862, + "eval_steps_per_second": 0.918, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.611240863800049, + "learning_rate": 2.845247766043867e-06, + "loss": 1.5563, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.604238510131836, + "learning_rate": 2.843216896831844e-06, + "loss": 1.5143, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.531214475631714, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.5579, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.6469998359680176, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.6049, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 3.1636979579925537, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.6072, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.6377527713775635, + "learning_rate": 2.835093419983753e-06, + "loss": 1.574, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.6896755695343018, + "learning_rate": 2.833062550771731e-06, + "loss": 1.5498, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.6172642707824707, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.5413, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.7607314586639404, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.5816, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.5872104167938232, + "learning_rate": 2.826969943135662e-06, + "loss": 1.5868, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.7989377975463867, + "eval_runtime": 46.6112, + "eval_samples_per_second": 21.454, + "eval_steps_per_second": 0.901, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.517178773880005, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.5875, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.531451940536499, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.5583, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.6367409229278564, + "learning_rate": 2.820877335499594e-06, + "loss": 1.5605, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.496851682662964, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.5227, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.583904981613159, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.5434, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.7455286979675293, + "learning_rate": 2.814784727863526e-06, + "loss": 1.5732, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.6893272399902344, + "learning_rate": 2.812753858651503e-06, + "loss": 1.5807, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 2.6912310123443604, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.5484, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 2.6918725967407227, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.535, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 2.808300256729126, + "learning_rate": 2.806661251015435e-06, + "loss": 1.6278, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.8001086711883545, + "eval_runtime": 50.3556, + "eval_samples_per_second": 19.859, + "eval_steps_per_second": 0.834, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.702766180038452, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.6223, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 2.890214681625366, + "learning_rate": 2.802599512591389e-06, + "loss": 1.68, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.6849842071533203, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.5836, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.678279161453247, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.529, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 2.782949686050415, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.5814, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.8183817863464355, + "learning_rate": 2.794476035743298e-06, + "loss": 1.6401, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.917726516723633, + "learning_rate": 2.792445166531276e-06, + "loss": 1.5936, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 2.8861331939697266, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.6213, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.7103989124298096, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.5537, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.876448392868042, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.5464, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.7939362525939941, + "eval_runtime": 47.8308, + "eval_samples_per_second": 20.907, + "eval_steps_per_second": 0.878, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.613130569458008, + "learning_rate": 2.784321689683185e-06, + "loss": 1.5427, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.553765058517456, + "learning_rate": 2.782290820471162e-06, + "loss": 1.5496, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.7586376667022705, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.6221, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.6437292098999023, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.6075, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.6939804553985596, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.5918, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.6177403926849365, + "learning_rate": 2.774167343623071e-06, + "loss": 1.6353, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 2.8386614322662354, + "learning_rate": 2.772136474411048e-06, + "loss": 1.5548, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.6705095767974854, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.497, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.790919065475464, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.5945, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.7193119525909424, + "learning_rate": 2.76604386677498e-06, + "loss": 1.5627, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.795873999595642, + "eval_runtime": 49.9423, + "eval_samples_per_second": 20.023, + "eval_steps_per_second": 0.841, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 2.871537208557129, + "learning_rate": 2.764012997562957e-06, + "loss": 1.6027, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.6842691898345947, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.5327, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 2.6385788917541504, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.6181, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.6933207511901855, + "learning_rate": 2.757920389926889e-06, + "loss": 1.5561, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 2.814086437225342, + "learning_rate": 2.755889520714866e-06, + "loss": 1.5519, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.6112890243530273, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.52, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.7777466773986816, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.5885, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.782341241836548, + "learning_rate": 2.749796913078798e-06, + "loss": 1.609, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.576822519302368, + "learning_rate": 2.747766043866775e-06, + "loss": 1.5808, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.6026647090911865, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.626, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.7908697128295898, + "eval_runtime": 47.8943, + "eval_samples_per_second": 20.879, + "eval_steps_per_second": 0.877, + "step": 1400 + }, + { + "epoch": 1.5353424657534247, + "grad_norm": 2.645002603530884, + "learning_rate": 2.7437043054427297e-06, + "loss": 1.6664, + "step": 1401 + }, + { + "epoch": 1.5364383561643835, + "grad_norm": 2.676917552947998, + "learning_rate": 2.741673436230707e-06, + "loss": 1.572, + "step": 1402 + }, + { + "epoch": 1.5375342465753423, + "grad_norm": 2.6958320140838623, + "learning_rate": 2.739642567018684e-06, + "loss": 1.5712, + "step": 1403 + }, + { + "epoch": 1.5386301369863014, + "grad_norm": 2.8492650985717773, + "learning_rate": 2.7376116978066615e-06, + "loss": 1.5796, + "step": 1404 + }, + { + "epoch": 1.5397260273972604, + "grad_norm": 2.7714121341705322, + "learning_rate": 2.7355808285946388e-06, + "loss": 1.568, + "step": 1405 + }, + { + "epoch": 1.5408219178082192, + "grad_norm": 2.6548564434051514, + "learning_rate": 2.733549959382616e-06, + "loss": 1.5754, + "step": 1406 + }, + { + "epoch": 1.541917808219178, + "grad_norm": 2.6842844486236572, + "learning_rate": 2.7315190901705933e-06, + "loss": 1.6325, + "step": 1407 + }, + { + "epoch": 1.543013698630137, + "grad_norm": 2.700601816177368, + "learning_rate": 2.7294882209585706e-06, + "loss": 1.5986, + "step": 1408 + }, + { + "epoch": 1.5441095890410959, + "grad_norm": 2.5768840312957764, + "learning_rate": 2.727457351746548e-06, + "loss": 1.5543, + "step": 1409 + }, + { + "epoch": 1.5452054794520547, + "grad_norm": 2.633925199508667, + "learning_rate": 2.725426482534525e-06, + "loss": 1.5324, + "step": 1410 + }, + { + "epoch": 1.5452054794520547, + "eval_loss": 1.793561577796936, + "eval_runtime": 44.8327, + "eval_samples_per_second": 22.305, + "eval_steps_per_second": 0.937, + "step": 1410 + }, + { + "epoch": 1.5463013698630137, + "grad_norm": 2.5587072372436523, + "learning_rate": 2.723395613322502e-06, + "loss": 1.5725, + "step": 1411 + }, + { + "epoch": 1.5473972602739727, + "grad_norm": 2.8127360343933105, + "learning_rate": 2.7213647441104797e-06, + "loss": 1.5697, + "step": 1412 + }, + { + "epoch": 1.5484931506849315, + "grad_norm": 2.649578332901001, + "learning_rate": 2.7193338748984565e-06, + "loss": 1.5611, + "step": 1413 + }, + { + "epoch": 1.5495890410958904, + "grad_norm": 2.792600393295288, + "learning_rate": 2.7173030056864342e-06, + "loss": 1.7033, + "step": 1414 + }, + { + "epoch": 1.5506849315068494, + "grad_norm": 2.945256233215332, + "learning_rate": 2.715272136474411e-06, + "loss": 1.5325, + "step": 1415 + }, + { + "epoch": 1.5517808219178082, + "grad_norm": 2.815673589706421, + "learning_rate": 2.713241267262389e-06, + "loss": 1.5764, + "step": 1416 + }, + { + "epoch": 1.552876712328767, + "grad_norm": 2.5527024269104004, + "learning_rate": 2.7112103980503656e-06, + "loss": 1.5647, + "step": 1417 + }, + { + "epoch": 1.553972602739726, + "grad_norm": 2.725724697113037, + "learning_rate": 2.7091795288383433e-06, + "loss": 1.5979, + "step": 1418 + }, + { + "epoch": 1.555068493150685, + "grad_norm": 2.7697906494140625, + "learning_rate": 2.70714865962632e-06, + "loss": 1.5521, + "step": 1419 + }, + { + "epoch": 1.5561643835616439, + "grad_norm": 2.5871148109436035, + "learning_rate": 2.705117790414298e-06, + "loss": 1.5812, + "step": 1420 + }, + { + "epoch": 1.5561643835616439, + "eval_loss": 1.798141360282898, + "eval_runtime": 51.1729, + "eval_samples_per_second": 19.542, + "eval_steps_per_second": 0.821, + "step": 1420 + }, + { + "epoch": 1.5572602739726027, + "grad_norm": 2.5430808067321777, + "learning_rate": 2.7030869212022747e-06, + "loss": 1.6085, + "step": 1421 + }, + { + "epoch": 1.5583561643835617, + "grad_norm": 2.878905773162842, + "learning_rate": 2.7010560519902516e-06, + "loss": 1.5544, + "step": 1422 + }, + { + "epoch": 1.5594520547945205, + "grad_norm": 2.7553086280822754, + "learning_rate": 2.6990251827782293e-06, + "loss": 1.5712, + "step": 1423 + }, + { + "epoch": 1.5605479452054793, + "grad_norm": 2.636136054992676, + "learning_rate": 2.696994313566206e-06, + "loss": 1.5574, + "step": 1424 + }, + { + "epoch": 1.5616438356164384, + "grad_norm": 2.8743226528167725, + "learning_rate": 2.694963444354184e-06, + "loss": 1.6279, + "step": 1425 + }, + { + "epoch": 1.5627397260273974, + "grad_norm": 2.7313430309295654, + "learning_rate": 2.6929325751421607e-06, + "loss": 1.5717, + "step": 1426 + }, + { + "epoch": 1.563835616438356, + "grad_norm": 2.6373813152313232, + "learning_rate": 2.6909017059301384e-06, + "loss": 1.5935, + "step": 1427 + }, + { + "epoch": 1.564931506849315, + "grad_norm": 2.7440435886383057, + "learning_rate": 2.6888708367181153e-06, + "loss": 1.5402, + "step": 1428 + }, + { + "epoch": 1.566027397260274, + "grad_norm": 2.5901665687561035, + "learning_rate": 2.686839967506093e-06, + "loss": 1.5946, + "step": 1429 + }, + { + "epoch": 1.5671232876712329, + "grad_norm": 2.680370807647705, + "learning_rate": 2.68480909829407e-06, + "loss": 1.5068, + "step": 1430 + }, + { + "epoch": 1.5671232876712329, + "eval_loss": 1.7998759746551514, + "eval_runtime": 49.5336, + "eval_samples_per_second": 20.188, + "eval_steps_per_second": 0.848, + "step": 1430 + }, + { + "epoch": 1.5682191780821917, + "grad_norm": 2.6154282093048096, + "learning_rate": 2.6827782290820475e-06, + "loss": 1.4984, + "step": 1431 + }, + { + "epoch": 1.5693150684931507, + "grad_norm": 2.981612205505371, + "learning_rate": 2.6807473598700244e-06, + "loss": 1.6224, + "step": 1432 + }, + { + "epoch": 1.5704109589041098, + "grad_norm": 3.0227465629577637, + "learning_rate": 2.678716490658002e-06, + "loss": 1.5942, + "step": 1433 + }, + { + "epoch": 1.5715068493150683, + "grad_norm": 2.803877115249634, + "learning_rate": 2.676685621445979e-06, + "loss": 1.5821, + "step": 1434 + }, + { + "epoch": 1.5726027397260274, + "grad_norm": 2.7260076999664307, + "learning_rate": 2.674654752233956e-06, + "loss": 1.5788, + "step": 1435 + }, + { + "epoch": 1.5736986301369864, + "grad_norm": 2.672656297683716, + "learning_rate": 2.6726238830219335e-06, + "loss": 1.5798, + "step": 1436 + }, + { + "epoch": 1.5747945205479452, + "grad_norm": 2.7088775634765625, + "learning_rate": 2.6705930138099107e-06, + "loss": 1.6142, + "step": 1437 + }, + { + "epoch": 1.575890410958904, + "grad_norm": 2.5411477088928223, + "learning_rate": 2.668562144597888e-06, + "loss": 1.4999, + "step": 1438 + }, + { + "epoch": 1.576986301369863, + "grad_norm": 2.5658116340637207, + "learning_rate": 2.6665312753858653e-06, + "loss": 1.5898, + "step": 1439 + }, + { + "epoch": 1.5780821917808219, + "grad_norm": 2.626997232437134, + "learning_rate": 2.6645004061738426e-06, + "loss": 1.5795, + "step": 1440 + }, + { + "epoch": 1.5780821917808219, + "eval_loss": 1.7931181192398071, + "eval_runtime": 49.7844, + "eval_samples_per_second": 20.087, + "eval_steps_per_second": 0.844, + "step": 1440 + }, + { + "epoch": 1.5791780821917807, + "grad_norm": 2.737011194229126, + "learning_rate": 2.66246953696182e-06, + "loss": 1.5275, + "step": 1441 + }, + { + "epoch": 1.5802739726027397, + "grad_norm": 2.6428184509277344, + "learning_rate": 2.660438667749797e-06, + "loss": 1.5429, + "step": 1442 + }, + { + "epoch": 1.5813698630136988, + "grad_norm": 2.6428184509277344, + "learning_rate": 2.660438667749797e-06, + "loss": 1.5771, + "step": 1443 + }, + { + "epoch": 1.5824657534246576, + "grad_norm": 3.038092613220215, + "learning_rate": 2.6584077985377744e-06, + "loss": 1.579, + "step": 1444 + }, + { + "epoch": 1.5835616438356164, + "grad_norm": 2.4511592388153076, + "learning_rate": 2.6563769293257517e-06, + "loss": 1.5409, + "step": 1445 + }, + { + "epoch": 1.5846575342465754, + "grad_norm": 2.576601505279541, + "learning_rate": 2.654346060113729e-06, + "loss": 1.5411, + "step": 1446 + }, + { + "epoch": 1.5857534246575342, + "grad_norm": 2.54603910446167, + "learning_rate": 2.6523151909017062e-06, + "loss": 1.5455, + "step": 1447 + }, + { + "epoch": 1.586849315068493, + "grad_norm": 2.6850075721740723, + "learning_rate": 2.6502843216896835e-06, + "loss": 1.6355, + "step": 1448 + }, + { + "epoch": 1.587945205479452, + "grad_norm": 2.625603675842285, + "learning_rate": 2.6482534524776603e-06, + "loss": 1.5954, + "step": 1449 + }, + { + "epoch": 1.589041095890411, + "grad_norm": 2.755312919616699, + "learning_rate": 2.646222583265638e-06, + "loss": 1.5867, + "step": 1450 + }, + { + "epoch": 1.589041095890411, + "eval_loss": 1.7949016094207764, + "eval_runtime": 48.7453, + "eval_samples_per_second": 20.515, + "eval_steps_per_second": 0.862, + "step": 1450 + }, + { + "epoch": 1.59013698630137, + "grad_norm": 2.6282010078430176, + "learning_rate": 2.644191714053615e-06, + "loss": 1.5539, + "step": 1451 + }, + { + "epoch": 1.5912328767123287, + "grad_norm": 2.5556142330169678, + "learning_rate": 2.6421608448415926e-06, + "loss": 1.5324, + "step": 1452 + }, + { + "epoch": 1.5923287671232877, + "grad_norm": 2.9253478050231934, + "learning_rate": 2.6401299756295694e-06, + "loss": 1.58, + "step": 1453 + }, + { + "epoch": 1.5934246575342466, + "grad_norm": 2.8228557109832764, + "learning_rate": 2.638099106417547e-06, + "loss": 1.5671, + "step": 1454 + }, + { + "epoch": 1.5945205479452054, + "grad_norm": 2.7850286960601807, + "learning_rate": 2.636068237205524e-06, + "loss": 1.6302, + "step": 1455 + }, + { + "epoch": 1.5956164383561644, + "grad_norm": 2.5926599502563477, + "learning_rate": 2.6340373679935017e-06, + "loss": 1.4949, + "step": 1456 + }, + { + "epoch": 1.5967123287671234, + "grad_norm": 2.7452011108398438, + "learning_rate": 2.6320064987814785e-06, + "loss": 1.5828, + "step": 1457 + }, + { + "epoch": 1.5978082191780822, + "grad_norm": 2.626157283782959, + "learning_rate": 2.6299756295694562e-06, + "loss": 1.6194, + "step": 1458 + }, + { + "epoch": 1.598904109589041, + "grad_norm": 2.7487716674804688, + "learning_rate": 2.627944760357433e-06, + "loss": 1.635, + "step": 1459 + }, + { + "epoch": 1.6, + "grad_norm": 2.738173723220825, + "learning_rate": 2.6259138911454104e-06, + "loss": 1.615, + "step": 1460 + }, + { + "epoch": 1.6, + "eval_loss": 1.789223313331604, + "eval_runtime": 45.8318, + "eval_samples_per_second": 21.819, + "eval_steps_per_second": 0.916, + "step": 1460 + }, + { + "epoch": 1.601095890410959, + "grad_norm": 2.535860061645508, + "learning_rate": 2.6238830219333877e-06, + "loss": 1.5307, + "step": 1461 + }, + { + "epoch": 1.6021917808219177, + "grad_norm": 2.745260000228882, + "learning_rate": 2.621852152721365e-06, + "loss": 1.5883, + "step": 1462 + }, + { + "epoch": 1.6032876712328767, + "grad_norm": 2.7904720306396484, + "learning_rate": 2.619821283509342e-06, + "loss": 1.5309, + "step": 1463 + }, + { + "epoch": 1.6043835616438358, + "grad_norm": 2.70133900642395, + "learning_rate": 2.6177904142973195e-06, + "loss": 1.5906, + "step": 1464 + }, + { + "epoch": 1.6054794520547944, + "grad_norm": 2.61118745803833, + "learning_rate": 2.6157595450852968e-06, + "loss": 1.6168, + "step": 1465 + }, + { + "epoch": 1.6065753424657534, + "grad_norm": 2.841352939605713, + "learning_rate": 2.613728675873274e-06, + "loss": 1.6228, + "step": 1466 + }, + { + "epoch": 1.6076712328767124, + "grad_norm": 2.5176472663879395, + "learning_rate": 2.6116978066612513e-06, + "loss": 1.5735, + "step": 1467 + }, + { + "epoch": 1.6087671232876712, + "grad_norm": 2.6179540157318115, + "learning_rate": 2.6096669374492286e-06, + "loss": 1.5506, + "step": 1468 + }, + { + "epoch": 1.60986301369863, + "grad_norm": 2.59338116645813, + "learning_rate": 2.607636068237206e-06, + "loss": 1.5964, + "step": 1469 + }, + { + "epoch": 1.610958904109589, + "grad_norm": 2.569547414779663, + "learning_rate": 2.605605199025183e-06, + "loss": 1.5716, + "step": 1470 + }, + { + "epoch": 1.610958904109589, + "eval_loss": 1.7863329648971558, + "eval_runtime": 43.4508, + "eval_samples_per_second": 23.015, + "eval_steps_per_second": 0.967, + "step": 1470 + }, + { + "epoch": 1.612054794520548, + "grad_norm": 2.603998899459839, + "learning_rate": 2.6035743298131604e-06, + "loss": 1.6011, + "step": 1471 + }, + { + "epoch": 1.6131506849315067, + "grad_norm": 2.7348268032073975, + "learning_rate": 2.6015434606011377e-06, + "loss": 1.6082, + "step": 1472 + }, + { + "epoch": 1.6142465753424657, + "grad_norm": 2.5288398265838623, + "learning_rate": 2.5995125913891145e-06, + "loss": 1.5756, + "step": 1473 + }, + { + "epoch": 1.6153424657534248, + "grad_norm": 2.914170742034912, + "learning_rate": 2.5974817221770922e-06, + "loss": 1.5961, + "step": 1474 + }, + { + "epoch": 1.6164383561643836, + "grad_norm": 2.7135908603668213, + "learning_rate": 2.595450852965069e-06, + "loss": 1.6216, + "step": 1475 + }, + { + "epoch": 1.6175342465753424, + "grad_norm": 2.5617594718933105, + "learning_rate": 2.593419983753047e-06, + "loss": 1.5639, + "step": 1476 + }, + { + "epoch": 1.6186301369863014, + "grad_norm": 2.931432008743286, + "learning_rate": 2.5913891145410236e-06, + "loss": 1.6058, + "step": 1477 + }, + { + "epoch": 1.6197260273972602, + "grad_norm": 2.577707052230835, + "learning_rate": 2.5893582453290013e-06, + "loss": 1.5524, + "step": 1478 + }, + { + "epoch": 1.620821917808219, + "grad_norm": 2.5421078205108643, + "learning_rate": 2.587327376116978e-06, + "loss": 1.5715, + "step": 1479 + }, + { + "epoch": 1.621917808219178, + "grad_norm": 2.6675302982330322, + "learning_rate": 2.585296506904956e-06, + "loss": 1.5514, + "step": 1480 + }, + { + "epoch": 1.621917808219178, + "eval_loss": 1.7871757745742798, + "eval_runtime": 43.254, + "eval_samples_per_second": 23.119, + "eval_steps_per_second": 0.971, + "step": 1480 + }, + { + "epoch": 1.6230136986301371, + "grad_norm": 2.6179966926574707, + "learning_rate": 2.5832656376929327e-06, + "loss": 1.5026, + "step": 1481 + }, + { + "epoch": 1.624109589041096, + "grad_norm": 2.8699069023132324, + "learning_rate": 2.5812347684809104e-06, + "loss": 1.6569, + "step": 1482 + }, + { + "epoch": 1.6252054794520547, + "grad_norm": 2.5897462368011475, + "learning_rate": 2.5792038992688873e-06, + "loss": 1.4955, + "step": 1483 + }, + { + "epoch": 1.6263013698630138, + "grad_norm": 2.618124485015869, + "learning_rate": 2.577173030056865e-06, + "loss": 1.5972, + "step": 1484 + }, + { + "epoch": 1.6273972602739726, + "grad_norm": 2.6729276180267334, + "learning_rate": 2.575142160844842e-06, + "loss": 1.5994, + "step": 1485 + }, + { + "epoch": 1.6284931506849314, + "grad_norm": 2.52236008644104, + "learning_rate": 2.5731112916328187e-06, + "loss": 1.536, + "step": 1486 + }, + { + "epoch": 1.6295890410958904, + "grad_norm": 2.5696699619293213, + "learning_rate": 2.5710804224207964e-06, + "loss": 1.5055, + "step": 1487 + }, + { + "epoch": 1.6306849315068495, + "grad_norm": 2.6239490509033203, + "learning_rate": 2.5690495532087733e-06, + "loss": 1.579, + "step": 1488 + }, + { + "epoch": 1.6317808219178083, + "grad_norm": 2.6407887935638428, + "learning_rate": 2.567018683996751e-06, + "loss": 1.541, + "step": 1489 + }, + { + "epoch": 1.632876712328767, + "grad_norm": 2.747039556503296, + "learning_rate": 2.564987814784728e-06, + "loss": 1.5592, + "step": 1490 + }, + { + "epoch": 1.632876712328767, + "eval_loss": 1.7917380332946777, + "eval_runtime": 43.2553, + "eval_samples_per_second": 23.119, + "eval_steps_per_second": 0.971, + "step": 1490 + }, + { + "epoch": 1.633972602739726, + "grad_norm": 2.574068069458008, + "learning_rate": 2.5629569455727055e-06, + "loss": 1.5694, + "step": 1491 + }, + { + "epoch": 1.635068493150685, + "grad_norm": 2.7107486724853516, + "learning_rate": 2.5609260763606824e-06, + "loss": 1.6631, + "step": 1492 + }, + { + "epoch": 1.6361643835616437, + "grad_norm": 2.603778839111328, + "learning_rate": 2.55889520714866e-06, + "loss": 1.5579, + "step": 1493 + }, + { + "epoch": 1.6372602739726028, + "grad_norm": 2.770524740219116, + "learning_rate": 2.556864337936637e-06, + "loss": 1.5619, + "step": 1494 + }, + { + "epoch": 1.6383561643835618, + "grad_norm": 2.6948940753936768, + "learning_rate": 2.5548334687246146e-06, + "loss": 1.6224, + "step": 1495 + }, + { + "epoch": 1.6394520547945204, + "grad_norm": 2.6838672161102295, + "learning_rate": 2.5528025995125915e-06, + "loss": 1.5727, + "step": 1496 + }, + { + "epoch": 1.6405479452054794, + "grad_norm": 2.6357474327087402, + "learning_rate": 2.550771730300569e-06, + "loss": 1.5985, + "step": 1497 + }, + { + "epoch": 1.6416438356164385, + "grad_norm": 2.8199994564056396, + "learning_rate": 2.548740861088546e-06, + "loss": 1.5065, + "step": 1498 + }, + { + "epoch": 1.6427397260273973, + "grad_norm": 2.595996141433716, + "learning_rate": 2.5467099918765233e-06, + "loss": 1.4976, + "step": 1499 + }, + { + "epoch": 1.643835616438356, + "grad_norm": 2.9895756244659424, + "learning_rate": 2.5446791226645006e-06, + "loss": 1.6085, + "step": 1500 + }, + { + "epoch": 1.643835616438356, + "eval_loss": 1.7889667749404907, + "eval_runtime": 42.5759, + "eval_samples_per_second": 23.487, + "eval_steps_per_second": 0.986, + "step": 1500 + }, + { + "epoch": 1.644931506849315, + "grad_norm": 2.686858654022217, + "learning_rate": 2.542648253452478e-06, + "loss": 1.5811, + "step": 1501 + }, + { + "epoch": 1.6460273972602741, + "grad_norm": 2.674805164337158, + "learning_rate": 2.540617384240455e-06, + "loss": 1.5365, + "step": 1502 + }, + { + "epoch": 1.6471232876712327, + "grad_norm": 2.6717612743377686, + "learning_rate": 2.5385865150284324e-06, + "loss": 1.558, + "step": 1503 + }, + { + "epoch": 1.6482191780821918, + "grad_norm": 2.8866565227508545, + "learning_rate": 2.5365556458164097e-06, + "loss": 1.5649, + "step": 1504 + }, + { + "epoch": 1.6493150684931508, + "grad_norm": 2.622196912765503, + "learning_rate": 2.534524776604387e-06, + "loss": 1.4744, + "step": 1505 + }, + { + "epoch": 1.6504109589041096, + "grad_norm": 2.6443939208984375, + "learning_rate": 2.5324939073923642e-06, + "loss": 1.5541, + "step": 1506 + }, + { + "epoch": 1.6515068493150684, + "grad_norm": 2.7976269721984863, + "learning_rate": 2.5304630381803415e-06, + "loss": 1.5654, + "step": 1507 + }, + { + "epoch": 1.6526027397260274, + "grad_norm": 2.8624062538146973, + "learning_rate": 2.5284321689683188e-06, + "loss": 1.5672, + "step": 1508 + }, + { + "epoch": 1.6536986301369863, + "grad_norm": 2.75897479057312, + "learning_rate": 2.526401299756296e-06, + "loss": 1.5536, + "step": 1509 + }, + { + "epoch": 1.654794520547945, + "grad_norm": 2.671563148498535, + "learning_rate": 2.524370430544273e-06, + "loss": 1.5327, + "step": 1510 + }, + { + "epoch": 1.654794520547945, + "eval_loss": 1.789809226989746, + "eval_runtime": 42.1012, + "eval_samples_per_second": 23.752, + "eval_steps_per_second": 0.998, + "step": 1510 + }, + { + "epoch": 1.655890410958904, + "grad_norm": 3.337616205215454, + "learning_rate": 2.5223395613322506e-06, + "loss": 1.5794, + "step": 1511 + }, + { + "epoch": 1.6569863013698631, + "grad_norm": 2.7513420581817627, + "learning_rate": 2.5203086921202274e-06, + "loss": 1.5561, + "step": 1512 + }, + { + "epoch": 1.658082191780822, + "grad_norm": 2.9580318927764893, + "learning_rate": 2.518277822908205e-06, + "loss": 1.572, + "step": 1513 + }, + { + "epoch": 1.6591780821917808, + "grad_norm": 3.01985764503479, + "learning_rate": 2.516246953696182e-06, + "loss": 1.585, + "step": 1514 + }, + { + "epoch": 1.6602739726027398, + "grad_norm": 2.762683629989624, + "learning_rate": 2.5142160844841597e-06, + "loss": 1.5817, + "step": 1515 + }, + { + "epoch": 1.6613698630136986, + "grad_norm": 2.6364688873291016, + "learning_rate": 2.5121852152721365e-06, + "loss": 1.5848, + "step": 1516 + }, + { + "epoch": 1.6624657534246574, + "grad_norm": 2.559342861175537, + "learning_rate": 2.5101543460601142e-06, + "loss": 1.512, + "step": 1517 + }, + { + "epoch": 1.6635616438356164, + "grad_norm": 2.657026529312134, + "learning_rate": 2.508123476848091e-06, + "loss": 1.6042, + "step": 1518 + }, + { + "epoch": 1.6646575342465755, + "grad_norm": 2.721721649169922, + "learning_rate": 2.506092607636069e-06, + "loss": 1.5585, + "step": 1519 + }, + { + "epoch": 1.6657534246575343, + "grad_norm": 2.5659472942352295, + "learning_rate": 2.5040617384240457e-06, + "loss": 1.5456, + "step": 1520 + }, + { + "epoch": 1.6657534246575343, + "eval_loss": 1.7858935594558716, + "eval_runtime": 43.2977, + "eval_samples_per_second": 23.096, + "eval_steps_per_second": 0.97, + "step": 1520 + }, + { + "epoch": 1.666849315068493, + "grad_norm": 2.689267158508301, + "learning_rate": 2.5020308692120234e-06, + "loss": 1.616, + "step": 1521 + }, + { + "epoch": 1.6679452054794521, + "grad_norm": 2.7258920669555664, + "learning_rate": 2.5e-06, + "loss": 1.5499, + "step": 1522 + }, + { + "epoch": 1.669041095890411, + "grad_norm": 2.9586708545684814, + "learning_rate": 2.4979691307879775e-06, + "loss": 1.5476, + "step": 1523 + }, + { + "epoch": 1.6701369863013698, + "grad_norm": 2.8209455013275146, + "learning_rate": 2.4959382615759548e-06, + "loss": 1.6178, + "step": 1524 + }, + { + "epoch": 1.6712328767123288, + "grad_norm": 2.6040000915527344, + "learning_rate": 2.493907392363932e-06, + "loss": 1.5612, + "step": 1525 + }, + { + "epoch": 1.6723287671232878, + "grad_norm": 2.551750659942627, + "learning_rate": 2.4918765231519093e-06, + "loss": 1.5724, + "step": 1526 + }, + { + "epoch": 1.6734246575342466, + "grad_norm": 2.681349992752075, + "learning_rate": 2.4898456539398866e-06, + "loss": 1.6296, + "step": 1527 + }, + { + "epoch": 1.6745205479452054, + "grad_norm": 2.674236536026001, + "learning_rate": 2.487814784727864e-06, + "loss": 1.5584, + "step": 1528 + }, + { + "epoch": 1.6756164383561645, + "grad_norm": 2.648228406906128, + "learning_rate": 2.4857839155158407e-06, + "loss": 1.5591, + "step": 1529 + }, + { + "epoch": 1.6767123287671233, + "grad_norm": 2.62040114402771, + "learning_rate": 2.483753046303818e-06, + "loss": 1.5627, + "step": 1530 + }, + { + "epoch": 1.6767123287671233, + "eval_loss": 1.7849520444869995, + "eval_runtime": 43.6712, + "eval_samples_per_second": 22.898, + "eval_steps_per_second": 0.962, + "step": 1530 + }, + { + "epoch": 1.677808219178082, + "grad_norm": 2.7421514987945557, + "learning_rate": 2.4817221770917953e-06, + "loss": 1.5925, + "step": 1531 + }, + { + "epoch": 1.6789041095890411, + "grad_norm": 3.0015368461608887, + "learning_rate": 2.4796913078797725e-06, + "loss": 1.5954, + "step": 1532 + }, + { + "epoch": 1.6800000000000002, + "grad_norm": 2.5985584259033203, + "learning_rate": 2.47766043866775e-06, + "loss": 1.5948, + "step": 1533 + }, + { + "epoch": 1.6810958904109587, + "grad_norm": 2.55576229095459, + "learning_rate": 2.475629569455727e-06, + "loss": 1.5397, + "step": 1534 + }, + { + "epoch": 1.6821917808219178, + "grad_norm": 2.4863014221191406, + "learning_rate": 2.4735987002437044e-06, + "loss": 1.5366, + "step": 1535 + }, + { + "epoch": 1.6832876712328768, + "grad_norm": 2.5521163940429688, + "learning_rate": 2.4715678310316816e-06, + "loss": 1.5527, + "step": 1536 + }, + { + "epoch": 1.6843835616438356, + "grad_norm": 2.6156206130981445, + "learning_rate": 2.469536961819659e-06, + "loss": 1.5657, + "step": 1537 + }, + { + "epoch": 1.6854794520547944, + "grad_norm": 3.497575044631958, + "learning_rate": 2.467506092607636e-06, + "loss": 1.5461, + "step": 1538 + }, + { + "epoch": 1.6865753424657535, + "grad_norm": 2.9800186157226562, + "learning_rate": 2.4654752233956135e-06, + "loss": 1.6193, + "step": 1539 + }, + { + "epoch": 1.6876712328767123, + "grad_norm": 2.5514957904815674, + "learning_rate": 2.4634443541835907e-06, + "loss": 1.5715, + "step": 1540 + }, + { + "epoch": 1.6876712328767123, + "eval_loss": 1.7836337089538574, + "eval_runtime": 45.9798, + "eval_samples_per_second": 21.749, + "eval_steps_per_second": 0.913, + "step": 1540 + }, + { + "epoch": 1.688767123287671, + "grad_norm": 2.6997995376586914, + "learning_rate": 2.461413484971568e-06, + "loss": 1.6398, + "step": 1541 + }, + { + "epoch": 1.6898630136986301, + "grad_norm": 2.8419151306152344, + "learning_rate": 2.4593826157595453e-06, + "loss": 1.5212, + "step": 1542 + }, + { + "epoch": 1.6909589041095892, + "grad_norm": 3.348747968673706, + "learning_rate": 2.4573517465475226e-06, + "loss": 1.5462, + "step": 1543 + }, + { + "epoch": 1.692054794520548, + "grad_norm": 2.4968204498291016, + "learning_rate": 2.4553208773355e-06, + "loss": 1.5652, + "step": 1544 + }, + { + "epoch": 1.6931506849315068, + "grad_norm": 2.534217357635498, + "learning_rate": 2.453290008123477e-06, + "loss": 1.5908, + "step": 1545 + }, + { + "epoch": 1.6942465753424658, + "grad_norm": 2.585479974746704, + "learning_rate": 2.4512591389114544e-06, + "loss": 1.5888, + "step": 1546 + }, + { + "epoch": 1.6953424657534246, + "grad_norm": 2.7593765258789062, + "learning_rate": 2.4492282696994317e-06, + "loss": 1.5538, + "step": 1547 + }, + { + "epoch": 1.6964383561643834, + "grad_norm": 2.7336785793304443, + "learning_rate": 2.447197400487409e-06, + "loss": 1.5795, + "step": 1548 + }, + { + "epoch": 1.6975342465753425, + "grad_norm": 3.5008962154388428, + "learning_rate": 2.4451665312753862e-06, + "loss": 1.5996, + "step": 1549 + }, + { + "epoch": 1.6986301369863015, + "grad_norm": 2.5645105838775635, + "learning_rate": 2.4431356620633635e-06, + "loss": 1.5865, + "step": 1550 + }, + { + "epoch": 1.6986301369863015, + "eval_loss": 1.7836376428604126, + "eval_runtime": 45.7525, + "eval_samples_per_second": 21.857, + "eval_steps_per_second": 0.918, + "step": 1550 + }, + { + "epoch": 1.6997260273972603, + "grad_norm": 2.62166690826416, + "learning_rate": 2.4411047928513408e-06, + "loss": 1.5788, + "step": 1551 + }, + { + "epoch": 1.7008219178082191, + "grad_norm": 2.7162153720855713, + "learning_rate": 2.439073923639318e-06, + "loss": 1.5215, + "step": 1552 + }, + { + "epoch": 1.7019178082191782, + "grad_norm": 2.7269327640533447, + "learning_rate": 2.4370430544272953e-06, + "loss": 1.5109, + "step": 1553 + }, + { + "epoch": 1.703013698630137, + "grad_norm": 2.9356398582458496, + "learning_rate": 2.435012185215272e-06, + "loss": 1.5491, + "step": 1554 + }, + { + "epoch": 1.7041095890410958, + "grad_norm": 2.738612174987793, + "learning_rate": 2.4329813160032495e-06, + "loss": 1.5703, + "step": 1555 + }, + { + "epoch": 1.7052054794520548, + "grad_norm": 2.671029567718506, + "learning_rate": 2.4309504467912267e-06, + "loss": 1.5546, + "step": 1556 + }, + { + "epoch": 1.7063013698630138, + "grad_norm": 2.821136713027954, + "learning_rate": 2.428919577579204e-06, + "loss": 1.6421, + "step": 1557 + }, + { + "epoch": 1.7073972602739727, + "grad_norm": 2.562530517578125, + "learning_rate": 2.4268887083671813e-06, + "loss": 1.5436, + "step": 1558 + }, + { + "epoch": 1.7084931506849315, + "grad_norm": 2.560969114303589, + "learning_rate": 2.4248578391551586e-06, + "loss": 1.5375, + "step": 1559 + }, + { + "epoch": 1.7095890410958905, + "grad_norm": 2.6686229705810547, + "learning_rate": 2.422826969943136e-06, + "loss": 1.5184, + "step": 1560 + }, + { + "epoch": 1.7095890410958905, + "eval_loss": 1.7841863632202148, + "eval_runtime": 44.8027, + "eval_samples_per_second": 22.32, + "eval_steps_per_second": 0.937, + "step": 1560 + }, + { + "epoch": 1.7106849315068493, + "grad_norm": 2.5086288452148438, + "learning_rate": 2.420796100731113e-06, + "loss": 1.5809, + "step": 1561 + }, + { + "epoch": 1.7117808219178081, + "grad_norm": 2.536912679672241, + "learning_rate": 2.4187652315190904e-06, + "loss": 1.5822, + "step": 1562 + }, + { + "epoch": 1.7128767123287671, + "grad_norm": 2.660067081451416, + "learning_rate": 2.4167343623070677e-06, + "loss": 1.5535, + "step": 1563 + }, + { + "epoch": 1.7139726027397262, + "grad_norm": 2.8859071731567383, + "learning_rate": 2.414703493095045e-06, + "loss": 1.5636, + "step": 1564 + }, + { + "epoch": 1.7150684931506848, + "grad_norm": 2.773730993270874, + "learning_rate": 2.4126726238830222e-06, + "loss": 1.5466, + "step": 1565 + }, + { + "epoch": 1.7161643835616438, + "grad_norm": 2.7574939727783203, + "learning_rate": 2.410641754670999e-06, + "loss": 1.5978, + "step": 1566 + }, + { + "epoch": 1.7172602739726028, + "grad_norm": 2.7123234272003174, + "learning_rate": 2.4086108854589763e-06, + "loss": 1.551, + "step": 1567 + }, + { + "epoch": 1.7183561643835616, + "grad_norm": 2.726815938949585, + "learning_rate": 2.4065800162469536e-06, + "loss": 1.5627, + "step": 1568 + }, + { + "epoch": 1.7194520547945205, + "grad_norm": 2.7364342212677, + "learning_rate": 2.404549147034931e-06, + "loss": 1.5967, + "step": 1569 + }, + { + "epoch": 1.7205479452054795, + "grad_norm": 2.6859424114227295, + "learning_rate": 2.402518277822908e-06, + "loss": 1.5347, + "step": 1570 + }, + { + "epoch": 1.7205479452054795, + "eval_loss": 1.7833315134048462, + "eval_runtime": 45.1817, + "eval_samples_per_second": 22.133, + "eval_steps_per_second": 0.93, + "step": 1570 + }, + { + "epoch": 1.7216438356164385, + "grad_norm": 2.7983601093292236, + "learning_rate": 2.4004874086108854e-06, + "loss": 1.5828, + "step": 1571 + }, + { + "epoch": 1.7227397260273971, + "grad_norm": 2.62542462348938, + "learning_rate": 2.3984565393988627e-06, + "loss": 1.5417, + "step": 1572 + }, + { + "epoch": 1.7238356164383561, + "grad_norm": 2.8035542964935303, + "learning_rate": 2.39642567018684e-06, + "loss": 1.5539, + "step": 1573 + }, + { + "epoch": 1.7249315068493152, + "grad_norm": 2.8192529678344727, + "learning_rate": 2.3943948009748173e-06, + "loss": 1.5824, + "step": 1574 + }, + { + "epoch": 1.726027397260274, + "grad_norm": 2.63417387008667, + "learning_rate": 2.3923639317627945e-06, + "loss": 1.5296, + "step": 1575 + }, + { + "epoch": 1.7271232876712328, + "grad_norm": 2.596536874771118, + "learning_rate": 2.390333062550772e-06, + "loss": 1.4545, + "step": 1576 + }, + { + "epoch": 1.7282191780821918, + "grad_norm": 2.8503546714782715, + "learning_rate": 2.388302193338749e-06, + "loss": 1.6365, + "step": 1577 + }, + { + "epoch": 1.7293150684931506, + "grad_norm": 2.6589956283569336, + "learning_rate": 2.3862713241267264e-06, + "loss": 1.566, + "step": 1578 + }, + { + "epoch": 1.7304109589041095, + "grad_norm": 2.728832721710205, + "learning_rate": 2.3842404549147037e-06, + "loss": 1.5242, + "step": 1579 + }, + { + "epoch": 1.7315068493150685, + "grad_norm": 2.7927703857421875, + "learning_rate": 2.382209585702681e-06, + "loss": 1.6126, + "step": 1580 + }, + { + "epoch": 1.7315068493150685, + "eval_loss": 1.78191077709198, + "eval_runtime": 46.2407, + "eval_samples_per_second": 21.626, + "eval_steps_per_second": 0.908, + "step": 1580 + }, + { + "epoch": 1.7326027397260275, + "grad_norm": 2.800701379776001, + "learning_rate": 2.380178716490658e-06, + "loss": 1.5566, + "step": 1581 + }, + { + "epoch": 1.7336986301369863, + "grad_norm": 2.746800661087036, + "learning_rate": 2.3781478472786355e-06, + "loss": 1.5979, + "step": 1582 + }, + { + "epoch": 1.7347945205479451, + "grad_norm": 2.7532665729522705, + "learning_rate": 2.3761169780666128e-06, + "loss": 1.6007, + "step": 1583 + }, + { + "epoch": 1.7358904109589042, + "grad_norm": 2.7455027103424072, + "learning_rate": 2.37408610885459e-06, + "loss": 1.5478, + "step": 1584 + }, + { + "epoch": 1.736986301369863, + "grad_norm": 2.693115711212158, + "learning_rate": 2.3720552396425673e-06, + "loss": 1.6249, + "step": 1585 + }, + { + "epoch": 1.7380821917808218, + "grad_norm": 2.6274261474609375, + "learning_rate": 2.3700243704305446e-06, + "loss": 1.574, + "step": 1586 + }, + { + "epoch": 1.7391780821917808, + "grad_norm": 2.6467201709747314, + "learning_rate": 2.367993501218522e-06, + "loss": 1.5556, + "step": 1587 + }, + { + "epoch": 1.7402739726027399, + "grad_norm": 2.7470309734344482, + "learning_rate": 2.365962632006499e-06, + "loss": 1.5932, + "step": 1588 + }, + { + "epoch": 1.7413698630136987, + "grad_norm": 2.694892406463623, + "learning_rate": 2.3639317627944764e-06, + "loss": 1.5889, + "step": 1589 + }, + { + "epoch": 1.7424657534246575, + "grad_norm": 2.614192008972168, + "learning_rate": 2.3619008935824537e-06, + "loss": 1.5858, + "step": 1590 + }, + { + "epoch": 1.7424657534246575, + "eval_loss": 1.7806529998779297, + "eval_runtime": 47.8714, + "eval_samples_per_second": 20.889, + "eval_steps_per_second": 0.877, + "step": 1590 + }, + { + "epoch": 1.7435616438356165, + "grad_norm": 2.5324199199676514, + "learning_rate": 2.3598700243704305e-06, + "loss": 1.5872, + "step": 1591 + }, + { + "epoch": 1.7446575342465753, + "grad_norm": 2.6094815731048584, + "learning_rate": 2.357839155158408e-06, + "loss": 1.5729, + "step": 1592 + }, + { + "epoch": 1.7457534246575341, + "grad_norm": 2.702822208404541, + "learning_rate": 2.355808285946385e-06, + "loss": 1.5945, + "step": 1593 + }, + { + "epoch": 1.7468493150684932, + "grad_norm": 2.7433812618255615, + "learning_rate": 2.3537774167343624e-06, + "loss": 1.5772, + "step": 1594 + }, + { + "epoch": 1.7479452054794522, + "grad_norm": 2.6817004680633545, + "learning_rate": 2.3517465475223396e-06, + "loss": 1.5803, + "step": 1595 + }, + { + "epoch": 1.749041095890411, + "grad_norm": 2.8029403686523438, + "learning_rate": 2.349715678310317e-06, + "loss": 1.5449, + "step": 1596 + }, + { + "epoch": 1.7501369863013698, + "grad_norm": 2.82515287399292, + "learning_rate": 2.347684809098294e-06, + "loss": 1.5967, + "step": 1597 + }, + { + "epoch": 1.7512328767123289, + "grad_norm": 2.7368521690368652, + "learning_rate": 2.3456539398862715e-06, + "loss": 1.5325, + "step": 1598 + }, + { + "epoch": 1.7523287671232877, + "grad_norm": 2.6093783378601074, + "learning_rate": 2.3436230706742487e-06, + "loss": 1.5384, + "step": 1599 + }, + { + "epoch": 1.7534246575342465, + "grad_norm": 2.6561474800109863, + "learning_rate": 2.341592201462226e-06, + "loss": 1.5497, + "step": 1600 + }, + { + "epoch": 1.7534246575342465, + "eval_loss": 1.7796835899353027, + "eval_runtime": 44.7534, + "eval_samples_per_second": 22.345, + "eval_steps_per_second": 0.938, + "step": 1600 + }, + { + "epoch": 1.7545205479452055, + "grad_norm": 2.601712942123413, + "learning_rate": 2.3395613322502033e-06, + "loss": 1.5184, + "step": 1601 + }, + { + "epoch": 1.7556164383561645, + "grad_norm": 2.5231258869171143, + "learning_rate": 2.3375304630381806e-06, + "loss": 1.5392, + "step": 1602 + }, + { + "epoch": 1.7567123287671231, + "grad_norm": 2.6895570755004883, + "learning_rate": 2.335499593826158e-06, + "loss": 1.5912, + "step": 1603 + }, + { + "epoch": 1.7578082191780822, + "grad_norm": 2.751396417617798, + "learning_rate": 2.333468724614135e-06, + "loss": 1.5943, + "step": 1604 + }, + { + "epoch": 1.7589041095890412, + "grad_norm": 2.7620153427124023, + "learning_rate": 2.3314378554021124e-06, + "loss": 1.5301, + "step": 1605 + }, + { + "epoch": 1.76, + "grad_norm": 2.6002280712127686, + "learning_rate": 2.3294069861900897e-06, + "loss": 1.5238, + "step": 1606 + }, + { + "epoch": 1.7610958904109588, + "grad_norm": 2.8312838077545166, + "learning_rate": 2.327376116978067e-06, + "loss": 1.5872, + "step": 1607 + }, + { + "epoch": 1.7621917808219179, + "grad_norm": 2.6640195846557617, + "learning_rate": 2.3253452477660442e-06, + "loss": 1.5416, + "step": 1608 + }, + { + "epoch": 1.7632876712328767, + "grad_norm": 2.8075945377349854, + "learning_rate": 2.3233143785540215e-06, + "loss": 1.5823, + "step": 1609 + }, + { + "epoch": 1.7643835616438355, + "grad_norm": 2.6641950607299805, + "learning_rate": 2.3212835093419988e-06, + "loss": 1.5285, + "step": 1610 + }, + { + "epoch": 1.7643835616438355, + "eval_loss": 1.7819304466247559, + "eval_runtime": 45.6364, + "eval_samples_per_second": 21.912, + "eval_steps_per_second": 0.92, + "step": 1610 + }, + { + "epoch": 1.7654794520547945, + "grad_norm": 2.970372438430786, + "learning_rate": 2.319252640129976e-06, + "loss": 1.6271, + "step": 1611 + }, + { + "epoch": 1.7665753424657535, + "grad_norm": 2.596376657485962, + "learning_rate": 2.3172217709179533e-06, + "loss": 1.5898, + "step": 1612 + }, + { + "epoch": 1.7676712328767124, + "grad_norm": 2.6436660289764404, + "learning_rate": 2.3151909017059306e-06, + "loss": 1.5074, + "step": 1613 + }, + { + "epoch": 1.7687671232876712, + "grad_norm": 2.7881598472595215, + "learning_rate": 2.313160032493908e-06, + "loss": 1.5761, + "step": 1614 + }, + { + "epoch": 1.7698630136986302, + "grad_norm": 2.8422069549560547, + "learning_rate": 2.311129163281885e-06, + "loss": 1.542, + "step": 1615 + }, + { + "epoch": 1.770958904109589, + "grad_norm": 2.6202616691589355, + "learning_rate": 2.309098294069862e-06, + "loss": 1.5212, + "step": 1616 + }, + { + "epoch": 1.7720547945205478, + "grad_norm": 2.7313668727874756, + "learning_rate": 2.3070674248578393e-06, + "loss": 1.5608, + "step": 1617 + }, + { + "epoch": 1.7731506849315068, + "grad_norm": 2.8208658695220947, + "learning_rate": 2.3050365556458166e-06, + "loss": 1.589, + "step": 1618 + }, + { + "epoch": 1.7742465753424659, + "grad_norm": 2.670839786529541, + "learning_rate": 2.303005686433794e-06, + "loss": 1.5565, + "step": 1619 + }, + { + "epoch": 1.7753424657534247, + "grad_norm": 2.6585769653320312, + "learning_rate": 2.300974817221771e-06, + "loss": 1.5148, + "step": 1620 + }, + { + "epoch": 1.7753424657534247, + "eval_loss": 1.7786930799484253, + "eval_runtime": 44.9833, + "eval_samples_per_second": 22.23, + "eval_steps_per_second": 0.934, + "step": 1620 + }, + { + "epoch": 1.7764383561643835, + "grad_norm": 2.6813182830810547, + "learning_rate": 2.2989439480097484e-06, + "loss": 1.5796, + "step": 1621 + }, + { + "epoch": 1.7775342465753425, + "grad_norm": 2.620898485183716, + "learning_rate": 2.2969130787977257e-06, + "loss": 1.5945, + "step": 1622 + }, + { + "epoch": 1.7786301369863013, + "grad_norm": 2.5621166229248047, + "learning_rate": 2.294882209585703e-06, + "loss": 1.5334, + "step": 1623 + }, + { + "epoch": 1.7797260273972602, + "grad_norm": 2.7039835453033447, + "learning_rate": 2.2928513403736802e-06, + "loss": 1.5651, + "step": 1624 + }, + { + "epoch": 1.7808219178082192, + "grad_norm": 2.730381965637207, + "learning_rate": 2.2908204711616575e-06, + "loss": 1.606, + "step": 1625 + }, + { + "epoch": 1.7819178082191782, + "grad_norm": 2.58190655708313, + "learning_rate": 2.2887896019496348e-06, + "loss": 1.5494, + "step": 1626 + }, + { + "epoch": 1.783013698630137, + "grad_norm": 2.643354654312134, + "learning_rate": 2.286758732737612e-06, + "loss": 1.5565, + "step": 1627 + }, + { + "epoch": 1.7841095890410958, + "grad_norm": 2.6504781246185303, + "learning_rate": 2.284727863525589e-06, + "loss": 1.5953, + "step": 1628 + }, + { + "epoch": 1.7852054794520549, + "grad_norm": 2.6131460666656494, + "learning_rate": 2.282696994313566e-06, + "loss": 1.5623, + "step": 1629 + }, + { + "epoch": 1.7863013698630137, + "grad_norm": 2.768601655960083, + "learning_rate": 2.2806661251015434e-06, + "loss": 1.5527, + "step": 1630 + }, + { + "epoch": 1.7863013698630137, + "eval_loss": 1.7815148830413818, + "eval_runtime": 49.2594, + "eval_samples_per_second": 20.301, + "eval_steps_per_second": 0.853, + "step": 1630 + }, + { + "epoch": 1.7873972602739725, + "grad_norm": 2.598111152648926, + "learning_rate": 2.2786352558895207e-06, + "loss": 1.5117, + "step": 1631 + }, + { + "epoch": 1.7884931506849315, + "grad_norm": 2.6045982837677, + "learning_rate": 2.276604386677498e-06, + "loss": 1.5367, + "step": 1632 + }, + { + "epoch": 1.7895890410958906, + "grad_norm": 2.737748146057129, + "learning_rate": 2.2745735174654753e-06, + "loss": 1.6116, + "step": 1633 + }, + { + "epoch": 1.7906849315068492, + "grad_norm": 2.6198277473449707, + "learning_rate": 2.2725426482534526e-06, + "loss": 1.5657, + "step": 1634 + }, + { + "epoch": 1.7917808219178082, + "grad_norm": 2.848696708679199, + "learning_rate": 2.27051177904143e-06, + "loss": 1.5509, + "step": 1635 + }, + { + "epoch": 1.7928767123287672, + "grad_norm": 2.7039992809295654, + "learning_rate": 2.268480909829407e-06, + "loss": 1.5698, + "step": 1636 + }, + { + "epoch": 1.793972602739726, + "grad_norm": 2.7331910133361816, + "learning_rate": 2.2664500406173844e-06, + "loss": 1.5935, + "step": 1637 + }, + { + "epoch": 1.7950684931506848, + "grad_norm": 2.922910451889038, + "learning_rate": 2.2644191714053617e-06, + "loss": 1.5553, + "step": 1638 + }, + { + "epoch": 1.7961643835616439, + "grad_norm": 2.7473456859588623, + "learning_rate": 2.262388302193339e-06, + "loss": 1.5449, + "step": 1639 + }, + { + "epoch": 1.7972602739726027, + "grad_norm": 2.749073028564453, + "learning_rate": 2.260357432981316e-06, + "loss": 1.5723, + "step": 1640 + }, + { + "epoch": 1.7972602739726027, + "eval_loss": 1.7806612253189087, + "eval_runtime": 46.7453, + "eval_samples_per_second": 21.393, + "eval_steps_per_second": 0.898, + "step": 1640 + }, + { + "epoch": 1.7983561643835615, + "grad_norm": 2.6906771659851074, + "learning_rate": 2.2583265637692935e-06, + "loss": 1.5738, + "step": 1641 + }, + { + "epoch": 1.7994520547945205, + "grad_norm": 2.6401400566101074, + "learning_rate": 2.2562956945572708e-06, + "loss": 1.5637, + "step": 1642 + }, + { + "epoch": 1.8005479452054796, + "grad_norm": 2.739522695541382, + "learning_rate": 2.254264825345248e-06, + "loss": 1.5699, + "step": 1643 + }, + { + "epoch": 1.8016438356164384, + "grad_norm": 2.6827986240386963, + "learning_rate": 2.2522339561332253e-06, + "loss": 1.5748, + "step": 1644 + }, + { + "epoch": 1.8027397260273972, + "grad_norm": 2.859121561050415, + "learning_rate": 2.2502030869212026e-06, + "loss": 1.5988, + "step": 1645 + }, + { + "epoch": 1.8038356164383562, + "grad_norm": 2.5924315452575684, + "learning_rate": 2.24817221770918e-06, + "loss": 1.5845, + "step": 1646 + }, + { + "epoch": 1.804931506849315, + "grad_norm": 2.5433928966522217, + "learning_rate": 2.246141348497157e-06, + "loss": 1.5487, + "step": 1647 + }, + { + "epoch": 1.8060273972602738, + "grad_norm": 2.591356039047241, + "learning_rate": 2.2441104792851344e-06, + "loss": 1.568, + "step": 1648 + }, + { + "epoch": 1.8071232876712329, + "grad_norm": 2.648468255996704, + "learning_rate": 2.2420796100731117e-06, + "loss": 1.5921, + "step": 1649 + }, + { + "epoch": 1.808219178082192, + "grad_norm": 2.613450765609741, + "learning_rate": 2.240048740861089e-06, + "loss": 1.4625, + "step": 1650 + }, + { + "epoch": 1.808219178082192, + "eval_loss": 1.7814301252365112, + "eval_runtime": 47.0872, + "eval_samples_per_second": 21.237, + "eval_steps_per_second": 0.892, + "step": 1650 + }, + { + "epoch": 1.8093150684931507, + "grad_norm": 2.558629035949707, + "learning_rate": 2.2380178716490662e-06, + "loss": 1.586, + "step": 1651 + }, + { + "epoch": 1.8104109589041095, + "grad_norm": 2.6624038219451904, + "learning_rate": 2.2359870024370435e-06, + "loss": 1.5637, + "step": 1652 + }, + { + "epoch": 1.8115068493150686, + "grad_norm": 2.6640877723693848, + "learning_rate": 2.2339561332250204e-06, + "loss": 1.5585, + "step": 1653 + }, + { + "epoch": 1.8126027397260274, + "grad_norm": 2.6179189682006836, + "learning_rate": 2.2319252640129976e-06, + "loss": 1.5051, + "step": 1654 + }, + { + "epoch": 1.8136986301369862, + "grad_norm": 2.6445083618164062, + "learning_rate": 2.229894394800975e-06, + "loss": 1.4866, + "step": 1655 + }, + { + "epoch": 1.8147945205479452, + "grad_norm": 2.5804684162139893, + "learning_rate": 2.227863525588952e-06, + "loss": 1.5251, + "step": 1656 + }, + { + "epoch": 1.8158904109589042, + "grad_norm": 2.603933572769165, + "learning_rate": 2.2258326563769295e-06, + "loss": 1.5107, + "step": 1657 + }, + { + "epoch": 1.816986301369863, + "grad_norm": 2.7108941078186035, + "learning_rate": 2.2238017871649067e-06, + "loss": 1.5602, + "step": 1658 + }, + { + "epoch": 1.8180821917808219, + "grad_norm": 2.727480888366699, + "learning_rate": 2.221770917952884e-06, + "loss": 1.5918, + "step": 1659 + }, + { + "epoch": 1.819178082191781, + "grad_norm": 2.6257543563842773, + "learning_rate": 2.2197400487408613e-06, + "loss": 1.5265, + "step": 1660 + }, + { + "epoch": 1.819178082191781, + "eval_loss": 1.7834619283676147, + "eval_runtime": 48.8556, + "eval_samples_per_second": 20.468, + "eval_steps_per_second": 0.86, + "step": 1660 + }, + { + "epoch": 1.8202739726027397, + "grad_norm": 2.8918514251708984, + "learning_rate": 2.2177091795288386e-06, + "loss": 1.5413, + "step": 1661 + }, + { + "epoch": 1.8213698630136985, + "grad_norm": 2.6081483364105225, + "learning_rate": 2.215678310316816e-06, + "loss": 1.5716, + "step": 1662 + }, + { + "epoch": 1.8224657534246576, + "grad_norm": 2.559870719909668, + "learning_rate": 2.213647441104793e-06, + "loss": 1.5277, + "step": 1663 + }, + { + "epoch": 1.8235616438356166, + "grad_norm": 2.703738212585449, + "learning_rate": 2.2116165718927704e-06, + "loss": 1.5563, + "step": 1664 + }, + { + "epoch": 1.8246575342465754, + "grad_norm": 2.902691602706909, + "learning_rate": 2.2095857026807473e-06, + "loss": 1.5241, + "step": 1665 + }, + { + "epoch": 1.8257534246575342, + "grad_norm": 2.6989998817443848, + "learning_rate": 2.2075548334687245e-06, + "loss": 1.5491, + "step": 1666 + }, + { + "epoch": 1.8268493150684932, + "grad_norm": 2.5755252838134766, + "learning_rate": 2.205523964256702e-06, + "loss": 1.5736, + "step": 1667 + }, + { + "epoch": 1.827945205479452, + "grad_norm": 2.6535956859588623, + "learning_rate": 2.203493095044679e-06, + "loss": 1.5822, + "step": 1668 + }, + { + "epoch": 1.8290410958904109, + "grad_norm": 2.577737808227539, + "learning_rate": 2.2014622258326564e-06, + "loss": 1.5927, + "step": 1669 + }, + { + "epoch": 1.83013698630137, + "grad_norm": 2.608058214187622, + "learning_rate": 2.1994313566206336e-06, + "loss": 1.5472, + "step": 1670 + }, + { + "epoch": 1.83013698630137, + "eval_loss": 1.7814104557037354, + "eval_runtime": 45.5588, + "eval_samples_per_second": 21.95, + "eval_steps_per_second": 0.922, + "step": 1670 + }, + { + "epoch": 1.831232876712329, + "grad_norm": 2.8584563732147217, + "learning_rate": 2.197400487408611e-06, + "loss": 1.5168, + "step": 1671 + }, + { + "epoch": 1.8323287671232875, + "grad_norm": 2.6451761722564697, + "learning_rate": 2.195369618196588e-06, + "loss": 1.5447, + "step": 1672 + }, + { + "epoch": 1.8334246575342465, + "grad_norm": 2.8102355003356934, + "learning_rate": 2.1933387489845655e-06, + "loss": 1.5827, + "step": 1673 + }, + { + "epoch": 1.8345205479452056, + "grad_norm": 2.731776714324951, + "learning_rate": 2.1913078797725427e-06, + "loss": 1.6249, + "step": 1674 + }, + { + "epoch": 1.8356164383561644, + "grad_norm": 3.7058873176574707, + "learning_rate": 2.18927701056052e-06, + "loss": 1.5718, + "step": 1675 + }, + { + "epoch": 1.8367123287671232, + "grad_norm": 2.8634655475616455, + "learning_rate": 2.1872461413484973e-06, + "loss": 1.5592, + "step": 1676 + }, + { + "epoch": 1.8378082191780822, + "grad_norm": 2.7904224395751953, + "learning_rate": 2.1852152721364746e-06, + "loss": 1.5412, + "step": 1677 + }, + { + "epoch": 1.838904109589041, + "grad_norm": 2.6223487854003906, + "learning_rate": 2.183184402924452e-06, + "loss": 1.5754, + "step": 1678 + }, + { + "epoch": 1.8399999999999999, + "grad_norm": 2.8098909854888916, + "learning_rate": 2.181153533712429e-06, + "loss": 1.5605, + "step": 1679 + }, + { + "epoch": 1.841095890410959, + "grad_norm": 2.801152467727661, + "learning_rate": 2.1791226645004064e-06, + "loss": 1.5502, + "step": 1680 + }, + { + "epoch": 1.841095890410959, + "eval_loss": 1.7773977518081665, + "eval_runtime": 46.6529, + "eval_samples_per_second": 21.435, + "eval_steps_per_second": 0.9, + "step": 1680 + }, + { + "epoch": 1.842191780821918, + "grad_norm": 2.9683728218078613, + "learning_rate": 2.1770917952883837e-06, + "loss": 1.6183, + "step": 1681 + }, + { + "epoch": 1.8432876712328767, + "grad_norm": 2.740091562271118, + "learning_rate": 2.175060926076361e-06, + "loss": 1.6229, + "step": 1682 + }, + { + "epoch": 1.8443835616438355, + "grad_norm": 2.718059778213501, + "learning_rate": 2.1730300568643382e-06, + "loss": 1.5476, + "step": 1683 + }, + { + "epoch": 1.8454794520547946, + "grad_norm": 2.9417316913604736, + "learning_rate": 2.1709991876523155e-06, + "loss": 1.4917, + "step": 1684 + }, + { + "epoch": 1.8465753424657534, + "grad_norm": 2.7604305744171143, + "learning_rate": 2.1689683184402928e-06, + "loss": 1.5743, + "step": 1685 + }, + { + "epoch": 1.8476712328767122, + "grad_norm": 2.6518256664276123, + "learning_rate": 2.16693744922827e-06, + "loss": 1.5306, + "step": 1686 + }, + { + "epoch": 1.8487671232876712, + "grad_norm": 2.6570379734039307, + "learning_rate": 2.1649065800162473e-06, + "loss": 1.5377, + "step": 1687 + }, + { + "epoch": 1.8498630136986303, + "grad_norm": 2.655618906021118, + "learning_rate": 2.1628757108042246e-06, + "loss": 1.5317, + "step": 1688 + }, + { + "epoch": 1.850958904109589, + "grad_norm": 2.7146053314208984, + "learning_rate": 2.160844841592202e-06, + "loss": 1.5331, + "step": 1689 + }, + { + "epoch": 1.8520547945205479, + "grad_norm": 2.7635397911071777, + "learning_rate": 2.1588139723801787e-06, + "loss": 1.5383, + "step": 1690 + }, + { + "epoch": 1.8520547945205479, + "eval_loss": 1.7732168436050415, + "eval_runtime": 46.4051, + "eval_samples_per_second": 21.549, + "eval_steps_per_second": 0.905, + "step": 1690 + }, + { + "epoch": 1.853150684931507, + "grad_norm": 2.6774237155914307, + "learning_rate": 2.156783103168156e-06, + "loss": 1.5154, + "step": 1691 + }, + { + "epoch": 1.8542465753424657, + "grad_norm": 2.7335660457611084, + "learning_rate": 2.1547522339561333e-06, + "loss": 1.5765, + "step": 1692 + }, + { + "epoch": 1.8553424657534245, + "grad_norm": 2.942945957183838, + "learning_rate": 2.1527213647441106e-06, + "loss": 1.6017, + "step": 1693 + }, + { + "epoch": 1.8564383561643836, + "grad_norm": 2.6062228679656982, + "learning_rate": 2.150690495532088e-06, + "loss": 1.5707, + "step": 1694 + }, + { + "epoch": 1.8575342465753426, + "grad_norm": 2.6039342880249023, + "learning_rate": 2.148659626320065e-06, + "loss": 1.5801, + "step": 1695 + }, + { + "epoch": 1.8586301369863014, + "grad_norm": 2.6357412338256836, + "learning_rate": 2.1466287571080424e-06, + "loss": 1.5326, + "step": 1696 + }, + { + "epoch": 1.8597260273972602, + "grad_norm": 2.6265313625335693, + "learning_rate": 2.1445978878960197e-06, + "loss": 1.5463, + "step": 1697 + }, + { + "epoch": 1.8608219178082193, + "grad_norm": 2.6953911781311035, + "learning_rate": 2.142567018683997e-06, + "loss": 1.607, + "step": 1698 + }, + { + "epoch": 1.861917808219178, + "grad_norm": 2.6380672454833984, + "learning_rate": 2.140536149471974e-06, + "loss": 1.5227, + "step": 1699 + }, + { + "epoch": 1.8630136986301369, + "grad_norm": 2.846090078353882, + "learning_rate": 2.1385052802599515e-06, + "loss": 1.5391, + "step": 1700 + }, + { + "epoch": 1.8630136986301369, + "eval_loss": 1.7703944444656372, + "eval_runtime": 48.2568, + "eval_samples_per_second": 20.722, + "eval_steps_per_second": 0.87, + "step": 1700 + }, + { + "epoch": 1.864109589041096, + "grad_norm": 2.7720913887023926, + "learning_rate": 2.1364744110479288e-06, + "loss": 1.5936, + "step": 1701 + }, + { + "epoch": 1.865205479452055, + "grad_norm": 2.6953914165496826, + "learning_rate": 2.134443541835906e-06, + "loss": 1.56, + "step": 1702 + }, + { + "epoch": 1.8663013698630135, + "grad_norm": 2.594421863555908, + "learning_rate": 2.132412672623883e-06, + "loss": 1.5314, + "step": 1703 + }, + { + "epoch": 1.8673972602739726, + "grad_norm": 2.9509782791137695, + "learning_rate": 2.13038180341186e-06, + "loss": 1.5616, + "step": 1704 + }, + { + "epoch": 1.8684931506849316, + "grad_norm": 2.8655569553375244, + "learning_rate": 2.1283509341998374e-06, + "loss": 1.595, + "step": 1705 + }, + { + "epoch": 1.8695890410958904, + "grad_norm": 2.731106758117676, + "learning_rate": 2.1263200649878147e-06, + "loss": 1.551, + "step": 1706 + }, + { + "epoch": 1.8706849315068492, + "grad_norm": 2.564788579940796, + "learning_rate": 2.124289195775792e-06, + "loss": 1.548, + "step": 1707 + }, + { + "epoch": 1.8717808219178083, + "grad_norm": 2.6494832038879395, + "learning_rate": 2.1222583265637693e-06, + "loss": 1.4766, + "step": 1708 + }, + { + "epoch": 1.872876712328767, + "grad_norm": 2.5751700401306152, + "learning_rate": 2.1202274573517465e-06, + "loss": 1.5154, + "step": 1709 + }, + { + "epoch": 1.8739726027397259, + "grad_norm": 2.7771973609924316, + "learning_rate": 2.118196588139724e-06, + "loss": 1.5418, + "step": 1710 + }, + { + "epoch": 1.8739726027397259, + "eval_loss": 1.7701855897903442, + "eval_runtime": 42.8097, + "eval_samples_per_second": 23.359, + "eval_steps_per_second": 0.981, + "step": 1710 + }, + { + "epoch": 1.875068493150685, + "grad_norm": 2.6338229179382324, + "learning_rate": 2.116165718927701e-06, + "loss": 1.5661, + "step": 1711 + }, + { + "epoch": 1.876164383561644, + "grad_norm": 2.5472185611724854, + "learning_rate": 2.1141348497156784e-06, + "loss": 1.5201, + "step": 1712 + }, + { + "epoch": 1.8772602739726028, + "grad_norm": 2.643543243408203, + "learning_rate": 2.1121039805036556e-06, + "loss": 1.5226, + "step": 1713 + }, + { + "epoch": 1.8783561643835616, + "grad_norm": 2.650101900100708, + "learning_rate": 2.110073111291633e-06, + "loss": 1.543, + "step": 1714 + }, + { + "epoch": 1.8794520547945206, + "grad_norm": 2.7839369773864746, + "learning_rate": 2.10804224207961e-06, + "loss": 1.5831, + "step": 1715 + }, + { + "epoch": 1.8805479452054794, + "grad_norm": 2.8418030738830566, + "learning_rate": 2.1060113728675875e-06, + "loss": 1.5888, + "step": 1716 + }, + { + "epoch": 1.8816438356164382, + "grad_norm": 2.585972309112549, + "learning_rate": 2.1039805036555647e-06, + "loss": 1.4632, + "step": 1717 + }, + { + "epoch": 1.8827397260273973, + "grad_norm": 2.694457769393921, + "learning_rate": 2.101949634443542e-06, + "loss": 1.549, + "step": 1718 + }, + { + "epoch": 1.8838356164383563, + "grad_norm": 2.7106211185455322, + "learning_rate": 2.0999187652315193e-06, + "loss": 1.5184, + "step": 1719 + }, + { + "epoch": 1.884931506849315, + "grad_norm": 2.7540805339813232, + "learning_rate": 2.0978878960194966e-06, + "loss": 1.5588, + "step": 1720 + }, + { + "epoch": 1.884931506849315, + "eval_loss": 1.7676713466644287, + "eval_runtime": 43.4914, + "eval_samples_per_second": 22.993, + "eval_steps_per_second": 0.966, + "step": 1720 + }, + { + "epoch": 1.886027397260274, + "grad_norm": 2.702197313308716, + "learning_rate": 2.095857026807474e-06, + "loss": 1.5447, + "step": 1721 + }, + { + "epoch": 1.887123287671233, + "grad_norm": 2.716628313064575, + "learning_rate": 2.093826157595451e-06, + "loss": 1.5416, + "step": 1722 + }, + { + "epoch": 1.8882191780821918, + "grad_norm": 2.6568949222564697, + "learning_rate": 2.0917952883834284e-06, + "loss": 1.5423, + "step": 1723 + }, + { + "epoch": 1.8893150684931506, + "grad_norm": 2.8084635734558105, + "learning_rate": 2.0897644191714057e-06, + "loss": 1.598, + "step": 1724 + }, + { + "epoch": 1.8904109589041096, + "grad_norm": 2.67305064201355, + "learning_rate": 2.087733549959383e-06, + "loss": 1.5194, + "step": 1725 + }, + { + "epoch": 1.8915068493150686, + "grad_norm": 2.6713247299194336, + "learning_rate": 2.0857026807473602e-06, + "loss": 1.5022, + "step": 1726 + }, + { + "epoch": 1.8926027397260274, + "grad_norm": 2.912266731262207, + "learning_rate": 2.0836718115353375e-06, + "loss": 1.5869, + "step": 1727 + }, + { + "epoch": 1.8936986301369862, + "grad_norm": 2.6969048976898193, + "learning_rate": 2.0816409423233144e-06, + "loss": 1.5379, + "step": 1728 + }, + { + "epoch": 1.8947945205479453, + "grad_norm": 2.6329574584960938, + "learning_rate": 2.0796100731112916e-06, + "loss": 1.5457, + "step": 1729 + }, + { + "epoch": 1.895890410958904, + "grad_norm": 2.6367034912109375, + "learning_rate": 2.077579203899269e-06, + "loss": 1.4919, + "step": 1730 + }, + { + "epoch": 1.895890410958904, + "eval_loss": 1.7656192779541016, + "eval_runtime": 44.8406, + "eval_samples_per_second": 22.301, + "eval_steps_per_second": 0.937, + "step": 1730 + }, + { + "epoch": 1.896986301369863, + "grad_norm": 2.8109962940216064, + "learning_rate": 2.075548334687246e-06, + "loss": 1.5674, + "step": 1731 + }, + { + "epoch": 1.898082191780822, + "grad_norm": 2.5886316299438477, + "learning_rate": 2.0735174654752235e-06, + "loss": 1.4996, + "step": 1732 + }, + { + "epoch": 1.899178082191781, + "grad_norm": 2.6195716857910156, + "learning_rate": 2.0714865962632007e-06, + "loss": 1.5339, + "step": 1733 + }, + { + "epoch": 1.9002739726027398, + "grad_norm": 2.6813786029815674, + "learning_rate": 2.069455727051178e-06, + "loss": 1.5156, + "step": 1734 + }, + { + "epoch": 1.9013698630136986, + "grad_norm": 2.726902484893799, + "learning_rate": 2.0674248578391553e-06, + "loss": 1.5152, + "step": 1735 + }, + { + "epoch": 1.9024657534246576, + "grad_norm": 2.845074415206909, + "learning_rate": 2.0653939886271326e-06, + "loss": 1.5968, + "step": 1736 + }, + { + "epoch": 1.9035616438356164, + "grad_norm": 2.6655383110046387, + "learning_rate": 2.06336311941511e-06, + "loss": 1.5396, + "step": 1737 + }, + { + "epoch": 1.9046575342465752, + "grad_norm": 2.759390115737915, + "learning_rate": 2.061332250203087e-06, + "loss": 1.5071, + "step": 1738 + }, + { + "epoch": 1.9057534246575343, + "grad_norm": 2.6290149688720703, + "learning_rate": 2.0593013809910644e-06, + "loss": 1.552, + "step": 1739 + }, + { + "epoch": 1.9068493150684933, + "grad_norm": 2.6996374130249023, + "learning_rate": 2.0572705117790417e-06, + "loss": 1.5691, + "step": 1740 + }, + { + "epoch": 1.9068493150684933, + "eval_loss": 1.7706645727157593, + "eval_runtime": 43.2696, + "eval_samples_per_second": 23.111, + "eval_steps_per_second": 0.971, + "step": 1740 + }, + { + "epoch": 1.907945205479452, + "grad_norm": 2.574399948120117, + "learning_rate": 2.055239642567019e-06, + "loss": 1.5489, + "step": 1741 + }, + { + "epoch": 1.909041095890411, + "grad_norm": 2.6261112689971924, + "learning_rate": 2.0532087733549962e-06, + "loss": 1.579, + "step": 1742 + }, + { + "epoch": 1.91013698630137, + "grad_norm": 2.7730929851531982, + "learning_rate": 2.0511779041429735e-06, + "loss": 1.5049, + "step": 1743 + }, + { + "epoch": 1.9112328767123288, + "grad_norm": 2.706677198410034, + "learning_rate": 2.0491470349309508e-06, + "loss": 1.57, + "step": 1744 + }, + { + "epoch": 1.9123287671232876, + "grad_norm": 2.667128801345825, + "learning_rate": 2.047116165718928e-06, + "loss": 1.5261, + "step": 1745 + }, + { + "epoch": 1.9134246575342466, + "grad_norm": 2.7455623149871826, + "learning_rate": 2.0450852965069053e-06, + "loss": 1.5812, + "step": 1746 + }, + { + "epoch": 1.9145205479452054, + "grad_norm": 2.6654391288757324, + "learning_rate": 2.0430544272948826e-06, + "loss": 1.5428, + "step": 1747 + }, + { + "epoch": 1.9156164383561642, + "grad_norm": 2.6527786254882812, + "learning_rate": 2.04102355808286e-06, + "loss": 1.5019, + "step": 1748 + }, + { + "epoch": 1.9167123287671233, + "grad_norm": 2.640010118484497, + "learning_rate": 2.038992688870837e-06, + "loss": 1.5245, + "step": 1749 + }, + { + "epoch": 1.9178082191780823, + "grad_norm": 2.5960302352905273, + "learning_rate": 2.0369618196588144e-06, + "loss": 1.5343, + "step": 1750 + }, + { + "epoch": 1.9178082191780823, + "eval_loss": 1.767797827720642, + "eval_runtime": 44.0669, + "eval_samples_per_second": 22.693, + "eval_steps_per_second": 0.953, + "step": 1750 + }, + { + "epoch": 1.9189041095890411, + "grad_norm": 2.7159781455993652, + "learning_rate": 2.0349309504467917e-06, + "loss": 1.5708, + "step": 1751 + }, + { + "epoch": 1.92, + "grad_norm": 2.626908540725708, + "learning_rate": 2.032900081234769e-06, + "loss": 1.5749, + "step": 1752 + }, + { + "epoch": 1.921095890410959, + "grad_norm": 2.538393259048462, + "learning_rate": 2.030869212022746e-06, + "loss": 1.5292, + "step": 1753 + }, + { + "epoch": 1.9221917808219178, + "grad_norm": 2.562856912612915, + "learning_rate": 2.028838342810723e-06, + "loss": 1.5345, + "step": 1754 + }, + { + "epoch": 1.9232876712328766, + "grad_norm": 2.639636278152466, + "learning_rate": 2.0268074735987004e-06, + "loss": 1.5255, + "step": 1755 + }, + { + "epoch": 1.9243835616438356, + "grad_norm": 2.6585395336151123, + "learning_rate": 2.0247766043866777e-06, + "loss": 1.5743, + "step": 1756 + }, + { + "epoch": 1.9254794520547946, + "grad_norm": 2.6081788539886475, + "learning_rate": 2.022745735174655e-06, + "loss": 1.5438, + "step": 1757 + }, + { + "epoch": 1.9265753424657535, + "grad_norm": 2.9027206897735596, + "learning_rate": 2.020714865962632e-06, + "loss": 1.5647, + "step": 1758 + }, + { + "epoch": 1.9276712328767123, + "grad_norm": 2.73712158203125, + "learning_rate": 2.0186839967506095e-06, + "loss": 1.545, + "step": 1759 + }, + { + "epoch": 1.9287671232876713, + "grad_norm": 2.5918891429901123, + "learning_rate": 2.0166531275385868e-06, + "loss": 1.4988, + "step": 1760 + }, + { + "epoch": 1.9287671232876713, + "eval_loss": 1.7693538665771484, + "eval_runtime": 43.3757, + "eval_samples_per_second": 23.054, + "eval_steps_per_second": 0.968, + "step": 1760 + }, + { + "epoch": 1.9298630136986301, + "grad_norm": 2.7728114128112793, + "learning_rate": 2.014622258326564e-06, + "loss": 1.5617, + "step": 1761 + }, + { + "epoch": 1.930958904109589, + "grad_norm": 2.6768674850463867, + "learning_rate": 2.0125913891145413e-06, + "loss": 1.5483, + "step": 1762 + }, + { + "epoch": 1.932054794520548, + "grad_norm": 2.7239630222320557, + "learning_rate": 2.0105605199025186e-06, + "loss": 1.542, + "step": 1763 + }, + { + "epoch": 1.933150684931507, + "grad_norm": 2.774303436279297, + "learning_rate": 2.008529650690496e-06, + "loss": 1.5789, + "step": 1764 + }, + { + "epoch": 1.9342465753424658, + "grad_norm": 2.6659293174743652, + "learning_rate": 2.0064987814784727e-06, + "loss": 1.5331, + "step": 1765 + }, + { + "epoch": 1.9353424657534246, + "grad_norm": 2.535292148590088, + "learning_rate": 2.00446791226645e-06, + "loss": 1.509, + "step": 1766 + }, + { + "epoch": 1.9364383561643836, + "grad_norm": 2.532724380493164, + "learning_rate": 2.0024370430544273e-06, + "loss": 1.5627, + "step": 1767 + }, + { + "epoch": 1.9375342465753425, + "grad_norm": 2.6539125442504883, + "learning_rate": 2.0004061738424045e-06, + "loss": 1.5859, + "step": 1768 + }, + { + "epoch": 1.9386301369863013, + "grad_norm": 2.4748764038085938, + "learning_rate": 1.998375304630382e-06, + "loss": 1.5622, + "step": 1769 + }, + { + "epoch": 1.9397260273972603, + "grad_norm": 2.621659517288208, + "learning_rate": 1.996344435418359e-06, + "loss": 1.5631, + "step": 1770 + }, + { + "epoch": 1.9397260273972603, + "eval_loss": 1.7688086032867432, + "eval_runtime": 43.8443, + "eval_samples_per_second": 22.808, + "eval_steps_per_second": 0.958, + "step": 1770 + }, + { + "epoch": 1.9408219178082193, + "grad_norm": 2.9397199153900146, + "learning_rate": 1.9943135662063364e-06, + "loss": 1.5744, + "step": 1771 + }, + { + "epoch": 1.941917808219178, + "grad_norm": 2.80846905708313, + "learning_rate": 1.9922826969943136e-06, + "loss": 1.5725, + "step": 1772 + }, + { + "epoch": 1.943013698630137, + "grad_norm": 2.6658272743225098, + "learning_rate": 1.990251827782291e-06, + "loss": 1.576, + "step": 1773 + }, + { + "epoch": 1.944109589041096, + "grad_norm": 2.7540268898010254, + "learning_rate": 1.988220958570268e-06, + "loss": 1.5528, + "step": 1774 + }, + { + "epoch": 1.9452054794520548, + "grad_norm": 2.8099920749664307, + "learning_rate": 1.9861900893582455e-06, + "loss": 1.5593, + "step": 1775 + }, + { + "epoch": 1.9463013698630136, + "grad_norm": 2.668316602706909, + "learning_rate": 1.9841592201462227e-06, + "loss": 1.5144, + "step": 1776 + }, + { + "epoch": 1.9473972602739726, + "grad_norm": 2.729696750640869, + "learning_rate": 1.9821283509342e-06, + "loss": 1.5161, + "step": 1777 + }, + { + "epoch": 1.9484931506849315, + "grad_norm": 2.6628568172454834, + "learning_rate": 1.9800974817221773e-06, + "loss": 1.565, + "step": 1778 + }, + { + "epoch": 1.9495890410958903, + "grad_norm": 2.7558653354644775, + "learning_rate": 1.9780666125101546e-06, + "loss": 1.5312, + "step": 1779 + }, + { + "epoch": 1.9506849315068493, + "grad_norm": 3.152547836303711, + "learning_rate": 1.976035743298132e-06, + "loss": 1.6071, + "step": 1780 + }, + { + "epoch": 1.9506849315068493, + "eval_loss": 1.7690435647964478, + "eval_runtime": 43.2873, + "eval_samples_per_second": 23.101, + "eval_steps_per_second": 0.97, + "step": 1780 + }, + { + "epoch": 1.9517808219178083, + "grad_norm": 2.7526113986968994, + "learning_rate": 1.974004874086109e-06, + "loss": 1.5105, + "step": 1781 + }, + { + "epoch": 1.9528767123287671, + "grad_norm": 2.575759172439575, + "learning_rate": 1.9719740048740864e-06, + "loss": 1.5454, + "step": 1782 + }, + { + "epoch": 1.953972602739726, + "grad_norm": 2.6542654037475586, + "learning_rate": 1.9699431356620637e-06, + "loss": 1.5325, + "step": 1783 + }, + { + "epoch": 1.955068493150685, + "grad_norm": 2.7686262130737305, + "learning_rate": 1.967912266450041e-06, + "loss": 1.524, + "step": 1784 + }, + { + "epoch": 1.9561643835616438, + "grad_norm": 2.6677448749542236, + "learning_rate": 1.9658813972380182e-06, + "loss": 1.623, + "step": 1785 + }, + { + "epoch": 1.9572602739726026, + "grad_norm": 2.6429874897003174, + "learning_rate": 1.9638505280259955e-06, + "loss": 1.5256, + "step": 1786 + }, + { + "epoch": 1.9583561643835616, + "grad_norm": 2.701997756958008, + "learning_rate": 1.9618196588139728e-06, + "loss": 1.5488, + "step": 1787 + }, + { + "epoch": 1.9594520547945207, + "grad_norm": 2.773179292678833, + "learning_rate": 1.95978878960195e-06, + "loss": 1.6104, + "step": 1788 + }, + { + "epoch": 1.9605479452054795, + "grad_norm": 2.71045184135437, + "learning_rate": 1.9577579203899273e-06, + "loss": 1.5758, + "step": 1789 + }, + { + "epoch": 1.9616438356164383, + "grad_norm": 2.667008876800537, + "learning_rate": 1.955727051177904e-06, + "loss": 1.5277, + "step": 1790 + }, + { + "epoch": 1.9616438356164383, + "eval_loss": 1.7701371908187866, + "eval_runtime": 44.1016, + "eval_samples_per_second": 22.675, + "eval_steps_per_second": 0.952, + "step": 1790 + }, + { + "epoch": 1.9627397260273973, + "grad_norm": 2.729663372039795, + "learning_rate": 1.9536961819658815e-06, + "loss": 1.5251, + "step": 1791 + }, + { + "epoch": 1.9638356164383561, + "grad_norm": 2.706153631210327, + "learning_rate": 1.9516653127538587e-06, + "loss": 1.5196, + "step": 1792 + }, + { + "epoch": 1.964931506849315, + "grad_norm": 2.638580083847046, + "learning_rate": 1.949634443541836e-06, + "loss": 1.5065, + "step": 1793 + }, + { + "epoch": 1.966027397260274, + "grad_norm": 2.6905441284179688, + "learning_rate": 1.9476035743298133e-06, + "loss": 1.5766, + "step": 1794 + }, + { + "epoch": 1.967123287671233, + "grad_norm": 2.7412495613098145, + "learning_rate": 1.9455727051177906e-06, + "loss": 1.539, + "step": 1795 + }, + { + "epoch": 1.9682191780821918, + "grad_norm": 2.669023036956787, + "learning_rate": 1.943541835905768e-06, + "loss": 1.592, + "step": 1796 + }, + { + "epoch": 1.9693150684931506, + "grad_norm": 2.670112371444702, + "learning_rate": 1.941510966693745e-06, + "loss": 1.5132, + "step": 1797 + }, + { + "epoch": 1.9704109589041097, + "grad_norm": 2.7127676010131836, + "learning_rate": 1.9394800974817224e-06, + "loss": 1.57, + "step": 1798 + }, + { + "epoch": 1.9715068493150685, + "grad_norm": 2.7849738597869873, + "learning_rate": 1.9374492282696997e-06, + "loss": 1.6033, + "step": 1799 + }, + { + "epoch": 1.9726027397260273, + "grad_norm": 3.07655668258667, + "learning_rate": 1.935418359057677e-06, + "loss": 1.5624, + "step": 1800 + }, + { + "epoch": 1.9726027397260273, + "eval_loss": 1.7671096324920654, + "eval_runtime": 44.0624, + "eval_samples_per_second": 22.695, + "eval_steps_per_second": 0.953, + "step": 1800 + }, + { + "epoch": 1.9736986301369863, + "grad_norm": 2.83571195602417, + "learning_rate": 1.9333874898456542e-06, + "loss": 1.577, + "step": 1801 + }, + { + "epoch": 1.9747945205479454, + "grad_norm": 2.7236087322235107, + "learning_rate": 1.931356620633631e-06, + "loss": 1.4957, + "step": 1802 + }, + { + "epoch": 1.9758904109589042, + "grad_norm": 2.8082962036132812, + "learning_rate": 1.9293257514216083e-06, + "loss": 1.5472, + "step": 1803 + }, + { + "epoch": 1.976986301369863, + "grad_norm": 2.779731512069702, + "learning_rate": 1.9272948822095856e-06, + "loss": 1.5165, + "step": 1804 + }, + { + "epoch": 1.978082191780822, + "grad_norm": 2.6644365787506104, + "learning_rate": 1.925264012997563e-06, + "loss": 1.5459, + "step": 1805 + }, + { + "epoch": 1.9791780821917808, + "grad_norm": 2.776524782180786, + "learning_rate": 1.92323314378554e-06, + "loss": 1.5778, + "step": 1806 + }, + { + "epoch": 1.9802739726027396, + "grad_norm": 2.566277265548706, + "learning_rate": 1.9212022745735174e-06, + "loss": 1.5202, + "step": 1807 + }, + { + "epoch": 1.9813698630136987, + "grad_norm": 2.600104331970215, + "learning_rate": 1.9191714053614947e-06, + "loss": 1.5145, + "step": 1808 + }, + { + "epoch": 1.9824657534246577, + "grad_norm": 2.577695608139038, + "learning_rate": 1.917140536149472e-06, + "loss": 1.4915, + "step": 1809 + }, + { + "epoch": 1.9835616438356163, + "grad_norm": 2.828059434890747, + "learning_rate": 1.9151096669374493e-06, + "loss": 1.5766, + "step": 1810 + }, + { + "epoch": 1.9835616438356163, + "eval_loss": 1.7653249502182007, + "eval_runtime": 42.3408, + "eval_samples_per_second": 23.618, + "eval_steps_per_second": 0.992, + "step": 1810 + }, + { + "epoch": 1.9846575342465753, + "grad_norm": 2.889705181121826, + "learning_rate": 1.9130787977254266e-06, + "loss": 1.5565, + "step": 1811 + }, + { + "epoch": 1.9857534246575343, + "grad_norm": 2.6573760509490967, + "learning_rate": 1.911047928513404e-06, + "loss": 1.5183, + "step": 1812 + }, + { + "epoch": 1.9868493150684932, + "grad_norm": 2.6458022594451904, + "learning_rate": 1.909017059301381e-06, + "loss": 1.5164, + "step": 1813 + }, + { + "epoch": 1.987945205479452, + "grad_norm": 2.5978479385375977, + "learning_rate": 1.9069861900893586e-06, + "loss": 1.5524, + "step": 1814 + }, + { + "epoch": 1.989041095890411, + "grad_norm": 2.6573572158813477, + "learning_rate": 1.9049553208773354e-06, + "loss": 1.5342, + "step": 1815 + }, + { + "epoch": 1.9901369863013698, + "grad_norm": 2.4940648078918457, + "learning_rate": 1.9029244516653127e-06, + "loss": 1.489, + "step": 1816 + }, + { + "epoch": 1.9912328767123286, + "grad_norm": 2.7048654556274414, + "learning_rate": 1.90089358245329e-06, + "loss": 1.529, + "step": 1817 + }, + { + "epoch": 1.9923287671232877, + "grad_norm": 2.63424015045166, + "learning_rate": 1.8988627132412673e-06, + "loss": 1.5661, + "step": 1818 + }, + { + "epoch": 1.9934246575342467, + "grad_norm": 2.5225062370300293, + "learning_rate": 1.8968318440292445e-06, + "loss": 1.5065, + "step": 1819 + }, + { + "epoch": 1.9945205479452055, + "grad_norm": 2.708705425262451, + "learning_rate": 1.8948009748172218e-06, + "loss": 1.5609, + "step": 1820 + }, + { + "epoch": 1.9945205479452055, + "eval_loss": 1.7619582414627075, + "eval_runtime": 45.3201, + "eval_samples_per_second": 22.065, + "eval_steps_per_second": 0.927, + "step": 1820 + }, + { + "epoch": 1.9956164383561643, + "grad_norm": 2.819265365600586, + "learning_rate": 1.892770105605199e-06, + "loss": 1.5681, + "step": 1821 + }, + { + "epoch": 1.9967123287671233, + "grad_norm": 2.585503339767456, + "learning_rate": 1.8907392363931764e-06, + "loss": 1.5213, + "step": 1822 + }, + { + "epoch": 1.9978082191780822, + "grad_norm": 2.693864345550537, + "learning_rate": 1.8887083671811536e-06, + "loss": 1.5516, + "step": 1823 + }, + { + "epoch": 1.998904109589041, + "grad_norm": 2.4569215774536133, + "learning_rate": 1.886677497969131e-06, + "loss": 1.4611, + "step": 1824 + }, + { + "epoch": 2.0, + "grad_norm": 2.8584630489349365, + "learning_rate": 1.8846466287571082e-06, + "loss": 1.5591, + "step": 1825 + }, + { + "epoch": 2.001095890410959, + "grad_norm": 2.8206143379211426, + "learning_rate": 1.8826157595450855e-06, + "loss": 1.4731, + "step": 1826 + }, + { + "epoch": 2.0021917808219176, + "grad_norm": 2.8901588916778564, + "learning_rate": 1.8805848903330625e-06, + "loss": 1.6144, + "step": 1827 + }, + { + "epoch": 2.0032876712328767, + "grad_norm": 2.8405115604400635, + "learning_rate": 1.8785540211210398e-06, + "loss": 1.5211, + "step": 1828 + }, + { + "epoch": 2.0043835616438357, + "grad_norm": 2.944380283355713, + "learning_rate": 1.876523151909017e-06, + "loss": 1.4969, + "step": 1829 + }, + { + "epoch": 2.0054794520547947, + "grad_norm": 3.1297731399536133, + "learning_rate": 1.8744922826969944e-06, + "loss": 1.4857, + "step": 1830 + }, + { + "epoch": 2.0054794520547947, + "eval_loss": 1.770565390586853, + "eval_runtime": 44.0088, + "eval_samples_per_second": 22.723, + "eval_steps_per_second": 0.954, + "step": 1830 + }, + { + "epoch": 2.0065753424657533, + "grad_norm": 2.9419260025024414, + "learning_rate": 1.8724614134849716e-06, + "loss": 1.5027, + "step": 1831 + }, + { + "epoch": 2.0076712328767123, + "grad_norm": 2.9483847618103027, + "learning_rate": 1.870430544272949e-06, + "loss": 1.5937, + "step": 1832 + }, + { + "epoch": 2.0087671232876714, + "grad_norm": 3.0062618255615234, + "learning_rate": 1.8683996750609262e-06, + "loss": 1.491, + "step": 1833 + }, + { + "epoch": 2.00986301369863, + "grad_norm": 2.8788418769836426, + "learning_rate": 1.8663688058489035e-06, + "loss": 1.5111, + "step": 1834 + }, + { + "epoch": 2.010958904109589, + "grad_norm": 2.9052963256835938, + "learning_rate": 1.8643379366368807e-06, + "loss": 1.5174, + "step": 1835 + }, + { + "epoch": 2.012054794520548, + "grad_norm": 2.8395936489105225, + "learning_rate": 1.862307067424858e-06, + "loss": 1.5249, + "step": 1836 + }, + { + "epoch": 2.013150684931507, + "grad_norm": 3.0219051837921143, + "learning_rate": 1.8602761982128353e-06, + "loss": 1.5544, + "step": 1837 + }, + { + "epoch": 2.0142465753424657, + "grad_norm": 2.8336734771728516, + "learning_rate": 1.8582453290008126e-06, + "loss": 1.4943, + "step": 1838 + }, + { + "epoch": 2.0153424657534247, + "grad_norm": 3.003967046737671, + "learning_rate": 1.8562144597887898e-06, + "loss": 1.5156, + "step": 1839 + }, + { + "epoch": 2.0164383561643837, + "grad_norm": 2.9854485988616943, + "learning_rate": 1.854183590576767e-06, + "loss": 1.5589, + "step": 1840 + }, + { + "epoch": 2.0164383561643837, + "eval_loss": 1.7706938982009888, + "eval_runtime": 45.2999, + "eval_samples_per_second": 22.075, + "eval_steps_per_second": 0.927, + "step": 1840 + }, + { + "epoch": 2.0175342465753423, + "grad_norm": 2.933345079421997, + "learning_rate": 1.8521527213647442e-06, + "loss": 1.481, + "step": 1841 + }, + { + "epoch": 2.0186301369863013, + "grad_norm": 2.896928310394287, + "learning_rate": 1.8501218521527215e-06, + "loss": 1.4422, + "step": 1842 + }, + { + "epoch": 2.0197260273972604, + "grad_norm": 2.9579174518585205, + "learning_rate": 1.8480909829406987e-06, + "loss": 1.4867, + "step": 1843 + }, + { + "epoch": 2.020821917808219, + "grad_norm": 2.875720262527466, + "learning_rate": 1.846060113728676e-06, + "loss": 1.5221, + "step": 1844 + }, + { + "epoch": 2.021917808219178, + "grad_norm": 2.937431573867798, + "learning_rate": 1.8440292445166533e-06, + "loss": 1.4605, + "step": 1845 + }, + { + "epoch": 2.023013698630137, + "grad_norm": 2.9405369758605957, + "learning_rate": 1.8419983753046306e-06, + "loss": 1.5576, + "step": 1846 + }, + { + "epoch": 2.024109589041096, + "grad_norm": 2.9425089359283447, + "learning_rate": 1.8399675060926078e-06, + "loss": 1.5182, + "step": 1847 + }, + { + "epoch": 2.0252054794520546, + "grad_norm": 2.8655056953430176, + "learning_rate": 1.8379366368805851e-06, + "loss": 1.4949, + "step": 1848 + }, + { + "epoch": 2.0263013698630137, + "grad_norm": 2.8670871257781982, + "learning_rate": 1.8359057676685624e-06, + "loss": 1.486, + "step": 1849 + }, + { + "epoch": 2.0273972602739727, + "grad_norm": 2.9343950748443604, + "learning_rate": 1.8338748984565397e-06, + "loss": 1.4976, + "step": 1850 + }, + { + "epoch": 2.0273972602739727, + "eval_loss": 1.7705117464065552, + "eval_runtime": 47.3175, + "eval_samples_per_second": 21.134, + "eval_steps_per_second": 0.888, + "step": 1850 + }, + { + "epoch": 2.0284931506849313, + "grad_norm": 2.898251533508301, + "learning_rate": 1.831844029244517e-06, + "loss": 1.4714, + "step": 1851 + }, + { + "epoch": 2.0295890410958903, + "grad_norm": 2.8314175605773926, + "learning_rate": 1.829813160032494e-06, + "loss": 1.4865, + "step": 1852 + }, + { + "epoch": 2.0306849315068494, + "grad_norm": 3.0761542320251465, + "learning_rate": 1.8277822908204713e-06, + "loss": 1.5268, + "step": 1853 + }, + { + "epoch": 2.0317808219178084, + "grad_norm": 2.877169370651245, + "learning_rate": 1.8257514216084486e-06, + "loss": 1.5026, + "step": 1854 + }, + { + "epoch": 2.032876712328767, + "grad_norm": 2.954618453979492, + "learning_rate": 1.8237205523964258e-06, + "loss": 1.5078, + "step": 1855 + }, + { + "epoch": 2.033972602739726, + "grad_norm": 3.0036604404449463, + "learning_rate": 1.8216896831844031e-06, + "loss": 1.5461, + "step": 1856 + }, + { + "epoch": 2.035068493150685, + "grad_norm": 2.857048273086548, + "learning_rate": 1.8196588139723804e-06, + "loss": 1.4695, + "step": 1857 + }, + { + "epoch": 2.0361643835616436, + "grad_norm": 2.8631272315979004, + "learning_rate": 1.8176279447603577e-06, + "loss": 1.4735, + "step": 1858 + }, + { + "epoch": 2.0372602739726027, + "grad_norm": 2.895977735519409, + "learning_rate": 1.815597075548335e-06, + "loss": 1.5153, + "step": 1859 + }, + { + "epoch": 2.0383561643835617, + "grad_norm": 3.0079503059387207, + "learning_rate": 1.8135662063363122e-06, + "loss": 1.5402, + "step": 1860 + }, + { + "epoch": 2.0383561643835617, + "eval_loss": 1.770928978919983, + "eval_runtime": 46.3745, + "eval_samples_per_second": 21.564, + "eval_steps_per_second": 0.906, + "step": 1860 + }, + { + "epoch": 2.0394520547945207, + "grad_norm": 2.902636766433716, + "learning_rate": 1.8115353371242895e-06, + "loss": 1.439, + "step": 1861 + }, + { + "epoch": 2.0405479452054793, + "grad_norm": 2.9332778453826904, + "learning_rate": 1.8095044679122668e-06, + "loss": 1.481, + "step": 1862 + }, + { + "epoch": 2.0416438356164384, + "grad_norm": 2.922544479370117, + "learning_rate": 1.807473598700244e-06, + "loss": 1.47, + "step": 1863 + }, + { + "epoch": 2.0427397260273974, + "grad_norm": 2.9917609691619873, + "learning_rate": 1.805442729488221e-06, + "loss": 1.4861, + "step": 1864 + }, + { + "epoch": 2.043835616438356, + "grad_norm": 2.9225614070892334, + "learning_rate": 1.8034118602761982e-06, + "loss": 1.4959, + "step": 1865 + }, + { + "epoch": 2.044931506849315, + "grad_norm": 3.067898988723755, + "learning_rate": 1.8013809910641754e-06, + "loss": 1.5026, + "step": 1866 + }, + { + "epoch": 2.046027397260274, + "grad_norm": 2.971832752227783, + "learning_rate": 1.7993501218521527e-06, + "loss": 1.4676, + "step": 1867 + }, + { + "epoch": 2.047123287671233, + "grad_norm": 3.0682990550994873, + "learning_rate": 1.79731925264013e-06, + "loss": 1.5258, + "step": 1868 + }, + { + "epoch": 2.0482191780821917, + "grad_norm": 2.8473973274230957, + "learning_rate": 1.7952883834281073e-06, + "loss": 1.4772, + "step": 1869 + }, + { + "epoch": 2.0493150684931507, + "grad_norm": 2.86590838432312, + "learning_rate": 1.7932575142160846e-06, + "loss": 1.5319, + "step": 1870 + }, + { + "epoch": 2.0493150684931507, + "eval_loss": 1.7683560848236084, + "eval_runtime": 49.2497, + "eval_samples_per_second": 20.305, + "eval_steps_per_second": 0.853, + "step": 1870 + }, + { + "epoch": 2.0504109589041097, + "grad_norm": 3.0600473880767822, + "learning_rate": 1.7912266450040618e-06, + "loss": 1.5289, + "step": 1871 + }, + { + "epoch": 2.0515068493150683, + "grad_norm": 2.908846855163574, + "learning_rate": 1.789195775792039e-06, + "loss": 1.4469, + "step": 1872 + }, + { + "epoch": 2.0526027397260274, + "grad_norm": 3.080873489379883, + "learning_rate": 1.7871649065800164e-06, + "loss": 1.5531, + "step": 1873 + }, + { + "epoch": 2.0536986301369864, + "grad_norm": 2.9285154342651367, + "learning_rate": 1.7851340373679937e-06, + "loss": 1.4755, + "step": 1874 + }, + { + "epoch": 2.0547945205479454, + "grad_norm": 2.8980650901794434, + "learning_rate": 1.783103168155971e-06, + "loss": 1.4971, + "step": 1875 + }, + { + "epoch": 2.055890410958904, + "grad_norm": 2.875659942626953, + "learning_rate": 1.7810722989439482e-06, + "loss": 1.4934, + "step": 1876 + }, + { + "epoch": 2.056986301369863, + "grad_norm": 2.925330638885498, + "learning_rate": 1.7790414297319253e-06, + "loss": 1.4801, + "step": 1877 + }, + { + "epoch": 2.058082191780822, + "grad_norm": 2.844937801361084, + "learning_rate": 1.7770105605199025e-06, + "loss": 1.4709, + "step": 1878 + }, + { + "epoch": 2.0591780821917807, + "grad_norm": 2.938685655593872, + "learning_rate": 1.7749796913078798e-06, + "loss": 1.5025, + "step": 1879 + }, + { + "epoch": 2.0602739726027397, + "grad_norm": 3.1271140575408936, + "learning_rate": 1.772948822095857e-06, + "loss": 1.5504, + "step": 1880 + }, + { + "epoch": 2.0602739726027397, + "eval_loss": 1.769714593887329, + "eval_runtime": 47.8231, + "eval_samples_per_second": 20.91, + "eval_steps_per_second": 0.878, + "step": 1880 + }, + { + "epoch": 2.0613698630136987, + "grad_norm": 3.052131414413452, + "learning_rate": 1.7709179528838344e-06, + "loss": 1.5321, + "step": 1881 + }, + { + "epoch": 2.0624657534246573, + "grad_norm": 2.8789966106414795, + "learning_rate": 1.7688870836718116e-06, + "loss": 1.4956, + "step": 1882 + }, + { + "epoch": 2.0635616438356164, + "grad_norm": 2.887768030166626, + "learning_rate": 1.766856214459789e-06, + "loss": 1.492, + "step": 1883 + }, + { + "epoch": 2.0646575342465754, + "grad_norm": 2.8060758113861084, + "learning_rate": 1.7648253452477662e-06, + "loss": 1.4735, + "step": 1884 + }, + { + "epoch": 2.0657534246575344, + "grad_norm": 2.89400053024292, + "learning_rate": 1.7627944760357435e-06, + "loss": 1.5528, + "step": 1885 + }, + { + "epoch": 2.066849315068493, + "grad_norm": 2.9682323932647705, + "learning_rate": 1.7607636068237208e-06, + "loss": 1.5142, + "step": 1886 + }, + { + "epoch": 2.067945205479452, + "grad_norm": 2.8660190105438232, + "learning_rate": 1.758732737611698e-06, + "loss": 1.5137, + "step": 1887 + }, + { + "epoch": 2.069041095890411, + "grad_norm": 2.8982481956481934, + "learning_rate": 1.7567018683996753e-06, + "loss": 1.4972, + "step": 1888 + }, + { + "epoch": 2.0701369863013697, + "grad_norm": 2.8190507888793945, + "learning_rate": 1.7546709991876524e-06, + "loss": 1.4647, + "step": 1889 + }, + { + "epoch": 2.0712328767123287, + "grad_norm": 2.830960988998413, + "learning_rate": 1.7526401299756296e-06, + "loss": 1.4855, + "step": 1890 + }, + { + "epoch": 2.0712328767123287, + "eval_loss": 1.7688370943069458, + "eval_runtime": 46.7465, + "eval_samples_per_second": 21.392, + "eval_steps_per_second": 0.898, + "step": 1890 + }, + { + "epoch": 2.0723287671232877, + "grad_norm": 2.891460657119751, + "learning_rate": 1.750609260763607e-06, + "loss": 1.5053, + "step": 1891 + }, + { + "epoch": 2.0734246575342468, + "grad_norm": 2.9590890407562256, + "learning_rate": 1.7485783915515842e-06, + "loss": 1.4905, + "step": 1892 + }, + { + "epoch": 2.0745205479452054, + "grad_norm": 2.8448328971862793, + "learning_rate": 1.7465475223395615e-06, + "loss": 1.4483, + "step": 1893 + }, + { + "epoch": 2.0756164383561644, + "grad_norm": 3.167686700820923, + "learning_rate": 1.7445166531275387e-06, + "loss": 1.5786, + "step": 1894 + }, + { + "epoch": 2.0767123287671234, + "grad_norm": 3.040597915649414, + "learning_rate": 1.742485783915516e-06, + "loss": 1.5737, + "step": 1895 + }, + { + "epoch": 2.077808219178082, + "grad_norm": 2.9017035961151123, + "learning_rate": 1.7404549147034933e-06, + "loss": 1.4341, + "step": 1896 + }, + { + "epoch": 2.078904109589041, + "grad_norm": 3.0835607051849365, + "learning_rate": 1.7384240454914706e-06, + "loss": 1.5019, + "step": 1897 + }, + { + "epoch": 2.08, + "grad_norm": 3.01590633392334, + "learning_rate": 1.7363931762794478e-06, + "loss": 1.5231, + "step": 1898 + }, + { + "epoch": 2.081095890410959, + "grad_norm": 2.776460886001587, + "learning_rate": 1.7343623070674251e-06, + "loss": 1.4943, + "step": 1899 + }, + { + "epoch": 2.0821917808219177, + "grad_norm": 3.023834705352783, + "learning_rate": 1.7323314378554024e-06, + "loss": 1.5197, + "step": 1900 + }, + { + "epoch": 2.0821917808219177, + "eval_loss": 1.7668771743774414, + "eval_runtime": 45.0367, + "eval_samples_per_second": 22.204, + "eval_steps_per_second": 0.933, + "step": 1900 + }, + { + "epoch": 2.0832876712328767, + "grad_norm": 2.885643243789673, + "learning_rate": 1.7303005686433797e-06, + "loss": 1.5017, + "step": 1901 + }, + { + "epoch": 2.0843835616438358, + "grad_norm": 3.0691707134246826, + "learning_rate": 1.7282696994313567e-06, + "loss": 1.5033, + "step": 1902 + }, + { + "epoch": 2.0854794520547943, + "grad_norm": 2.8418657779693604, + "learning_rate": 1.726238830219334e-06, + "loss": 1.4954, + "step": 1903 + }, + { + "epoch": 2.0865753424657534, + "grad_norm": 2.9669976234436035, + "learning_rate": 1.7242079610073113e-06, + "loss": 1.5482, + "step": 1904 + }, + { + "epoch": 2.0876712328767124, + "grad_norm": 3.0346579551696777, + "learning_rate": 1.7221770917952886e-06, + "loss": 1.507, + "step": 1905 + }, + { + "epoch": 2.0887671232876714, + "grad_norm": 3.0112743377685547, + "learning_rate": 1.7201462225832658e-06, + "loss": 1.5197, + "step": 1906 + }, + { + "epoch": 2.08986301369863, + "grad_norm": 2.9679806232452393, + "learning_rate": 1.7181153533712431e-06, + "loss": 1.4265, + "step": 1907 + }, + { + "epoch": 2.090958904109589, + "grad_norm": 2.97170090675354, + "learning_rate": 1.7160844841592204e-06, + "loss": 1.4286, + "step": 1908 + }, + { + "epoch": 2.092054794520548, + "grad_norm": 2.9378933906555176, + "learning_rate": 1.7140536149471977e-06, + "loss": 1.4812, + "step": 1909 + }, + { + "epoch": 2.0931506849315067, + "grad_norm": 3.054729461669922, + "learning_rate": 1.712022745735175e-06, + "loss": 1.4993, + "step": 1910 + }, + { + "epoch": 2.0931506849315067, + "eval_loss": 1.7670947313308716, + "eval_runtime": 44.6325, + "eval_samples_per_second": 22.405, + "eval_steps_per_second": 0.941, + "step": 1910 + }, + { + "epoch": 2.0942465753424657, + "grad_norm": 3.0674521923065186, + "learning_rate": 1.7099918765231522e-06, + "loss": 1.495, + "step": 1911 + }, + { + "epoch": 2.0953424657534248, + "grad_norm": 2.848553419113159, + "learning_rate": 1.7079610073111295e-06, + "loss": 1.4556, + "step": 1912 + }, + { + "epoch": 2.0964383561643833, + "grad_norm": 2.9111032485961914, + "learning_rate": 1.7059301380991068e-06, + "loss": 1.5083, + "step": 1913 + }, + { + "epoch": 2.0975342465753424, + "grad_norm": 2.9358608722686768, + "learning_rate": 1.7038992688870836e-06, + "loss": 1.5028, + "step": 1914 + }, + { + "epoch": 2.0986301369863014, + "grad_norm": 2.965364933013916, + "learning_rate": 1.701868399675061e-06, + "loss": 1.5176, + "step": 1915 + }, + { + "epoch": 2.0997260273972604, + "grad_norm": 3.0749897956848145, + "learning_rate": 1.6998375304630382e-06, + "loss": 1.5682, + "step": 1916 + }, + { + "epoch": 2.100821917808219, + "grad_norm": 2.9876911640167236, + "learning_rate": 1.6978066612510155e-06, + "loss": 1.5129, + "step": 1917 + }, + { + "epoch": 2.101917808219178, + "grad_norm": 3.027039051055908, + "learning_rate": 1.6957757920389927e-06, + "loss": 1.4892, + "step": 1918 + }, + { + "epoch": 2.103013698630137, + "grad_norm": 2.8717336654663086, + "learning_rate": 1.69374492282697e-06, + "loss": 1.4716, + "step": 1919 + }, + { + "epoch": 2.1041095890410957, + "grad_norm": 3.011690616607666, + "learning_rate": 1.6917140536149473e-06, + "loss": 1.5576, + "step": 1920 + }, + { + "epoch": 2.1041095890410957, + "eval_loss": 1.7682307958602905, + "eval_runtime": 44.7041, + "eval_samples_per_second": 22.369, + "eval_steps_per_second": 0.94, + "step": 1920 + }, + { + "epoch": 2.1052054794520547, + "grad_norm": 2.858067035675049, + "learning_rate": 1.6896831844029246e-06, + "loss": 1.4797, + "step": 1921 + }, + { + "epoch": 2.1063013698630138, + "grad_norm": 2.9333767890930176, + "learning_rate": 1.6876523151909018e-06, + "loss": 1.5089, + "step": 1922 + }, + { + "epoch": 2.107397260273973, + "grad_norm": 2.996248483657837, + "learning_rate": 1.6856214459788791e-06, + "loss": 1.4719, + "step": 1923 + }, + { + "epoch": 2.1084931506849314, + "grad_norm": 3.006612539291382, + "learning_rate": 1.6835905767668564e-06, + "loss": 1.4827, + "step": 1924 + }, + { + "epoch": 2.1095890410958904, + "grad_norm": 2.9358556270599365, + "learning_rate": 1.6815597075548337e-06, + "loss": 1.5401, + "step": 1925 + }, + { + "epoch": 2.1106849315068494, + "grad_norm": 2.9943015575408936, + "learning_rate": 1.679528838342811e-06, + "loss": 1.4834, + "step": 1926 + }, + { + "epoch": 2.111780821917808, + "grad_norm": 2.850550889968872, + "learning_rate": 1.677497969130788e-06, + "loss": 1.4825, + "step": 1927 + }, + { + "epoch": 2.112876712328767, + "grad_norm": 3.026198148727417, + "learning_rate": 1.6754670999187653e-06, + "loss": 1.5069, + "step": 1928 + }, + { + "epoch": 2.113972602739726, + "grad_norm": 2.9569458961486816, + "learning_rate": 1.6734362307067426e-06, + "loss": 1.4499, + "step": 1929 + }, + { + "epoch": 2.115068493150685, + "grad_norm": 2.8670272827148438, + "learning_rate": 1.6714053614947198e-06, + "loss": 1.4289, + "step": 1930 + }, + { + "epoch": 2.115068493150685, + "eval_loss": 1.7686749696731567, + "eval_runtime": 47.0063, + "eval_samples_per_second": 21.274, + "eval_steps_per_second": 0.893, + "step": 1930 + }, + { + "epoch": 2.1161643835616437, + "grad_norm": 3.057095527648926, + "learning_rate": 1.669374492282697e-06, + "loss": 1.5302, + "step": 1931 + }, + { + "epoch": 2.1172602739726027, + "grad_norm": 3.1339385509490967, + "learning_rate": 1.6673436230706744e-06, + "loss": 1.5144, + "step": 1932 + }, + { + "epoch": 2.118356164383562, + "grad_norm": 2.980172872543335, + "learning_rate": 1.6653127538586517e-06, + "loss": 1.4901, + "step": 1933 + }, + { + "epoch": 2.1194520547945204, + "grad_norm": 2.9081473350524902, + "learning_rate": 1.663281884646629e-06, + "loss": 1.4866, + "step": 1934 + }, + { + "epoch": 2.1205479452054794, + "grad_norm": 2.963780403137207, + "learning_rate": 1.6612510154346062e-06, + "loss": 1.5503, + "step": 1935 + }, + { + "epoch": 2.1216438356164384, + "grad_norm": 3.2041895389556885, + "learning_rate": 1.6592201462225835e-06, + "loss": 1.5175, + "step": 1936 + }, + { + "epoch": 2.1227397260273975, + "grad_norm": 3.2994704246520996, + "learning_rate": 1.6571892770105608e-06, + "loss": 1.5509, + "step": 1937 + }, + { + "epoch": 2.123835616438356, + "grad_norm": 3.2698915004730225, + "learning_rate": 1.655158407798538e-06, + "loss": 1.4997, + "step": 1938 + }, + { + "epoch": 2.124931506849315, + "grad_norm": 2.9475982189178467, + "learning_rate": 1.653127538586515e-06, + "loss": 1.523, + "step": 1939 + }, + { + "epoch": 2.126027397260274, + "grad_norm": 2.901801824569702, + "learning_rate": 1.6510966693744924e-06, + "loss": 1.5401, + "step": 1940 + }, + { + "epoch": 2.126027397260274, + "eval_loss": 1.7699897289276123, + "eval_runtime": 48.7484, + "eval_samples_per_second": 20.514, + "eval_steps_per_second": 0.862, + "step": 1940 + }, + { + "epoch": 2.1271232876712327, + "grad_norm": 3.064338207244873, + "learning_rate": 1.6490658001624696e-06, + "loss": 1.5291, + "step": 1941 + }, + { + "epoch": 2.1282191780821917, + "grad_norm": 3.1033506393432617, + "learning_rate": 1.647034930950447e-06, + "loss": 1.4794, + "step": 1942 + }, + { + "epoch": 2.1293150684931508, + "grad_norm": 3.041693925857544, + "learning_rate": 1.6450040617384242e-06, + "loss": 1.5038, + "step": 1943 + }, + { + "epoch": 2.1304109589041094, + "grad_norm": 3.2308413982391357, + "learning_rate": 1.6429731925264015e-06, + "loss": 1.527, + "step": 1944 + }, + { + "epoch": 2.1315068493150684, + "grad_norm": 2.9788589477539062, + "learning_rate": 1.6409423233143788e-06, + "loss": 1.4852, + "step": 1945 + }, + { + "epoch": 2.1326027397260274, + "grad_norm": 2.8036675453186035, + "learning_rate": 1.638911454102356e-06, + "loss": 1.4811, + "step": 1946 + }, + { + "epoch": 2.1336986301369865, + "grad_norm": 2.80544376373291, + "learning_rate": 1.6368805848903333e-06, + "loss": 1.5276, + "step": 1947 + }, + { + "epoch": 2.134794520547945, + "grad_norm": 2.9585835933685303, + "learning_rate": 1.6348497156783106e-06, + "loss": 1.4817, + "step": 1948 + }, + { + "epoch": 2.135890410958904, + "grad_norm": 3.0835964679718018, + "learning_rate": 1.6328188464662879e-06, + "loss": 1.4888, + "step": 1949 + }, + { + "epoch": 2.136986301369863, + "grad_norm": 2.9466328620910645, + "learning_rate": 1.6307879772542651e-06, + "loss": 1.5099, + "step": 1950 + }, + { + "epoch": 2.136986301369863, + "eval_loss": 1.7669609785079956, + "eval_runtime": 47.0054, + "eval_samples_per_second": 21.274, + "eval_steps_per_second": 0.894, + "step": 1950 + }, + { + "epoch": 2.138082191780822, + "grad_norm": 2.984405040740967, + "learning_rate": 1.628757108042242e-06, + "loss": 1.5386, + "step": 1951 + }, + { + "epoch": 2.1391780821917807, + "grad_norm": 2.9967305660247803, + "learning_rate": 1.6267262388302193e-06, + "loss": 1.478, + "step": 1952 + }, + { + "epoch": 2.1402739726027398, + "grad_norm": 2.985135316848755, + "learning_rate": 1.6246953696181965e-06, + "loss": 1.4987, + "step": 1953 + }, + { + "epoch": 2.141369863013699, + "grad_norm": 2.9715938568115234, + "learning_rate": 1.6226645004061738e-06, + "loss": 1.5812, + "step": 1954 + }, + { + "epoch": 2.1424657534246574, + "grad_norm": 3.09761381149292, + "learning_rate": 1.620633631194151e-06, + "loss": 1.5401, + "step": 1955 + }, + { + "epoch": 2.1435616438356164, + "grad_norm": 2.969388008117676, + "learning_rate": 1.6186027619821284e-06, + "loss": 1.4886, + "step": 1956 + }, + { + "epoch": 2.1446575342465755, + "grad_norm": 2.9492249488830566, + "learning_rate": 1.6165718927701056e-06, + "loss": 1.5215, + "step": 1957 + }, + { + "epoch": 2.145753424657534, + "grad_norm": 2.909186840057373, + "learning_rate": 1.614541023558083e-06, + "loss": 1.5114, + "step": 1958 + }, + { + "epoch": 2.146849315068493, + "grad_norm": 3.070688247680664, + "learning_rate": 1.6125101543460602e-06, + "loss": 1.5016, + "step": 1959 + }, + { + "epoch": 2.147945205479452, + "grad_norm": 3.0163958072662354, + "learning_rate": 1.6104792851340375e-06, + "loss": 1.528, + "step": 1960 + }, + { + "epoch": 2.147945205479452, + "eval_loss": 1.7661935091018677, + "eval_runtime": 48.2729, + "eval_samples_per_second": 20.716, + "eval_steps_per_second": 0.87, + "step": 1960 + }, + { + "epoch": 2.149041095890411, + "grad_norm": 2.894883871078491, + "learning_rate": 1.6084484159220147e-06, + "loss": 1.483, + "step": 1961 + }, + { + "epoch": 2.1501369863013697, + "grad_norm": 2.8487191200256348, + "learning_rate": 1.606417546709992e-06, + "loss": 1.4889, + "step": 1962 + }, + { + "epoch": 2.1512328767123288, + "grad_norm": 2.929729700088501, + "learning_rate": 1.6043866774979693e-06, + "loss": 1.4976, + "step": 1963 + }, + { + "epoch": 2.152328767123288, + "grad_norm": 2.942971706390381, + "learning_rate": 1.6023558082859464e-06, + "loss": 1.4727, + "step": 1964 + }, + { + "epoch": 2.1534246575342464, + "grad_norm": 2.9945068359375, + "learning_rate": 1.6003249390739236e-06, + "loss": 1.5032, + "step": 1965 + }, + { + "epoch": 2.1545205479452054, + "grad_norm": 2.9212305545806885, + "learning_rate": 1.598294069861901e-06, + "loss": 1.4777, + "step": 1966 + }, + { + "epoch": 2.1556164383561645, + "grad_norm": 3.138329029083252, + "learning_rate": 1.5962632006498782e-06, + "loss": 1.5577, + "step": 1967 + }, + { + "epoch": 2.1567123287671235, + "grad_norm": 3.079740047454834, + "learning_rate": 1.5942323314378555e-06, + "loss": 1.465, + "step": 1968 + }, + { + "epoch": 2.157808219178082, + "grad_norm": 3.0435216426849365, + "learning_rate": 1.5922014622258327e-06, + "loss": 1.4641, + "step": 1969 + }, + { + "epoch": 2.158904109589041, + "grad_norm": 2.8816394805908203, + "learning_rate": 1.59017059301381e-06, + "loss": 1.5141, + "step": 1970 + }, + { + "epoch": 2.158904109589041, + "eval_loss": 1.7663989067077637, + "eval_runtime": 49.1495, + "eval_samples_per_second": 20.346, + "eval_steps_per_second": 0.855, + "step": 1970 + }, + { + "epoch": 2.16, + "grad_norm": 2.8964526653289795, + "learning_rate": 1.5881397238017873e-06, + "loss": 1.4978, + "step": 1971 + }, + { + "epoch": 2.1610958904109587, + "grad_norm": 2.9125728607177734, + "learning_rate": 1.5861088545897646e-06, + "loss": 1.4611, + "step": 1972 + }, + { + "epoch": 2.1621917808219178, + "grad_norm": 2.9918034076690674, + "learning_rate": 1.5840779853777418e-06, + "loss": 1.5017, + "step": 1973 + }, + { + "epoch": 2.163287671232877, + "grad_norm": 3.020989418029785, + "learning_rate": 1.5820471161657191e-06, + "loss": 1.4773, + "step": 1974 + }, + { + "epoch": 2.1643835616438354, + "grad_norm": 3.05708384513855, + "learning_rate": 1.5800162469536964e-06, + "loss": 1.5253, + "step": 1975 + }, + { + "epoch": 2.1654794520547944, + "grad_norm": 3.099055767059326, + "learning_rate": 1.5779853777416735e-06, + "loss": 1.5212, + "step": 1976 + }, + { + "epoch": 2.1665753424657535, + "grad_norm": 3.072061061859131, + "learning_rate": 1.5759545085296507e-06, + "loss": 1.5006, + "step": 1977 + }, + { + "epoch": 2.1676712328767125, + "grad_norm": 3.069732904434204, + "learning_rate": 1.573923639317628e-06, + "loss": 1.5016, + "step": 1978 + }, + { + "epoch": 2.168767123287671, + "grad_norm": 3.081413507461548, + "learning_rate": 1.5718927701056053e-06, + "loss": 1.5307, + "step": 1979 + }, + { + "epoch": 2.16986301369863, + "grad_norm": 2.9821321964263916, + "learning_rate": 1.5698619008935826e-06, + "loss": 1.5279, + "step": 1980 + }, + { + "epoch": 2.16986301369863, + "eval_loss": 1.7625439167022705, + "eval_runtime": 46.6457, + "eval_samples_per_second": 21.438, + "eval_steps_per_second": 0.9, + "step": 1980 + }, + { + "epoch": 2.170958904109589, + "grad_norm": 2.807304859161377, + "learning_rate": 1.5678310316815598e-06, + "loss": 1.4665, + "step": 1981 + }, + { + "epoch": 2.172054794520548, + "grad_norm": 3.033357620239258, + "learning_rate": 1.5658001624695371e-06, + "loss": 1.524, + "step": 1982 + }, + { + "epoch": 2.1731506849315068, + "grad_norm": 2.8620829582214355, + "learning_rate": 1.5637692932575144e-06, + "loss": 1.4889, + "step": 1983 + }, + { + "epoch": 2.174246575342466, + "grad_norm": 2.9625964164733887, + "learning_rate": 1.5617384240454917e-06, + "loss": 1.5098, + "step": 1984 + }, + { + "epoch": 2.175342465753425, + "grad_norm": 3.011361598968506, + "learning_rate": 1.559707554833469e-06, + "loss": 1.5342, + "step": 1985 + }, + { + "epoch": 2.1764383561643834, + "grad_norm": 2.950828790664673, + "learning_rate": 1.5576766856214462e-06, + "loss": 1.435, + "step": 1986 + }, + { + "epoch": 2.1775342465753424, + "grad_norm": 2.8914756774902344, + "learning_rate": 1.5556458164094235e-06, + "loss": 1.4756, + "step": 1987 + }, + { + "epoch": 2.1786301369863015, + "grad_norm": 2.8352556228637695, + "learning_rate": 1.5536149471974008e-06, + "loss": 1.4534, + "step": 1988 + }, + { + "epoch": 2.17972602739726, + "grad_norm": 2.9009487628936768, + "learning_rate": 1.5515840779853778e-06, + "loss": 1.4922, + "step": 1989 + }, + { + "epoch": 2.180821917808219, + "grad_norm": 3.05477237701416, + "learning_rate": 1.549553208773355e-06, + "loss": 1.4857, + "step": 1990 + }, + { + "epoch": 2.180821917808219, + "eval_loss": 1.766425371170044, + "eval_runtime": 51.2126, + "eval_samples_per_second": 19.526, + "eval_steps_per_second": 0.82, + "step": 1990 + }, + { + "epoch": 2.181917808219178, + "grad_norm": 2.9161787033081055, + "learning_rate": 1.5475223395613324e-06, + "loss": 1.4691, + "step": 1991 + }, + { + "epoch": 2.183013698630137, + "grad_norm": 3.073545455932617, + "learning_rate": 1.5454914703493097e-06, + "loss": 1.5333, + "step": 1992 + }, + { + "epoch": 2.1841095890410958, + "grad_norm": 3.025035858154297, + "learning_rate": 1.543460601137287e-06, + "loss": 1.4907, + "step": 1993 + }, + { + "epoch": 2.185205479452055, + "grad_norm": 3.068497896194458, + "learning_rate": 1.5414297319252642e-06, + "loss": 1.5087, + "step": 1994 + }, + { + "epoch": 2.186301369863014, + "grad_norm": 2.9628260135650635, + "learning_rate": 1.5393988627132415e-06, + "loss": 1.5039, + "step": 1995 + }, + { + "epoch": 2.1873972602739724, + "grad_norm": 2.9438061714172363, + "learning_rate": 1.5373679935012188e-06, + "loss": 1.4428, + "step": 1996 + }, + { + "epoch": 2.1884931506849314, + "grad_norm": 3.060307025909424, + "learning_rate": 1.535337124289196e-06, + "loss": 1.4867, + "step": 1997 + }, + { + "epoch": 2.1895890410958905, + "grad_norm": 3.1308751106262207, + "learning_rate": 1.5333062550771733e-06, + "loss": 1.5643, + "step": 1998 + }, + { + "epoch": 2.1906849315068495, + "grad_norm": 2.943803310394287, + "learning_rate": 1.5312753858651506e-06, + "loss": 1.513, + "step": 1999 + }, + { + "epoch": 2.191780821917808, + "grad_norm": 3.041837453842163, + "learning_rate": 1.5292445166531279e-06, + "loss": 1.5016, + "step": 2000 + }, + { + "epoch": 2.191780821917808, + "eval_loss": 1.7659627199172974, + "eval_runtime": 49.0365, + "eval_samples_per_second": 20.393, + "eval_steps_per_second": 0.857, + "step": 2000 + }, + { + "epoch": 2.192876712328767, + "grad_norm": 3.006897211074829, + "learning_rate": 1.5272136474411047e-06, + "loss": 1.528, + "step": 2001 + }, + { + "epoch": 2.193972602739726, + "grad_norm": 3.0074305534362793, + "learning_rate": 1.525182778229082e-06, + "loss": 1.5213, + "step": 2002 + }, + { + "epoch": 2.1950684931506848, + "grad_norm": 2.864231586456299, + "learning_rate": 1.5231519090170593e-06, + "loss": 1.4897, + "step": 2003 + }, + { + "epoch": 2.196164383561644, + "grad_norm": 3.0037953853607178, + "learning_rate": 1.5211210398050365e-06, + "loss": 1.4673, + "step": 2004 + }, + { + "epoch": 2.197260273972603, + "grad_norm": 2.880462169647217, + "learning_rate": 1.5190901705930138e-06, + "loss": 1.509, + "step": 2005 + }, + { + "epoch": 2.198356164383562, + "grad_norm": 2.9022490978240967, + "learning_rate": 1.517059301380991e-06, + "loss": 1.4818, + "step": 2006 + }, + { + "epoch": 2.1994520547945204, + "grad_norm": 2.902884006500244, + "learning_rate": 1.5150284321689684e-06, + "loss": 1.4891, + "step": 2007 + }, + { + "epoch": 2.2005479452054795, + "grad_norm": 2.992396831512451, + "learning_rate": 1.5129975629569456e-06, + "loss": 1.4821, + "step": 2008 + }, + { + "epoch": 2.2016438356164385, + "grad_norm": 3.0909762382507324, + "learning_rate": 1.510966693744923e-06, + "loss": 1.5049, + "step": 2009 + }, + { + "epoch": 2.202739726027397, + "grad_norm": 3.001403570175171, + "learning_rate": 1.5089358245329002e-06, + "loss": 1.4659, + "step": 2010 + }, + { + "epoch": 2.202739726027397, + "eval_loss": 1.7672089338302612, + "eval_runtime": 44.2628, + "eval_samples_per_second": 22.592, + "eval_steps_per_second": 0.949, + "step": 2010 + }, + { + "epoch": 2.203835616438356, + "grad_norm": 3.038501501083374, + "learning_rate": 1.5069049553208775e-06, + "loss": 1.4948, + "step": 2011 + }, + { + "epoch": 2.204931506849315, + "grad_norm": 2.9623525142669678, + "learning_rate": 1.5048740861088547e-06, + "loss": 1.5136, + "step": 2012 + }, + { + "epoch": 2.206027397260274, + "grad_norm": 2.9490935802459717, + "learning_rate": 1.502843216896832e-06, + "loss": 1.4399, + "step": 2013 + }, + { + "epoch": 2.207123287671233, + "grad_norm": 2.9608194828033447, + "learning_rate": 1.500812347684809e-06, + "loss": 1.4995, + "step": 2014 + }, + { + "epoch": 2.208219178082192, + "grad_norm": 2.920863628387451, + "learning_rate": 1.4987814784727864e-06, + "loss": 1.4773, + "step": 2015 + }, + { + "epoch": 2.209315068493151, + "grad_norm": 2.898723840713501, + "learning_rate": 1.4967506092607636e-06, + "loss": 1.4659, + "step": 2016 + }, + { + "epoch": 2.2104109589041094, + "grad_norm": 3.0445916652679443, + "learning_rate": 1.494719740048741e-06, + "loss": 1.5103, + "step": 2017 + }, + { + "epoch": 2.2115068493150685, + "grad_norm": 3.143130302429199, + "learning_rate": 1.4926888708367182e-06, + "loss": 1.5337, + "step": 2018 + }, + { + "epoch": 2.2126027397260275, + "grad_norm": 3.12093448638916, + "learning_rate": 1.4906580016246955e-06, + "loss": 1.5341, + "step": 2019 + }, + { + "epoch": 2.213698630136986, + "grad_norm": 3.3940253257751465, + "learning_rate": 1.4886271324126727e-06, + "loss": 1.5457, + "step": 2020 + }, + { + "epoch": 2.213698630136986, + "eval_loss": 1.7669864892959595, + "eval_runtime": 49.4214, + "eval_samples_per_second": 20.234, + "eval_steps_per_second": 0.85, + "step": 2020 + }, + { + "epoch": 2.214794520547945, + "grad_norm": 2.9350290298461914, + "learning_rate": 1.48659626320065e-06, + "loss": 1.4216, + "step": 2021 + }, + { + "epoch": 2.215890410958904, + "grad_norm": 2.9831831455230713, + "learning_rate": 1.4845653939886273e-06, + "loss": 1.4389, + "step": 2022 + }, + { + "epoch": 2.216986301369863, + "grad_norm": 3.0144903659820557, + "learning_rate": 1.4825345247766046e-06, + "loss": 1.4437, + "step": 2023 + }, + { + "epoch": 2.2180821917808218, + "grad_norm": 3.202122449874878, + "learning_rate": 1.4805036555645818e-06, + "loss": 1.4842, + "step": 2024 + }, + { + "epoch": 2.219178082191781, + "grad_norm": 2.955124855041504, + "learning_rate": 1.4784727863525591e-06, + "loss": 1.4517, + "step": 2025 + }, + { + "epoch": 2.22027397260274, + "grad_norm": 2.9716339111328125, + "learning_rate": 1.4764419171405362e-06, + "loss": 1.5262, + "step": 2026 + }, + { + "epoch": 2.2213698630136984, + "grad_norm": 3.0919816493988037, + "learning_rate": 1.4744110479285135e-06, + "loss": 1.5283, + "step": 2027 + }, + { + "epoch": 2.2224657534246575, + "grad_norm": 2.9478485584259033, + "learning_rate": 1.4723801787164907e-06, + "loss": 1.5193, + "step": 2028 + }, + { + "epoch": 2.2235616438356165, + "grad_norm": 2.9672513008117676, + "learning_rate": 1.470349309504468e-06, + "loss": 1.5087, + "step": 2029 + }, + { + "epoch": 2.2246575342465755, + "grad_norm": 3.0786097049713135, + "learning_rate": 1.4683184402924453e-06, + "loss": 1.5374, + "step": 2030 + }, + { + "epoch": 2.2246575342465755, + "eval_loss": 1.7661728858947754, + "eval_runtime": 45.7712, + "eval_samples_per_second": 21.848, + "eval_steps_per_second": 0.918, + "step": 2030 + }, + { + "epoch": 2.225753424657534, + "grad_norm": 3.043802499771118, + "learning_rate": 1.4662875710804226e-06, + "loss": 1.5623, + "step": 2031 + }, + { + "epoch": 2.226849315068493, + "grad_norm": 2.9582507610321045, + "learning_rate": 1.4642567018683998e-06, + "loss": 1.4836, + "step": 2032 + }, + { + "epoch": 2.227945205479452, + "grad_norm": 2.937865734100342, + "learning_rate": 1.4622258326563771e-06, + "loss": 1.4809, + "step": 2033 + }, + { + "epoch": 2.2290410958904108, + "grad_norm": 3.0230515003204346, + "learning_rate": 1.4601949634443544e-06, + "loss": 1.4854, + "step": 2034 + }, + { + "epoch": 2.23013698630137, + "grad_norm": 3.106809139251709, + "learning_rate": 1.4581640942323317e-06, + "loss": 1.574, + "step": 2035 + }, + { + "epoch": 2.231232876712329, + "grad_norm": 2.911781072616577, + "learning_rate": 1.456133225020309e-06, + "loss": 1.5024, + "step": 2036 + }, + { + "epoch": 2.232328767123288, + "grad_norm": 3.054163694381714, + "learning_rate": 1.4541023558082862e-06, + "loss": 1.4918, + "step": 2037 + }, + { + "epoch": 2.2334246575342465, + "grad_norm": 3.1683154106140137, + "learning_rate": 1.4520714865962635e-06, + "loss": 1.5328, + "step": 2038 + }, + { + "epoch": 2.2345205479452055, + "grad_norm": 3.079185724258423, + "learning_rate": 1.4500406173842406e-06, + "loss": 1.5, + "step": 2039 + }, + { + "epoch": 2.2356164383561645, + "grad_norm": 2.98567271232605, + "learning_rate": 1.4480097481722178e-06, + "loss": 1.5124, + "step": 2040 + }, + { + "epoch": 2.2356164383561645, + "eval_loss": 1.7658758163452148, + "eval_runtime": 46.8933, + "eval_samples_per_second": 21.325, + "eval_steps_per_second": 0.896, + "step": 2040 + }, + { + "epoch": 2.236712328767123, + "grad_norm": 3.0830793380737305, + "learning_rate": 1.4459788789601951e-06, + "loss": 1.4775, + "step": 2041 + }, + { + "epoch": 2.237808219178082, + "grad_norm": 3.052293539047241, + "learning_rate": 1.4439480097481724e-06, + "loss": 1.4644, + "step": 2042 + }, + { + "epoch": 2.238904109589041, + "grad_norm": 3.1189992427825928, + "learning_rate": 1.4419171405361497e-06, + "loss": 1.5087, + "step": 2043 + }, + { + "epoch": 2.24, + "grad_norm": 2.9710934162139893, + "learning_rate": 1.439886271324127e-06, + "loss": 1.5125, + "step": 2044 + }, + { + "epoch": 2.241095890410959, + "grad_norm": 3.050168037414551, + "learning_rate": 1.4378554021121042e-06, + "loss": 1.493, + "step": 2045 + }, + { + "epoch": 2.242191780821918, + "grad_norm": 3.186884880065918, + "learning_rate": 1.4358245329000815e-06, + "loss": 1.5207, + "step": 2046 + }, + { + "epoch": 2.243287671232877, + "grad_norm": 2.940720319747925, + "learning_rate": 1.4337936636880588e-06, + "loss": 1.5205, + "step": 2047 + }, + { + "epoch": 2.2443835616438355, + "grad_norm": 3.028677225112915, + "learning_rate": 1.431762794476036e-06, + "loss": 1.4579, + "step": 2048 + }, + { + "epoch": 2.2454794520547945, + "grad_norm": 3.199021577835083, + "learning_rate": 1.4297319252640133e-06, + "loss": 1.4798, + "step": 2049 + }, + { + "epoch": 2.2465753424657535, + "grad_norm": 2.9849867820739746, + "learning_rate": 1.4277010560519906e-06, + "loss": 1.4747, + "step": 2050 + }, + { + "epoch": 2.2465753424657535, + "eval_loss": 1.7669358253479004, + "eval_runtime": 45.2867, + "eval_samples_per_second": 22.082, + "eval_steps_per_second": 0.927, + "step": 2050 + }, + { + "epoch": 2.247671232876712, + "grad_norm": 3.1557536125183105, + "learning_rate": 1.4256701868399674e-06, + "loss": 1.5369, + "step": 2051 + }, + { + "epoch": 2.248767123287671, + "grad_norm": 3.2142200469970703, + "learning_rate": 1.4236393176279447e-06, + "loss": 1.5014, + "step": 2052 + }, + { + "epoch": 2.24986301369863, + "grad_norm": 3.184725522994995, + "learning_rate": 1.421608448415922e-06, + "loss": 1.5707, + "step": 2053 + }, + { + "epoch": 2.250958904109589, + "grad_norm": 2.9449050426483154, + "learning_rate": 1.4195775792038993e-06, + "loss": 1.4718, + "step": 2054 + }, + { + "epoch": 2.252054794520548, + "grad_norm": 3.0560460090637207, + "learning_rate": 1.4175467099918765e-06, + "loss": 1.5043, + "step": 2055 + }, + { + "epoch": 2.253150684931507, + "grad_norm": 3.032956123352051, + "learning_rate": 1.4155158407798538e-06, + "loss": 1.5233, + "step": 2056 + }, + { + "epoch": 2.254246575342466, + "grad_norm": 3.041759490966797, + "learning_rate": 1.413484971567831e-06, + "loss": 1.4327, + "step": 2057 + }, + { + "epoch": 2.255342465753425, + "grad_norm": 3.0571510791778564, + "learning_rate": 1.4114541023558084e-06, + "loss": 1.4789, + "step": 2058 + }, + { + "epoch": 2.2564383561643835, + "grad_norm": 2.9762444496154785, + "learning_rate": 1.4094232331437856e-06, + "loss": 1.5244, + "step": 2059 + }, + { + "epoch": 2.2575342465753425, + "grad_norm": 3.0351779460906982, + "learning_rate": 1.407392363931763e-06, + "loss": 1.5731, + "step": 2060 + }, + { + "epoch": 2.2575342465753425, + "eval_loss": 1.768678903579712, + "eval_runtime": 44.6047, + "eval_samples_per_second": 22.419, + "eval_steps_per_second": 0.942, + "step": 2060 + }, + { + "epoch": 2.2586301369863016, + "grad_norm": 3.267486810684204, + "learning_rate": 1.4053614947197402e-06, + "loss": 1.4895, + "step": 2061 + }, + { + "epoch": 2.25972602739726, + "grad_norm": 2.893648862838745, + "learning_rate": 1.4033306255077175e-06, + "loss": 1.4383, + "step": 2062 + }, + { + "epoch": 2.260821917808219, + "grad_norm": 3.103024482727051, + "learning_rate": 1.4012997562956945e-06, + "loss": 1.5395, + "step": 2063 + }, + { + "epoch": 2.261917808219178, + "grad_norm": 3.0002236366271973, + "learning_rate": 1.3992688870836718e-06, + "loss": 1.5246, + "step": 2064 + }, + { + "epoch": 2.263013698630137, + "grad_norm": 3.029136896133423, + "learning_rate": 1.397238017871649e-06, + "loss": 1.5111, + "step": 2065 + }, + { + "epoch": 2.264109589041096, + "grad_norm": 3.0597546100616455, + "learning_rate": 1.3952071486596264e-06, + "loss": 1.4709, + "step": 2066 + }, + { + "epoch": 2.265205479452055, + "grad_norm": 2.8829214572906494, + "learning_rate": 1.3931762794476036e-06, + "loss": 1.5387, + "step": 2067 + }, + { + "epoch": 2.266301369863014, + "grad_norm": 3.011145830154419, + "learning_rate": 1.391145410235581e-06, + "loss": 1.4851, + "step": 2068 + }, + { + "epoch": 2.2673972602739725, + "grad_norm": 2.9189581871032715, + "learning_rate": 1.3891145410235582e-06, + "loss": 1.4108, + "step": 2069 + }, + { + "epoch": 2.2684931506849315, + "grad_norm": 2.9735352993011475, + "learning_rate": 1.3870836718115355e-06, + "loss": 1.497, + "step": 2070 + }, + { + "epoch": 2.2684931506849315, + "eval_loss": 1.7667678594589233, + "eval_runtime": 44.089, + "eval_samples_per_second": 22.681, + "eval_steps_per_second": 0.953, + "step": 2070 + }, + { + "epoch": 2.2695890410958905, + "grad_norm": 2.932903289794922, + "learning_rate": 1.3850528025995127e-06, + "loss": 1.4881, + "step": 2071 + }, + { + "epoch": 2.270684931506849, + "grad_norm": 2.925725221633911, + "learning_rate": 1.38302193338749e-06, + "loss": 1.4059, + "step": 2072 + }, + { + "epoch": 2.271780821917808, + "grad_norm": 3.072967290878296, + "learning_rate": 1.3809910641754673e-06, + "loss": 1.5202, + "step": 2073 + }, + { + "epoch": 2.272876712328767, + "grad_norm": 3.1604013442993164, + "learning_rate": 1.3789601949634446e-06, + "loss": 1.4577, + "step": 2074 + }, + { + "epoch": 2.2739726027397262, + "grad_norm": 3.0084853172302246, + "learning_rate": 1.3769293257514218e-06, + "loss": 1.4996, + "step": 2075 + }, + { + "epoch": 2.275068493150685, + "grad_norm": 2.9525721073150635, + "learning_rate": 1.374898456539399e-06, + "loss": 1.4543, + "step": 2076 + }, + { + "epoch": 2.276164383561644, + "grad_norm": 3.019083261489868, + "learning_rate": 1.3728675873273762e-06, + "loss": 1.5305, + "step": 2077 + }, + { + "epoch": 2.277260273972603, + "grad_norm": 3.2680931091308594, + "learning_rate": 1.3708367181153535e-06, + "loss": 1.5043, + "step": 2078 + }, + { + "epoch": 2.2783561643835615, + "grad_norm": 3.021989583969116, + "learning_rate": 1.3688058489033307e-06, + "loss": 1.5024, + "step": 2079 + }, + { + "epoch": 2.2794520547945205, + "grad_norm": 3.033827066421509, + "learning_rate": 1.366774979691308e-06, + "loss": 1.5203, + "step": 2080 + }, + { + "epoch": 2.2794520547945205, + "eval_loss": 1.7672016620635986, + "eval_runtime": 42.675, + "eval_samples_per_second": 23.433, + "eval_steps_per_second": 0.984, + "step": 2080 + }, + { + "epoch": 2.2805479452054795, + "grad_norm": 3.1422817707061768, + "learning_rate": 1.3647441104792853e-06, + "loss": 1.497, + "step": 2081 + }, + { + "epoch": 2.281643835616438, + "grad_norm": 3.00771427154541, + "learning_rate": 1.3627132412672626e-06, + "loss": 1.4547, + "step": 2082 + }, + { + "epoch": 2.282739726027397, + "grad_norm": 2.9592597484588623, + "learning_rate": 1.3606823720552398e-06, + "loss": 1.4977, + "step": 2083 + }, + { + "epoch": 2.283835616438356, + "grad_norm": 2.9128124713897705, + "learning_rate": 1.3586515028432171e-06, + "loss": 1.4454, + "step": 2084 + }, + { + "epoch": 2.2849315068493152, + "grad_norm": 2.852811574935913, + "learning_rate": 1.3566206336311944e-06, + "loss": 1.4915, + "step": 2085 + }, + { + "epoch": 2.286027397260274, + "grad_norm": 2.8820016384124756, + "learning_rate": 1.3545897644191717e-06, + "loss": 1.4803, + "step": 2086 + }, + { + "epoch": 2.287123287671233, + "grad_norm": 3.0272696018218994, + "learning_rate": 1.352558895207149e-06, + "loss": 1.5282, + "step": 2087 + }, + { + "epoch": 2.288219178082192, + "grad_norm": 2.969461679458618, + "learning_rate": 1.3505280259951258e-06, + "loss": 1.493, + "step": 2088 + }, + { + "epoch": 2.289315068493151, + "grad_norm": 3.001244306564331, + "learning_rate": 1.348497156783103e-06, + "loss": 1.5378, + "step": 2089 + }, + { + "epoch": 2.2904109589041095, + "grad_norm": 2.952277660369873, + "learning_rate": 1.3464662875710804e-06, + "loss": 1.465, + "step": 2090 + }, + { + "epoch": 2.2904109589041095, + "eval_loss": 1.7604496479034424, + "eval_runtime": 42.7902, + "eval_samples_per_second": 23.37, + "eval_steps_per_second": 0.982, + "step": 2090 + }, + { + "epoch": 2.2915068493150685, + "grad_norm": 2.944876194000244, + "learning_rate": 1.3444354183590576e-06, + "loss": 1.5244, + "step": 2091 + }, + { + "epoch": 2.2926027397260276, + "grad_norm": 2.93131422996521, + "learning_rate": 1.342404549147035e-06, + "loss": 1.4824, + "step": 2092 + }, + { + "epoch": 2.293698630136986, + "grad_norm": 3.04500675201416, + "learning_rate": 1.3403736799350122e-06, + "loss": 1.5019, + "step": 2093 + }, + { + "epoch": 2.294794520547945, + "grad_norm": 3.0568926334381104, + "learning_rate": 1.3383428107229895e-06, + "loss": 1.5274, + "step": 2094 + }, + { + "epoch": 2.2958904109589042, + "grad_norm": 3.0788938999176025, + "learning_rate": 1.3363119415109667e-06, + "loss": 1.5238, + "step": 2095 + }, + { + "epoch": 2.296986301369863, + "grad_norm": 2.9863524436950684, + "learning_rate": 1.334281072298944e-06, + "loss": 1.513, + "step": 2096 + }, + { + "epoch": 2.298082191780822, + "grad_norm": 2.9421253204345703, + "learning_rate": 1.3322502030869213e-06, + "loss": 1.4923, + "step": 2097 + }, + { + "epoch": 2.299178082191781, + "grad_norm": 3.0319464206695557, + "learning_rate": 1.3302193338748986e-06, + "loss": 1.5295, + "step": 2098 + }, + { + "epoch": 2.30027397260274, + "grad_norm": 2.970057249069214, + "learning_rate": 1.3281884646628758e-06, + "loss": 1.5139, + "step": 2099 + }, + { + "epoch": 2.3013698630136985, + "grad_norm": 2.9552175998687744, + "learning_rate": 1.3261575954508531e-06, + "loss": 1.4988, + "step": 2100 + }, + { + "epoch": 2.3013698630136985, + "eval_loss": 1.7599544525146484, + "eval_runtime": 43.9278, + "eval_samples_per_second": 22.765, + "eval_steps_per_second": 0.956, + "step": 2100 + } + ], + "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": 6.546349759660032e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2100/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-2200/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2200/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-2400/rng_state_0.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..7cebe108f7c295f538afca780f8bff98aa6232f2 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c7105b68cc7706a616e5184c049404c7551d0ad422b5b1a292aa19facaa3d0f +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/rng_state_1.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..618e221236c3debbaad7ada8aebf8a3335d91f6e --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a994ddb08d50eccc51680080b7c8b587407a3900fb66b09b933d12402a0da4b +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/rng_state_2.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..f378850538621142c5cd15d9d6f50bc9a0f23b02 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0fd0c5d9fbbbbb845dde3ee7d26db4ac79f6c6261f597d751fb6b57476387a4b +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/scheduler.pt b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f69433a32ae7a51ab2e602cee3669921e616cc78 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74f626111a5ad05a79e900ea0a6b60ce2c34064434b5e9605e0bc5ff18dbb0d1 +size 627 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/tokenizer.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..309789d5ccd06f418fbcfb7472a6a9affda68101 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/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-2400/training_args.bin b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..6db59a81c34edebb00ef6a59f5669f3babfbb622 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-2400/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-300/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/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_local3_10M_seed0/runs/checkpoint-300/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/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_local3_10M_seed0/runs/checkpoint-300/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/latest new file mode 100644 index 0000000000000000000000000000000000000000..6761b575fffac7f1984044dcb6446b3a51da04c8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/latest @@ -0,0 +1 @@ +global_step300 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/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_local3_10M_seed0/runs/checkpoint-300/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/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_local3_10M_seed0/runs/checkpoint-300/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/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_local3_10M_seed0/runs/checkpoint-300/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..47601d2fd30543c3d78da6a2ba05864ac5a381b8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/trainer_state.json @@ -0,0 +1,2373 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.3287671232876712, + "eval_steps": 10, + "global_step": 300, + "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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.7659518718719482, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.279, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.433785915374756, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.3368, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 3.1486687660217285, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3896, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.391956090927124, + "learning_rate": 1.624087591240876e-06, + "loss": 2.297, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.6649584770202637, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.2415, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.5604891777038574, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2688, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.47900652885437, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2933, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.632723569869995, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.2143, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.8085947036743164, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2748, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.3927979469299316, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2685, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.355207920074463, + "eval_runtime": 42.3505, + "eval_samples_per_second": 23.612, + "eval_steps_per_second": 0.992, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.6170592308044434, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.3121, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.699814796447754, + "learning_rate": 1.77007299270073e-06, + "loss": 2.2853, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.3450610637664795, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3297, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.458219528198242, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2639, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.7101449966430664, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2134, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.5734996795654297, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2347, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 2.5200352668762207, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2299, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5022480487823486, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2253, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.190117359161377, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.2272, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.414524555206299, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2424, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.317634344100952, + "eval_runtime": 43.5742, + "eval_samples_per_second": 22.949, + "eval_steps_per_second": 0.964, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.3226804733276367, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1449, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.733006477355957, + "learning_rate": 1.952554744525548e-06, + "loss": 2.2027, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 2.5586233139038086, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2007, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.669914722442627, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2024, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.6148102283477783, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2496, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 2.331369638442993, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1814, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.3051886558532715, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2015, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.6890218257904053, + "learning_rate": 2.062043795620438e-06, + "loss": 2.184, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.4975883960723877, + "learning_rate": 2.08029197080292e-06, + "loss": 2.2439, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.5848891735076904, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1413, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.280966281890869, + "eval_runtime": 43.1001, + "eval_samples_per_second": 23.202, + "eval_steps_per_second": 0.974, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.2104, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.3356540203094482, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2312, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 2.3616602420806885, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.1399, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.581623077392578, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1456, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.547288656234741, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.184, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6386160850524902, + "learning_rate": 2.208029197080292e-06, + "loss": 2.1658, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.593065023422241, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1443, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.530632972717285, + "learning_rate": 2.244525547445256e-06, + "loss": 2.194, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.3553175926208496, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.1918, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 2.7536532878875732, + "learning_rate": 2.281021897810219e-06, + "loss": 2.1477, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2608063220977783, + "eval_runtime": 42.7938, + "eval_samples_per_second": 23.368, + "eval_steps_per_second": 0.981, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.160094738006592, + "learning_rate": 2.299270072992701e-06, + "loss": 2.133, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.4474735260009766, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1468, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.6654434204101562, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.1194, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.726611375808716, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1518, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 2.3560168743133545, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1537, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.664065361022949, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0888, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.3946099281311035, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.0998, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.7812652587890625, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1231, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.64595890045166, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1746, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.788524866104126, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0731, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.2340502738952637, + "eval_runtime": 42.3104, + "eval_samples_per_second": 23.635, + "eval_steps_per_second": 0.993, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 2.9249846935272217, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0564, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.5507798194885254, + "learning_rate": 2.5e-06, + "loss": 2.1714, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.1604716777801514, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0701, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.5429840087890625, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.0662, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8296737670898438, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.1378, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.9396755695343018, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1069, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.736814022064209, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1076, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 2.774909496307373, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1286, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.0861318111419678, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1499, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.624016284942627, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0989, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.215052604675293, + "eval_runtime": 43.9956, + "eval_samples_per_second": 22.73, + "eval_steps_per_second": 0.955, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.801500082015991, + "learning_rate": 2.664233576642336e-06, + "loss": 2.0795, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 2.2107114791870117, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.1035, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.030381441116333, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.0794, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 2.9719064235687256, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1496, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.688839912414551, + "learning_rate": 2.737226277372263e-06, + "loss": 2.1152, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 2.9350979328155518, + "learning_rate": 2.755474452554745e-06, + "loss": 2.0773, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.8958046436309814, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.1805, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.0781922340393066, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.1249, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.032059907913208, + "learning_rate": 2.81021897810219e-06, + "loss": 2.0632, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.798758029937744, + "learning_rate": 2.828467153284672e-06, + "loss": 2.1001, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2005105018615723, + "eval_runtime": 43.5111, + "eval_samples_per_second": 22.983, + "eval_steps_per_second": 0.965, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.7674927711486816, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.0211, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 2.6168575286865234, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0599, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 2.933353900909424, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.0169, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.81402587890625, + "learning_rate": 2.901459854014599e-06, + "loss": 2.122, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.5592572689056396, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.0957, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.738271951675415, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.0868, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.5362703800201416, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1237, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.6304945945739746, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.0825, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.040717601776123, + "learning_rate": 2.992700729927008e-06, + "loss": 2.0388, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.6226463317871094, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0163, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.186411142349243, + "eval_runtime": 44.625, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.5307068824768066, + "learning_rate": 3.029197080291971e-06, + "loss": 2.1097, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 2.9160444736480713, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.0946, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.5250024795532227, + "learning_rate": 3.065693430656935e-06, + "loss": 2.0709, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.6963064670562744, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.0937, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 2.845707416534424, + "learning_rate": 3.102189781021898e-06, + "loss": 2.0851, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.8985090255737305, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.0176, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 2.734377861022949, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.9931, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.5125555992126465, + "learning_rate": 3.156934306569343e-06, + "loss": 2.1133, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.0393502712249756, + "learning_rate": 3.175182481751825e-06, + "loss": 2.0636, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.6410164833068848, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0304, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.171868085861206, + "eval_runtime": 44.2249, + "eval_samples_per_second": 22.612, + "eval_steps_per_second": 0.95, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.7937498092651367, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0518, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7801201343536377, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.026, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7963638305664062, + "learning_rate": 3.248175182481752e-06, + "loss": 1.9642, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6330482959747314, + "learning_rate": 3.266423357664234e-06, + "loss": 2.0545, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.9771625995635986, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0445, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.6112730503082275, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0616, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.732865333557129, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0403, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.9113688468933105, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0382, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.4516773223876953, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.0416, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.781735897064209, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0348, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.160499334335327, + "eval_runtime": 44.5652, + "eval_samples_per_second": 22.439, + "eval_steps_per_second": 0.942, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.9990546703338623, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0324, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.783071279525757, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0518, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.659040689468384, + "learning_rate": 3.43065693430657e-06, + "loss": 1.9722, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 2.799798011779785, + "learning_rate": 3.4489051094890513e-06, + "loss": 2.0362, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 3.025458812713623, + "learning_rate": 3.467153284671533e-06, + "loss": 1.9863, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 2.686415195465088, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0201, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.716064929962158, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0249, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.637653350830078, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0403, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 2.7686848640441895, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0626, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.721956729888916, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.1077, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.152338981628418, + "eval_runtime": 43.5455, + "eval_samples_per_second": 22.964, + "eval_steps_per_second": 0.965, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.92571759223938, + "learning_rate": 3.576642335766424e-06, + "loss": 2.0928, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.9189202785491943, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9315, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 2.990233898162842, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9427, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.702348232269287, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9768, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6949572563171387, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.9954, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.108405113220215, + "learning_rate": 3.667883211678833e-06, + "loss": 1.9631, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.738922119140625, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9644, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.7419424057006836, + "learning_rate": 3.704379562043796e-06, + "loss": 2.0675, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9853813648223877, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0245, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.4614126682281494, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.9665, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.1342060565948486, + "eval_runtime": 46.3954, + "eval_samples_per_second": 21.554, + "eval_steps_per_second": 0.905, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.6787819862365723, + "learning_rate": 3.7591240875912412e-06, + "loss": 2.0063, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 2.563525438308716, + "learning_rate": 3.777372262773723e-06, + "loss": 1.998, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.4981701374053955, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.9738, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 2.5357978343963623, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0221, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.5365488529205322, + "learning_rate": 3.832116788321168e-06, + "loss": 1.9664, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 2.669468641281128, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9577, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 2.6950953006744385, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9786, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8249378204345703, + "learning_rate": 3.886861313868613e-06, + "loss": 2.0041, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.5924317836761475, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0338, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.1282334327697754, + "learning_rate": 3.923357664233577e-06, + "loss": 1.9698, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.1221420764923096, + "eval_runtime": 47.1127, + "eval_samples_per_second": 21.226, + "eval_steps_per_second": 0.891, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.7878193855285645, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0446, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.7137255668640137, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0575, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.286421060562134, + "learning_rate": 3.978102189781022e-06, + "loss": 1.9131, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.5466175079345703, + "learning_rate": 3.996350364963504e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.792483329772949, + "learning_rate": 4.014598540145986e-06, + "loss": 1.912, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.7533867359161377, + "learning_rate": 4.032846715328467e-06, + "loss": 1.9742, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 2.848405361175537, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9815, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.1034414768218994, + "learning_rate": 4.069343065693431e-06, + "loss": 1.9409, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.5109779834747314, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9676, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.486506938934326, + "learning_rate": 4.105839416058395e-06, + "loss": 2.0353, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1102187633514404, + "eval_runtime": 45.384, + "eval_samples_per_second": 22.034, + "eval_steps_per_second": 0.925, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.419269323348999, + "learning_rate": 4.124087591240876e-06, + "loss": 1.9622, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.4967262744903564, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.9523, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 2.7311460971832275, + "learning_rate": 4.16058394160584e-06, + "loss": 2.0234, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.241004467010498, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.907, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.821537733078003, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9761, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 2.5711426734924316, + "learning_rate": 4.215328467153285e-06, + "loss": 1.9411, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.936594009399414, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9432, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.6091108322143555, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9545, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.522674798965454, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9246, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.984649181365967, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9417, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1031947135925293, + "eval_runtime": 47.5409, + "eval_samples_per_second": 21.035, + "eval_steps_per_second": 0.883, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.8250324726104736, + "learning_rate": 4.306569343065693e-06, + "loss": 1.9061, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.7654290199279785, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9653, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.7091264724731445, + "learning_rate": 4.343065693430658e-06, + "loss": 1.9482, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 2.8019583225250244, + "learning_rate": 4.361313868613139e-06, + "loss": 1.9332, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.147368907928467, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9836, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.7050259113311768, + "learning_rate": 4.397810218978102e-06, + "loss": 1.9607, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.738743543624878, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9374, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.122777223587036, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9249, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.850745439529419, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9707, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.391998291015625, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.9703, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.0887911319732666, + "eval_runtime": 48.5149, + "eval_samples_per_second": 20.612, + "eval_steps_per_second": 0.866, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 2.9131698608398438, + "learning_rate": 4.489051094890512e-06, + "loss": 1.9989, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.1045937538146973, + "learning_rate": 4.507299270072993e-06, + "loss": 1.9718, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.7984158992767334, + "learning_rate": 4.525547445255475e-06, + "loss": 1.947, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.6223504543304443, + "learning_rate": 4.543795620437956e-06, + "loss": 1.8942, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.8221523761749268, + "learning_rate": 4.562043795620438e-06, + "loss": 2.0045, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.6657445430755615, + "learning_rate": 4.580291970802921e-06, + "loss": 2.0053, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 2.793455123901367, + "learning_rate": 4.598540145985402e-06, + "loss": 1.9577, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.551201820373535, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9256, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 2.739194631576538, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9265, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.2917332649230957, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9664, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.086106777191162, + "eval_runtime": 48.995, + "eval_samples_per_second": 20.41, + "eval_steps_per_second": 0.857, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 2.6489369869232178, + "learning_rate": 4.671532846715329e-06, + "loss": 1.9629, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 2.60017466545105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.932, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.569115400314331, + "learning_rate": 4.708029197080292e-06, + "loss": 1.9902, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 2.7225773334503174, + "learning_rate": 4.726277372262774e-06, + "loss": 1.9619, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.5064749717712402, + "learning_rate": 4.744525547445255e-06, + "loss": 1.8921, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.612931966781616, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9584, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.2524304389953613, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9675, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.8852503299713135, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9107, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 2.654776096343994, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9465, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.679086685180664, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9468, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.0842835903167725, + "eval_runtime": 51.7803, + "eval_samples_per_second": 19.312, + "eval_steps_per_second": 0.811, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.5110855102539062, + "learning_rate": 4.854014598540146e-06, + "loss": 1.88, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.1786117553710938, + "learning_rate": 4.872262773722628e-06, + "loss": 1.8994, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.388328790664673, + "learning_rate": 4.89051094890511e-06, + "loss": 1.8914, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 2.784834384918213, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.923, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.470426559448242, + "learning_rate": 4.927007299270074e-06, + "loss": 1.944, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.690580129623413, + "learning_rate": 4.945255474452555e-06, + "loss": 1.8978, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.869737386703491, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9964, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.9965591430664062, + "learning_rate": 4.981751824817518e-06, + "loss": 1.9706, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.6665658950805664, + "learning_rate": 5e-06, + "loss": 1.9313, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.9231722354888916, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9504, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.068364143371582, + "eval_runtime": 47.8334, + "eval_samples_per_second": 20.906, + "eval_steps_per_second": 0.878, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.6094462871551514, + "learning_rate": 4.995938261575955e-06, + "loss": 1.862, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.780848741531372, + "learning_rate": 4.993907392363932e-06, + "loss": 1.95, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.948662757873535, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.878, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 2.9622998237609863, + "learning_rate": 4.989845653939887e-06, + "loss": 1.8954, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 2.6401216983795166, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9021, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.029592275619507, + "learning_rate": 4.985783915515841e-06, + "loss": 1.8967, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.828392505645752, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9134, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.613926649093628, + "learning_rate": 4.981722177091796e-06, + "loss": 1.8829, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.6173207759857178, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9225, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.4657015800476074, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9142, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.059906482696533, + "eval_runtime": 49.6267, + "eval_samples_per_second": 20.15, + "eval_steps_per_second": 0.846, + "step": 300 + } + ], + "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": 9.35192822808576e+16, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-300/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-400/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-400/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-400/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/latest new file mode 100644 index 0000000000000000000000000000000000000000..e5bdf58d4f29d34e909da25905fad376f73e7c29 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/latest @@ -0,0 +1 @@ +global_step400 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-400/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-400/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-400/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..89ab2c9fe299bd8f6acbeccf13139d165800bcdd --- /dev/null +++ b/babylm_shuffle_local3_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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.7659518718719482, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.279, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.433785915374756, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.3368, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 3.1486687660217285, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3896, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.391956090927124, + "learning_rate": 1.624087591240876e-06, + "loss": 2.297, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.6649584770202637, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.2415, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.5604891777038574, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2688, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.47900652885437, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2933, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.632723569869995, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.2143, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.8085947036743164, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2748, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.3927979469299316, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2685, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.355207920074463, + "eval_runtime": 42.3505, + "eval_samples_per_second": 23.612, + "eval_steps_per_second": 0.992, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.6170592308044434, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.3121, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.699814796447754, + "learning_rate": 1.77007299270073e-06, + "loss": 2.2853, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.3450610637664795, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3297, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.458219528198242, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2639, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.7101449966430664, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2134, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.5734996795654297, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2347, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 2.5200352668762207, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2299, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5022480487823486, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2253, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.190117359161377, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.2272, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.414524555206299, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2424, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.317634344100952, + "eval_runtime": 43.5742, + "eval_samples_per_second": 22.949, + "eval_steps_per_second": 0.964, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.3226804733276367, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1449, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.733006477355957, + "learning_rate": 1.952554744525548e-06, + "loss": 2.2027, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 2.5586233139038086, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2007, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.669914722442627, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2024, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.6148102283477783, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2496, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 2.331369638442993, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1814, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.3051886558532715, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2015, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.6890218257904053, + "learning_rate": 2.062043795620438e-06, + "loss": 2.184, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.4975883960723877, + "learning_rate": 2.08029197080292e-06, + "loss": 2.2439, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.5848891735076904, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1413, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.280966281890869, + "eval_runtime": 43.1001, + "eval_samples_per_second": 23.202, + "eval_steps_per_second": 0.974, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.2104, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.3356540203094482, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2312, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 2.3616602420806885, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.1399, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.581623077392578, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1456, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.547288656234741, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.184, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6386160850524902, + "learning_rate": 2.208029197080292e-06, + "loss": 2.1658, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.593065023422241, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1443, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.530632972717285, + "learning_rate": 2.244525547445256e-06, + "loss": 2.194, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.3553175926208496, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.1918, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 2.7536532878875732, + "learning_rate": 2.281021897810219e-06, + "loss": 2.1477, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2608063220977783, + "eval_runtime": 42.7938, + "eval_samples_per_second": 23.368, + "eval_steps_per_second": 0.981, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.160094738006592, + "learning_rate": 2.299270072992701e-06, + "loss": 2.133, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.4474735260009766, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1468, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.6654434204101562, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.1194, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.726611375808716, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1518, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 2.3560168743133545, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1537, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.664065361022949, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0888, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.3946099281311035, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.0998, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.7812652587890625, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1231, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.64595890045166, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1746, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.788524866104126, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0731, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.2340502738952637, + "eval_runtime": 42.3104, + "eval_samples_per_second": 23.635, + "eval_steps_per_second": 0.993, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 2.9249846935272217, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0564, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.5507798194885254, + "learning_rate": 2.5e-06, + "loss": 2.1714, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.1604716777801514, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0701, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.5429840087890625, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.0662, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8296737670898438, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.1378, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.9396755695343018, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1069, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.736814022064209, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1076, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 2.774909496307373, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1286, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.0861318111419678, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1499, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.624016284942627, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0989, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.215052604675293, + "eval_runtime": 43.9956, + "eval_samples_per_second": 22.73, + "eval_steps_per_second": 0.955, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.801500082015991, + "learning_rate": 2.664233576642336e-06, + "loss": 2.0795, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 2.2107114791870117, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.1035, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.030381441116333, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.0794, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 2.9719064235687256, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1496, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.688839912414551, + "learning_rate": 2.737226277372263e-06, + "loss": 2.1152, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 2.9350979328155518, + "learning_rate": 2.755474452554745e-06, + "loss": 2.0773, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.8958046436309814, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.1805, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.0781922340393066, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.1249, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.032059907913208, + "learning_rate": 2.81021897810219e-06, + "loss": 2.0632, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.798758029937744, + "learning_rate": 2.828467153284672e-06, + "loss": 2.1001, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2005105018615723, + "eval_runtime": 43.5111, + "eval_samples_per_second": 22.983, + "eval_steps_per_second": 0.965, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.7674927711486816, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.0211, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 2.6168575286865234, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0599, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 2.933353900909424, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.0169, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.81402587890625, + "learning_rate": 2.901459854014599e-06, + "loss": 2.122, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.5592572689056396, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.0957, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.738271951675415, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.0868, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.5362703800201416, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1237, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.6304945945739746, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.0825, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.040717601776123, + "learning_rate": 2.992700729927008e-06, + "loss": 2.0388, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.6226463317871094, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0163, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.186411142349243, + "eval_runtime": 44.625, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.5307068824768066, + "learning_rate": 3.029197080291971e-06, + "loss": 2.1097, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 2.9160444736480713, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.0946, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.5250024795532227, + "learning_rate": 3.065693430656935e-06, + "loss": 2.0709, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.6963064670562744, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.0937, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 2.845707416534424, + "learning_rate": 3.102189781021898e-06, + "loss": 2.0851, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.8985090255737305, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.0176, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 2.734377861022949, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.9931, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.5125555992126465, + "learning_rate": 3.156934306569343e-06, + "loss": 2.1133, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.0393502712249756, + "learning_rate": 3.175182481751825e-06, + "loss": 2.0636, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.6410164833068848, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0304, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.171868085861206, + "eval_runtime": 44.2249, + "eval_samples_per_second": 22.612, + "eval_steps_per_second": 0.95, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.7937498092651367, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0518, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7801201343536377, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.026, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7963638305664062, + "learning_rate": 3.248175182481752e-06, + "loss": 1.9642, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6330482959747314, + "learning_rate": 3.266423357664234e-06, + "loss": 2.0545, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.9771625995635986, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0445, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.6112730503082275, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0616, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.732865333557129, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0403, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.9113688468933105, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0382, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.4516773223876953, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.0416, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.781735897064209, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0348, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.160499334335327, + "eval_runtime": 44.5652, + "eval_samples_per_second": 22.439, + "eval_steps_per_second": 0.942, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.9990546703338623, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0324, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.783071279525757, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0518, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.659040689468384, + "learning_rate": 3.43065693430657e-06, + "loss": 1.9722, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 2.799798011779785, + "learning_rate": 3.4489051094890513e-06, + "loss": 2.0362, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 3.025458812713623, + "learning_rate": 3.467153284671533e-06, + "loss": 1.9863, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 2.686415195465088, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0201, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.716064929962158, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0249, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.637653350830078, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0403, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 2.7686848640441895, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0626, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.721956729888916, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.1077, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.152338981628418, + "eval_runtime": 43.5455, + "eval_samples_per_second": 22.964, + "eval_steps_per_second": 0.965, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.92571759223938, + "learning_rate": 3.576642335766424e-06, + "loss": 2.0928, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.9189202785491943, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9315, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 2.990233898162842, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9427, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.702348232269287, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9768, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6949572563171387, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.9954, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.108405113220215, + "learning_rate": 3.667883211678833e-06, + "loss": 1.9631, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.738922119140625, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9644, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.7419424057006836, + "learning_rate": 3.704379562043796e-06, + "loss": 2.0675, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9853813648223877, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0245, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.4614126682281494, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.9665, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.1342060565948486, + "eval_runtime": 46.3954, + "eval_samples_per_second": 21.554, + "eval_steps_per_second": 0.905, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.6787819862365723, + "learning_rate": 3.7591240875912412e-06, + "loss": 2.0063, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 2.563525438308716, + "learning_rate": 3.777372262773723e-06, + "loss": 1.998, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.4981701374053955, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.9738, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 2.5357978343963623, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0221, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.5365488529205322, + "learning_rate": 3.832116788321168e-06, + "loss": 1.9664, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 2.669468641281128, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9577, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 2.6950953006744385, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9786, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8249378204345703, + "learning_rate": 3.886861313868613e-06, + "loss": 2.0041, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.5924317836761475, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0338, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.1282334327697754, + "learning_rate": 3.923357664233577e-06, + "loss": 1.9698, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.1221420764923096, + "eval_runtime": 47.1127, + "eval_samples_per_second": 21.226, + "eval_steps_per_second": 0.891, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.7878193855285645, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0446, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.7137255668640137, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0575, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.286421060562134, + "learning_rate": 3.978102189781022e-06, + "loss": 1.9131, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.5466175079345703, + "learning_rate": 3.996350364963504e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.792483329772949, + "learning_rate": 4.014598540145986e-06, + "loss": 1.912, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.7533867359161377, + "learning_rate": 4.032846715328467e-06, + "loss": 1.9742, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 2.848405361175537, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9815, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.1034414768218994, + "learning_rate": 4.069343065693431e-06, + "loss": 1.9409, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.5109779834747314, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9676, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.486506938934326, + "learning_rate": 4.105839416058395e-06, + "loss": 2.0353, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1102187633514404, + "eval_runtime": 45.384, + "eval_samples_per_second": 22.034, + "eval_steps_per_second": 0.925, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.419269323348999, + "learning_rate": 4.124087591240876e-06, + "loss": 1.9622, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.4967262744903564, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.9523, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 2.7311460971832275, + "learning_rate": 4.16058394160584e-06, + "loss": 2.0234, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.241004467010498, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.907, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.821537733078003, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9761, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 2.5711426734924316, + "learning_rate": 4.215328467153285e-06, + "loss": 1.9411, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.936594009399414, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9432, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.6091108322143555, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9545, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.522674798965454, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9246, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.984649181365967, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9417, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1031947135925293, + "eval_runtime": 47.5409, + "eval_samples_per_second": 21.035, + "eval_steps_per_second": 0.883, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.8250324726104736, + "learning_rate": 4.306569343065693e-06, + "loss": 1.9061, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.7654290199279785, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9653, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.7091264724731445, + "learning_rate": 4.343065693430658e-06, + "loss": 1.9482, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 2.8019583225250244, + "learning_rate": 4.361313868613139e-06, + "loss": 1.9332, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.147368907928467, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9836, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.7050259113311768, + "learning_rate": 4.397810218978102e-06, + "loss": 1.9607, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.738743543624878, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9374, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.122777223587036, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9249, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.850745439529419, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9707, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.391998291015625, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.9703, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.0887911319732666, + "eval_runtime": 48.5149, + "eval_samples_per_second": 20.612, + "eval_steps_per_second": 0.866, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 2.9131698608398438, + "learning_rate": 4.489051094890512e-06, + "loss": 1.9989, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.1045937538146973, + "learning_rate": 4.507299270072993e-06, + "loss": 1.9718, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.7984158992767334, + "learning_rate": 4.525547445255475e-06, + "loss": 1.947, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.6223504543304443, + "learning_rate": 4.543795620437956e-06, + "loss": 1.8942, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.8221523761749268, + "learning_rate": 4.562043795620438e-06, + "loss": 2.0045, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.6657445430755615, + "learning_rate": 4.580291970802921e-06, + "loss": 2.0053, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 2.793455123901367, + "learning_rate": 4.598540145985402e-06, + "loss": 1.9577, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.551201820373535, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9256, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 2.739194631576538, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9265, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.2917332649230957, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9664, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.086106777191162, + "eval_runtime": 48.995, + "eval_samples_per_second": 20.41, + "eval_steps_per_second": 0.857, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 2.6489369869232178, + "learning_rate": 4.671532846715329e-06, + "loss": 1.9629, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 2.60017466545105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.932, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.569115400314331, + "learning_rate": 4.708029197080292e-06, + "loss": 1.9902, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 2.7225773334503174, + "learning_rate": 4.726277372262774e-06, + "loss": 1.9619, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.5064749717712402, + "learning_rate": 4.744525547445255e-06, + "loss": 1.8921, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.612931966781616, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9584, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.2524304389953613, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9675, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.8852503299713135, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9107, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 2.654776096343994, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9465, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.679086685180664, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9468, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.0842835903167725, + "eval_runtime": 51.7803, + "eval_samples_per_second": 19.312, + "eval_steps_per_second": 0.811, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.5110855102539062, + "learning_rate": 4.854014598540146e-06, + "loss": 1.88, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.1786117553710938, + "learning_rate": 4.872262773722628e-06, + "loss": 1.8994, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.388328790664673, + "learning_rate": 4.89051094890511e-06, + "loss": 1.8914, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 2.784834384918213, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.923, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.470426559448242, + "learning_rate": 4.927007299270074e-06, + "loss": 1.944, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.690580129623413, + "learning_rate": 4.945255474452555e-06, + "loss": 1.8978, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.869737386703491, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9964, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.9965591430664062, + "learning_rate": 4.981751824817518e-06, + "loss": 1.9706, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.6665658950805664, + "learning_rate": 5e-06, + "loss": 1.9313, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.9231722354888916, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9504, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.068364143371582, + "eval_runtime": 47.8334, + "eval_samples_per_second": 20.906, + "eval_steps_per_second": 0.878, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.6094462871551514, + "learning_rate": 4.995938261575955e-06, + "loss": 1.862, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.780848741531372, + "learning_rate": 4.993907392363932e-06, + "loss": 1.95, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.948662757873535, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.878, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 2.9622998237609863, + "learning_rate": 4.989845653939887e-06, + "loss": 1.8954, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 2.6401216983795166, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9021, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.029592275619507, + "learning_rate": 4.985783915515841e-06, + "loss": 1.8967, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.828392505645752, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9134, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.613926649093628, + "learning_rate": 4.981722177091796e-06, + "loss": 1.8829, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.6173207759857178, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9225, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.4657015800476074, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9142, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.059906482696533, + "eval_runtime": 49.6267, + "eval_samples_per_second": 20.15, + "eval_steps_per_second": 0.846, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7448673248291016, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9545, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.639096736907959, + "learning_rate": 4.973598700243705e-06, + "loss": 1.8664, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 2.6445884704589844, + "learning_rate": 4.971567831031681e-06, + "loss": 1.888, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.9130005836486816, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7717, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.48093318939209, + "learning_rate": 4.967506092607636e-06, + "loss": 1.8538, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.8918230533599854, + "learning_rate": 4.965475223395614e-06, + "loss": 1.886, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.511747121810913, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.8633, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.860029697418213, + "learning_rate": 4.961413484971569e-06, + "loss": 1.9063, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7304794788360596, + "learning_rate": 4.959382615759545e-06, + "loss": 1.8872, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.6569266319274902, + "learning_rate": 4.957351746547523e-06, + "loss": 1.8721, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0610575675964355, + "eval_runtime": 43.9292, + "eval_samples_per_second": 22.764, + "eval_steps_per_second": 0.956, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.1482994556427, + "learning_rate": 4.9553208773355e-06, + "loss": 2.0123, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.59568190574646, + "learning_rate": 4.953290008123478e-06, + "loss": 1.902, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.5897581577301025, + "learning_rate": 4.951259138911454e-06, + "loss": 1.8406, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.734032154083252, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.9605, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.8088371753692627, + "learning_rate": 4.947197400487409e-06, + "loss": 1.896, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.9678707122802734, + "learning_rate": 4.945166531275386e-06, + "loss": 1.8625, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.261200189590454, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8884, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 2.5811619758605957, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9311, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.6783649921417236, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9324, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.6533944606781006, + "learning_rate": 4.937043054427295e-06, + "loss": 1.8575, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.049668312072754, + "eval_runtime": 47.7347, + "eval_samples_per_second": 20.949, + "eval_steps_per_second": 0.88, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 3.0882792472839355, + "learning_rate": 4.935012185215272e-06, + "loss": 1.84, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.8259081840515137, + "learning_rate": 4.93298131600325e-06, + "loss": 1.9109, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.540908098220825, + "learning_rate": 4.930950446791227e-06, + "loss": 1.8906, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.9277396202087402, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9047, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.392449140548706, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.876, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 3.1890463829040527, + "learning_rate": 4.924857839155159e-06, + "loss": 1.8709, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.725710391998291, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8953, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.6287014484405518, + "learning_rate": 4.920796100731113e-06, + "loss": 1.8204, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.7316651344299316, + "learning_rate": 4.918765231519091e-06, + "loss": 1.882, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.5561037063598633, + "learning_rate": 4.916734362307068e-06, + "loss": 1.8331, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0412285327911377, + "eval_runtime": 50.237, + "eval_samples_per_second": 19.906, + "eval_steps_per_second": 0.836, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.974501132965088, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8414, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.4530839920043945, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8323, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.789569616317749, + "learning_rate": 4.910641754671e-06, + "loss": 1.8697, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.582704544067383, + "learning_rate": 4.908610885458977e-06, + "loss": 1.8592, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 2.776442289352417, + "learning_rate": 4.906580016246954e-06, + "loss": 1.912, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.670191764831543, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.8899, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.4534454345703125, + "learning_rate": 4.902518277822909e-06, + "loss": 1.8878, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.007469654083252, + "learning_rate": 4.900487408610886e-06, + "loss": 1.8977, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 2.856466293334961, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9062, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.817441463470459, + "learning_rate": 4.89642567018684e-06, + "loss": 1.9686, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.0344440937042236, + "eval_runtime": 47.2472, + "eval_samples_per_second": 21.165, + "eval_steps_per_second": 0.889, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.9732158184051514, + "learning_rate": 4.894394800974818e-06, + "loss": 1.8332, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 3.0413055419921875, + "learning_rate": 4.892363931762794e-06, + "loss": 1.9171, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.646749973297119, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.8403, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.1274876594543457, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8218, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.8122124671936035, + "learning_rate": 4.886271324126727e-06, + "loss": 1.9272, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.4608492851257324, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.8974, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.928083658218384, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.8646, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.6308164596557617, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8219, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7311816215515137, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8326, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.640162467956543, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8013, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0261101722717285, + "eval_runtime": 47.0605, + "eval_samples_per_second": 21.249, + "eval_steps_per_second": 0.892, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.8542518615722656, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8233, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.464303970336914, + "learning_rate": 4.872055239642567e-06, + "loss": 1.787, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8535103797912598, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8535, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.5060505867004395, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8733, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.6549222469329834, + "learning_rate": 4.865962632006499e-06, + "loss": 1.8047, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.8636224269866943, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8543, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.644256830215454, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8853, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.949490547180176, + "learning_rate": 4.859870024370431e-06, + "loss": 1.8951, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.4576282501220703, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9069, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.6140542030334473, + "learning_rate": 4.855808285946385e-06, + "loss": 1.8573, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0159800052642822, + "eval_runtime": 44.2447, + "eval_samples_per_second": 22.602, + "eval_steps_per_second": 0.949, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.525766134262085, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.844, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.5479896068573, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8572, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.628570795059204, + "learning_rate": 4.849715678310317e-06, + "loss": 1.865, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.2768666744232178, + "learning_rate": 4.847684809098294e-06, + "loss": 1.8355, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.635516881942749, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8603, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.5307435989379883, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8482, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.3688952922821045, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8894, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.6000115871429443, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.8045, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7782504558563232, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8283, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.6914777755737305, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8765, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0086047649383545, + "eval_runtime": 48.245, + "eval_samples_per_second": 20.728, + "eval_steps_per_second": 0.871, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.598365306854248, + "learning_rate": 4.833468724614135e-06, + "loss": 1.8461, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.6718530654907227, + "learning_rate": 4.831437855402113e-06, + "loss": 1.7453, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.393528699874878, + "learning_rate": 4.82940698619009e-06, + "loss": 1.8275, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 3.0622947216033936, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8658, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.431041955947876, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8601, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.939424514770508, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8649, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.8116796016693115, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8303, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.511960029602051, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8383, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 3.8372581005096436, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8111, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.511948347091675, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8448, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.9998281002044678, + "eval_runtime": 44.7624, + "eval_samples_per_second": 22.34, + "eval_steps_per_second": 0.938, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 3.5780487060546875, + "learning_rate": 4.813160032493907e-06, + "loss": 1.8161, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.81740403175354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.8633, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.8117666244506836, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8385, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.5791451930999756, + "learning_rate": 4.80706742485784e-06, + "loss": 1.896, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6816465854644775, + "learning_rate": 4.805036555645816e-06, + "loss": 1.8126, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.6419460773468018, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8753, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.666086196899414, + "learning_rate": 4.800974817221771e-06, + "loss": 1.8305, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8626585006713867, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8463, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.351825475692749, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.838, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.300717830657959, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8711, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.9904799461364746, + "eval_runtime": 43.9511, + "eval_samples_per_second": 22.753, + "eval_steps_per_second": 0.956, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.43884015083313, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8145, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.50593900680542, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8103, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.4643657207489014, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8162, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.4615302085876465, + "learning_rate": 4.786758732737612e-06, + "loss": 1.8637, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.6566433906555176, + "learning_rate": 4.784727863525589e-06, + "loss": 1.7805, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.5973432064056396, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8327, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.4996957778930664, + "learning_rate": 4.780666125101544e-06, + "loss": 1.7761, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 3.0266876220703125, + "learning_rate": 4.778635255889521e-06, + "loss": 1.8036, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.522369146347046, + "learning_rate": 4.776604386677498e-06, + "loss": 1.814, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.482727527618408, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8665, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.9888195991516113, + "eval_runtime": 44.0755, + "eval_samples_per_second": 22.688, + "eval_steps_per_second": 0.953, + "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_local3_10M_seed0/runs/checkpoint-400/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-400/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-700/config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-700/generation_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-700/latest b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/latest new file mode 100644 index 0000000000000000000000000000000000000000..a0f3e526aa9af2ac647b278f006bb9616843c5d6 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/latest @@ -0,0 +1 @@ +global_step700 \ No newline at end of file diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/model.safetensors.index.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-700/special_tokens_map.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-700/tokenizer_config.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/babylm_shuffle_local3_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_local3_10M_seed0/runs/checkpoint-700/trainer_state.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..917aca133175ed3ede67b903b58214d80f4ac9d1 --- /dev/null +++ b/babylm_shuffle_local3_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": 2.9089, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9026, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8829, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8952, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8723, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0066, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9114, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9456, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9227, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9777, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.956331491470337, + "eval_runtime": 43.1475, + "eval_samples_per_second": 23.176, + "eval_steps_per_second": 0.973, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9469, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.852, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.932, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9105, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 5.27911901473999, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9662, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 5.446037292480469, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9031, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 5.539210796356201, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.8999, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 5.492989540100098, + "learning_rate": 7.299270072992701e-08, + "loss": 2.936, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 5.78201961517334, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8898, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 5.155740737915039, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9726, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9550349712371826, + "eval_runtime": 46.2677, + "eval_samples_per_second": 21.613, + "eval_steps_per_second": 0.908, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 5.111966609954834, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.943, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 5.59846830368042, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8627, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 5.231739521026611, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9047, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.892563343048096, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.935, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 5.221184253692627, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9351, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.918729782104492, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.9341, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 4.572084903717041, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9393, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 4.296558380126953, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8892, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 4.179786205291748, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8798, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 4.410481929779053, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.8489, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.908602714538574, + "eval_runtime": 45.2643, + "eval_samples_per_second": 22.092, + "eval_steps_per_second": 0.928, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.663278102874756, + "learning_rate": 3.102189781021898e-07, + "loss": 2.8722, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.7895305156707764, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8605, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.7787935733795166, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8245, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.669219732284546, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8236, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.64031982421875, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8475, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 3.3962018489837646, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8712, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.9786815643310547, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8081, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.9515726566314697, + "learning_rate": 4.379562043795621e-07, + "loss": 2.7586, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 3.1295464038848877, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8621, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 3.0926554203033447, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7642, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.830131769180298, + "eval_runtime": 46.7688, + "eval_samples_per_second": 21.382, + "eval_steps_per_second": 0.898, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.8757333755493164, + "learning_rate": 4.927007299270073e-07, + "loss": 2.8001, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.7291815280914307, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8222, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.6528494358062744, + "learning_rate": 5.291970802919708e-07, + "loss": 2.7992, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.462028980255127, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8014, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.6123344898223877, + "learning_rate": 5.656934306569343e-07, + "loss": 2.7932, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.4687554836273193, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7365, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.6216812133789062, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7167, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.3418407440185547, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7478, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 2.260784149169922, + "learning_rate": 6.386861313868614e-07, + "loss": 2.6721, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.1247763633728027, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8029, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.744903802871704, + "eval_runtime": 46.0361, + "eval_samples_per_second": 21.722, + "eval_steps_per_second": 0.912, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.4009053707122803, + "learning_rate": 6.751824817518248e-07, + "loss": 2.6825, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 2.1920251846313477, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7179, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 2.0571298599243164, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6414, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.8794459104537964, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6493, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.7492607831954956, + "learning_rate": 7.48175182481752e-07, + "loss": 2.6689, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.8351389169692993, + "learning_rate": 7.664233576642337e-07, + "loss": 2.6771, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.946508526802063, + "learning_rate": 7.846715328467154e-07, + "loss": 2.6883, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.9688613414764404, + "learning_rate": 8.029197080291971e-07, + "loss": 2.664, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.7094988822937012, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6165, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.6592987775802612, + "learning_rate": 8.394160583941606e-07, + "loss": 2.5819, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.6823935508728027, + "eval_runtime": 49.1139, + "eval_samples_per_second": 20.361, + "eval_steps_per_second": 0.855, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7375643253326416, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6174, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6795835494995117, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6035, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.585369348526001, + "learning_rate": 8.94160583941606e-07, + "loss": 2.619, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.6931068897247314, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6323, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.712717890739441, + "learning_rate": 9.306569343065694e-07, + "loss": 2.5926, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.801835536956787, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6139, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.660003900527954, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6031, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.6886236667633057, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6414, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 1.8952292203903198, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.5794, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.953311562538147, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.5342, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.616886854171753, + "eval_runtime": 51.6884, + "eval_samples_per_second": 19.347, + "eval_steps_per_second": 0.813, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.733276605606079, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6102, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.240635871887207, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.4948, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 2.0552077293395996, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.4928, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.2784221172332764, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.5063, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 2.0558624267578125, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5365, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.0571038722991943, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.546, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.0617258548736572, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4652, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 1.8994337320327759, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.453, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.261740207672119, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4599, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.3452014923095703, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4489, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5349061489105225, + "eval_runtime": 47.1188, + "eval_samples_per_second": 21.223, + "eval_steps_per_second": 0.891, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.537118434906006, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.4532, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.4936294555664062, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.4836, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 3.0557146072387695, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4421, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.464334487915039, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4602, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.703789234161377, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4416, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.6780083179473877, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.3866, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.495685577392578, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4121, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 2.4785284996032715, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3287, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.5816562175750732, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.3787, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.7493233680725098, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.306, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4593827724456787, + "eval_runtime": 46.9805, + "eval_samples_per_second": 21.285, + "eval_steps_per_second": 0.894, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 2.702443838119507, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4027, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.529048204421997, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3646, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.9887475967407227, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.3913, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.357, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 2.066145896911621, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.3205, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 3.58024001121521, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3856, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.536687135696411, + "learning_rate": 1.496350364963504e-06, + "loss": 2.3602, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.3430354595184326, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3757, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.6249454021453857, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.3316, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.5809590816497803, + "learning_rate": 1.551094890510949e-06, + "loss": 2.2809, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.4053611755371094, + "eval_runtime": 46.9655, + "eval_samples_per_second": 21.292, + "eval_steps_per_second": 0.894, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.7659518718719482, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.279, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.433785915374756, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.3368, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 3.1486687660217285, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3896, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.391956090927124, + "learning_rate": 1.624087591240876e-06, + "loss": 2.297, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.6649584770202637, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.2415, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 2.5604891777038574, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2688, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.47900652885437, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2933, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.632723569869995, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.2143, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.8085947036743164, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2748, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.3927979469299316, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2685, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.355207920074463, + "eval_runtime": 42.3505, + "eval_samples_per_second": 23.612, + "eval_steps_per_second": 0.992, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 2.6170592308044434, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.3121, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.699814796447754, + "learning_rate": 1.77007299270073e-06, + "loss": 2.2853, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 2.3450610637664795, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3297, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.458219528198242, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.2639, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 2.7101449966430664, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2134, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.5734996795654297, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2347, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 2.5200352668762207, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2299, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.5022480487823486, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2253, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.190117359161377, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.2272, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.414524555206299, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2424, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.317634344100952, + "eval_runtime": 43.5742, + "eval_samples_per_second": 22.949, + "eval_steps_per_second": 0.964, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.3226804733276367, + "learning_rate": 1.934306569343066e-06, + "loss": 2.1449, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.733006477355957, + "learning_rate": 1.952554744525548e-06, + "loss": 2.2027, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 2.5586233139038086, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2007, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.669914722442627, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2024, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 2.6148102283477783, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2496, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 2.331369638442993, + "learning_rate": 2.025547445255475e-06, + "loss": 2.1814, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.3051886558532715, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2015, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.6890218257904053, + "learning_rate": 2.062043795620438e-06, + "loss": 2.184, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.4975883960723877, + "learning_rate": 2.08029197080292e-06, + "loss": 2.2439, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 2.5848891735076904, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.1413, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.280966281890869, + "eval_runtime": 43.1001, + "eval_samples_per_second": 23.202, + "eval_steps_per_second": 0.974, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.2104, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.3356540203094482, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2312, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 2.3616602420806885, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.1399, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.581623077392578, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1456, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 2.547288656234741, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.184, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6386160850524902, + "learning_rate": 2.208029197080292e-06, + "loss": 2.1658, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 2.593065023422241, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1443, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 2.530632972717285, + "learning_rate": 2.244525547445256e-06, + "loss": 2.194, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 2.3553175926208496, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.1918, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 2.7536532878875732, + "learning_rate": 2.281021897810219e-06, + "loss": 2.1477, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2608063220977783, + "eval_runtime": 42.7938, + "eval_samples_per_second": 23.368, + "eval_steps_per_second": 0.981, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.160094738006592, + "learning_rate": 2.299270072992701e-06, + "loss": 2.133, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.4474735260009766, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1468, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 2.6654434204101562, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.1194, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.726611375808716, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1518, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 2.3560168743133545, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1537, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 2.664065361022949, + "learning_rate": 2.39051094890511e-06, + "loss": 2.0888, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.3946099281311035, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.0998, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 2.7812652587890625, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1231, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 2.64595890045166, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1746, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.788524866104126, + "learning_rate": 2.463503649635037e-06, + "loss": 2.0731, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.2340502738952637, + "eval_runtime": 42.3104, + "eval_samples_per_second": 23.635, + "eval_steps_per_second": 0.993, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 2.9249846935272217, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0564, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.5507798194885254, + "learning_rate": 2.5e-06, + "loss": 2.1714, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 3.1604716777801514, + "learning_rate": 2.518248175182482e-06, + "loss": 2.0701, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.5429840087890625, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.0662, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8296737670898438, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.1378, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.9396755695343018, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1069, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.736814022064209, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1076, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 2.774909496307373, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1286, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.0861318111419678, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1499, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.624016284942627, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.0989, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.215052604675293, + "eval_runtime": 43.9956, + "eval_samples_per_second": 22.73, + "eval_steps_per_second": 0.955, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 2.801500082015991, + "learning_rate": 2.664233576642336e-06, + "loss": 2.0795, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 2.2107114791870117, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.1035, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 3.030381441116333, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.0794, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 2.9719064235687256, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1496, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.688839912414551, + "learning_rate": 2.737226277372263e-06, + "loss": 2.1152, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 2.9350979328155518, + "learning_rate": 2.755474452554745e-06, + "loss": 2.0773, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.8958046436309814, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.1805, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.0781922340393066, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.1249, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.032059907913208, + "learning_rate": 2.81021897810219e-06, + "loss": 2.0632, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.798758029937744, + "learning_rate": 2.828467153284672e-06, + "loss": 2.1001, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2005105018615723, + "eval_runtime": 43.5111, + "eval_samples_per_second": 22.983, + "eval_steps_per_second": 0.965, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.7674927711486816, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.0211, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 2.6168575286865234, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0599, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 2.933353900909424, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.0169, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.81402587890625, + "learning_rate": 2.901459854014599e-06, + "loss": 2.122, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.5592572689056396, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.0957, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.738271951675415, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.0868, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 2.5362703800201416, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1237, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.6304945945739746, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.0825, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.040717601776123, + "learning_rate": 2.992700729927008e-06, + "loss": 2.0388, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 2.6226463317871094, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0163, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.186411142349243, + "eval_runtime": 44.625, + "eval_samples_per_second": 22.409, + "eval_steps_per_second": 0.941, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.5307068824768066, + "learning_rate": 3.029197080291971e-06, + "loss": 2.1097, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 2.9160444736480713, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.0946, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.5250024795532227, + "learning_rate": 3.065693430656935e-06, + "loss": 2.0709, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.6963064670562744, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.0937, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 2.845707416534424, + "learning_rate": 3.102189781021898e-06, + "loss": 2.0851, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.8985090255737305, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.0176, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 2.734377861022949, + "learning_rate": 3.1386861313868614e-06, + "loss": 1.9931, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 3.5125555992126465, + "learning_rate": 3.156934306569343e-06, + "loss": 2.1133, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.0393502712249756, + "learning_rate": 3.175182481751825e-06, + "loss": 2.0636, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.6410164833068848, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0304, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.171868085861206, + "eval_runtime": 44.2249, + "eval_samples_per_second": 22.612, + "eval_steps_per_second": 0.95, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.7937498092651367, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0518, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7801201343536377, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.026, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7963638305664062, + "learning_rate": 3.248175182481752e-06, + "loss": 1.9642, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6330482959747314, + "learning_rate": 3.266423357664234e-06, + "loss": 2.0545, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.9771625995635986, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0445, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.6112730503082275, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0616, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.732865333557129, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0403, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.9113688468933105, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0382, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.4516773223876953, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.0416, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.781735897064209, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0348, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.160499334335327, + "eval_runtime": 44.5652, + "eval_samples_per_second": 22.439, + "eval_steps_per_second": 0.942, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.9990546703338623, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0324, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.783071279525757, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0518, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 2.659040689468384, + "learning_rate": 3.43065693430657e-06, + "loss": 1.9722, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 2.799798011779785, + "learning_rate": 3.4489051094890513e-06, + "loss": 2.0362, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 3.025458812713623, + "learning_rate": 3.467153284671533e-06, + "loss": 1.9863, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 2.686415195465088, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0201, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 2.716064929962158, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0249, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.637653350830078, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0403, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 2.7686848640441895, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0626, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.721956729888916, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.1077, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.152338981628418, + "eval_runtime": 43.5455, + "eval_samples_per_second": 22.964, + "eval_steps_per_second": 0.965, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.92571759223938, + "learning_rate": 3.576642335766424e-06, + "loss": 2.0928, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 2.9189202785491943, + "learning_rate": 3.5948905109489054e-06, + "loss": 1.9315, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 2.990233898162842, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9427, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.702348232269287, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9768, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.6949572563171387, + "learning_rate": 3.6496350364963505e-06, + "loss": 1.9954, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.108405113220215, + "learning_rate": 3.667883211678833e-06, + "loss": 1.9631, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.738922119140625, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9644, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.7419424057006836, + "learning_rate": 3.704379562043796e-06, + "loss": 2.0675, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9853813648223877, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0245, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.4614126682281494, + "learning_rate": 3.7408759124087594e-06, + "loss": 1.9665, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.1342060565948486, + "eval_runtime": 46.3954, + "eval_samples_per_second": 21.554, + "eval_steps_per_second": 0.905, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.6787819862365723, + "learning_rate": 3.7591240875912412e-06, + "loss": 2.0063, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 2.563525438308716, + "learning_rate": 3.777372262773723e-06, + "loss": 1.998, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 2.4981701374053955, + "learning_rate": 3.7956204379562045e-06, + "loss": 1.9738, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 2.5357978343963623, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0221, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.5365488529205322, + "learning_rate": 3.832116788321168e-06, + "loss": 1.9664, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 2.669468641281128, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9577, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 2.6950953006744385, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9786, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8249378204345703, + "learning_rate": 3.886861313868613e-06, + "loss": 2.0041, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 2.5924317836761475, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0338, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 3.1282334327697754, + "learning_rate": 3.923357664233577e-06, + "loss": 1.9698, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.1221420764923096, + "eval_runtime": 47.1127, + "eval_samples_per_second": 21.226, + "eval_steps_per_second": 0.891, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.7878193855285645, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0446, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.7137255668640137, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0575, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 3.286421060562134, + "learning_rate": 3.978102189781022e-06, + "loss": 1.9131, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.5466175079345703, + "learning_rate": 3.996350364963504e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.792483329772949, + "learning_rate": 4.014598540145986e-06, + "loss": 1.912, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.7533867359161377, + "learning_rate": 4.032846715328467e-06, + "loss": 1.9742, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 2.848405361175537, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9815, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 3.1034414768218994, + "learning_rate": 4.069343065693431e-06, + "loss": 1.9409, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.5109779834747314, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9676, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.486506938934326, + "learning_rate": 4.105839416058395e-06, + "loss": 2.0353, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1102187633514404, + "eval_runtime": 45.384, + "eval_samples_per_second": 22.034, + "eval_steps_per_second": 0.925, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.419269323348999, + "learning_rate": 4.124087591240876e-06, + "loss": 1.9622, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.4967262744903564, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.9523, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 2.7311460971832275, + "learning_rate": 4.16058394160584e-06, + "loss": 2.0234, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 3.241004467010498, + "learning_rate": 4.1788321167883214e-06, + "loss": 1.907, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.821537733078003, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9761, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 2.5711426734924316, + "learning_rate": 4.215328467153285e-06, + "loss": 1.9411, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.936594009399414, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9432, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.6091108322143555, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9545, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.522674798965454, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9246, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.984649181365967, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9417, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1031947135925293, + "eval_runtime": 47.5409, + "eval_samples_per_second": 21.035, + "eval_steps_per_second": 0.883, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.8250324726104736, + "learning_rate": 4.306569343065693e-06, + "loss": 1.9061, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.7654290199279785, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9653, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.7091264724731445, + "learning_rate": 4.343065693430658e-06, + "loss": 1.9482, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 2.8019583225250244, + "learning_rate": 4.361313868613139e-06, + "loss": 1.9332, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.147368907928467, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9836, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.7050259113311768, + "learning_rate": 4.397810218978102e-06, + "loss": 1.9607, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.738743543624878, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9374, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.122777223587036, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9249, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.850745439529419, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9707, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.391998291015625, + "learning_rate": 4.4708029197080295e-06, + "loss": 1.9703, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.0887911319732666, + "eval_runtime": 48.5149, + "eval_samples_per_second": 20.612, + "eval_steps_per_second": 0.866, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 2.9131698608398438, + "learning_rate": 4.489051094890512e-06, + "loss": 1.9989, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.1045937538146973, + "learning_rate": 4.507299270072993e-06, + "loss": 1.9718, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.7984158992767334, + "learning_rate": 4.525547445255475e-06, + "loss": 1.947, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.6223504543304443, + "learning_rate": 4.543795620437956e-06, + "loss": 1.8942, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.8221523761749268, + "learning_rate": 4.562043795620438e-06, + "loss": 2.0045, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.6657445430755615, + "learning_rate": 4.580291970802921e-06, + "loss": 2.0053, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 2.793455123901367, + "learning_rate": 4.598540145985402e-06, + "loss": 1.9577, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.551201820373535, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9256, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 2.739194631576538, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9265, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.2917332649230957, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9664, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.086106777191162, + "eval_runtime": 48.995, + "eval_samples_per_second": 20.41, + "eval_steps_per_second": 0.857, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 2.6489369869232178, + "learning_rate": 4.671532846715329e-06, + "loss": 1.9629, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 2.60017466545105, + "learning_rate": 4.689781021897811e-06, + "loss": 1.932, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.569115400314331, + "learning_rate": 4.708029197080292e-06, + "loss": 1.9902, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 2.7225773334503174, + "learning_rate": 4.726277372262774e-06, + "loss": 1.9619, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 2.5064749717712402, + "learning_rate": 4.744525547445255e-06, + "loss": 1.8921, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.612931966781616, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9584, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.2524304389953613, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9675, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.8852503299713135, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9107, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 2.654776096343994, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9465, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.679086685180664, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9468, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.0842835903167725, + "eval_runtime": 51.7803, + "eval_samples_per_second": 19.312, + "eval_steps_per_second": 0.811, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.5110855102539062, + "learning_rate": 4.854014598540146e-06, + "loss": 1.88, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 3.1786117553710938, + "learning_rate": 4.872262773722628e-06, + "loss": 1.8994, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.388328790664673, + "learning_rate": 4.89051094890511e-06, + "loss": 1.8914, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 2.784834384918213, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.923, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.470426559448242, + "learning_rate": 4.927007299270074e-06, + "loss": 1.944, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 2.690580129623413, + "learning_rate": 4.945255474452555e-06, + "loss": 1.8978, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.869737386703491, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9964, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.9965591430664062, + "learning_rate": 4.981751824817518e-06, + "loss": 1.9706, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.6665658950805664, + "learning_rate": 5e-06, + "loss": 1.9313, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.9231722354888916, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9504, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.068364143371582, + "eval_runtime": 47.8334, + "eval_samples_per_second": 20.906, + "eval_steps_per_second": 0.878, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.6094462871551514, + "learning_rate": 4.995938261575955e-06, + "loss": 1.862, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.780848741531372, + "learning_rate": 4.993907392363932e-06, + "loss": 1.95, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.948662757873535, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.878, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 2.9622998237609863, + "learning_rate": 4.989845653939887e-06, + "loss": 1.8954, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 2.6401216983795166, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9021, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 3.029592275619507, + "learning_rate": 4.985783915515841e-06, + "loss": 1.8967, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.828392505645752, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9134, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.613926649093628, + "learning_rate": 4.981722177091796e-06, + "loss": 1.8829, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.6173207759857178, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9225, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.4657015800476074, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9142, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.059906482696533, + "eval_runtime": 49.6267, + "eval_samples_per_second": 20.15, + "eval_steps_per_second": 0.846, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7448673248291016, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9545, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.639096736907959, + "learning_rate": 4.973598700243705e-06, + "loss": 1.8664, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 2.6445884704589844, + "learning_rate": 4.971567831031681e-06, + "loss": 1.888, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.9130005836486816, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.7717, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.48093318939209, + "learning_rate": 4.967506092607636e-06, + "loss": 1.8538, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.8918230533599854, + "learning_rate": 4.965475223395614e-06, + "loss": 1.886, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.511747121810913, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.8633, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.860029697418213, + "learning_rate": 4.961413484971569e-06, + "loss": 1.9063, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7304794788360596, + "learning_rate": 4.959382615759545e-06, + "loss": 1.8872, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.6569266319274902, + "learning_rate": 4.957351746547523e-06, + "loss": 1.8721, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0610575675964355, + "eval_runtime": 43.9292, + "eval_samples_per_second": 22.764, + "eval_steps_per_second": 0.956, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 3.1482994556427, + "learning_rate": 4.9553208773355e-06, + "loss": 2.0123, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 2.59568190574646, + "learning_rate": 4.953290008123478e-06, + "loss": 1.902, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.5897581577301025, + "learning_rate": 4.951259138911454e-06, + "loss": 1.8406, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.734032154083252, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.9605, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 2.8088371753692627, + "learning_rate": 4.947197400487409e-06, + "loss": 1.896, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.9678707122802734, + "learning_rate": 4.945166531275386e-06, + "loss": 1.8625, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.261200189590454, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8884, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 2.5811619758605957, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9311, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.6783649921417236, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9324, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.6533944606781006, + "learning_rate": 4.937043054427295e-06, + "loss": 1.8575, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.049668312072754, + "eval_runtime": 47.7347, + "eval_samples_per_second": 20.949, + "eval_steps_per_second": 0.88, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 3.0882792472839355, + "learning_rate": 4.935012185215272e-06, + "loss": 1.84, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.8259081840515137, + "learning_rate": 4.93298131600325e-06, + "loss": 1.9109, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.540908098220825, + "learning_rate": 4.930950446791227e-06, + "loss": 1.8906, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.9277396202087402, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9047, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.392449140548706, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.876, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 3.1890463829040527, + "learning_rate": 4.924857839155159e-06, + "loss": 1.8709, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.725710391998291, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8953, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.6287014484405518, + "learning_rate": 4.920796100731113e-06, + "loss": 1.8204, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.7316651344299316, + "learning_rate": 4.918765231519091e-06, + "loss": 1.882, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.5561037063598633, + "learning_rate": 4.916734362307068e-06, + "loss": 1.8331, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0412285327911377, + "eval_runtime": 50.237, + "eval_samples_per_second": 19.906, + "eval_steps_per_second": 0.836, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.974501132965088, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8414, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.4530839920043945, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8323, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.789569616317749, + "learning_rate": 4.910641754671e-06, + "loss": 1.8697, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.582704544067383, + "learning_rate": 4.908610885458977e-06, + "loss": 1.8592, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 2.776442289352417, + "learning_rate": 4.906580016246954e-06, + "loss": 1.912, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 2.670191764831543, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.8899, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.4534454345703125, + "learning_rate": 4.902518277822909e-06, + "loss": 1.8878, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.007469654083252, + "learning_rate": 4.900487408610886e-06, + "loss": 1.8977, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 2.856466293334961, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9062, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.817441463470459, + "learning_rate": 4.89642567018684e-06, + "loss": 1.9686, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.0344440937042236, + "eval_runtime": 47.2472, + "eval_samples_per_second": 21.165, + "eval_steps_per_second": 0.889, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 2.9732158184051514, + "learning_rate": 4.894394800974818e-06, + "loss": 1.8332, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 3.0413055419921875, + "learning_rate": 4.892363931762794e-06, + "loss": 1.9171, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.646749973297119, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.8403, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 3.1274876594543457, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8218, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.8122124671936035, + "learning_rate": 4.886271324126727e-06, + "loss": 1.9272, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.4608492851257324, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.8974, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.928083658218384, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.8646, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.6308164596557617, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8219, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7311816215515137, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8326, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.640162467956543, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8013, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0261101722717285, + "eval_runtime": 47.0605, + "eval_samples_per_second": 21.249, + "eval_steps_per_second": 0.892, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 2.8542518615722656, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8233, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 2.464303970336914, + "learning_rate": 4.872055239642567e-06, + "loss": 1.787, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.8535103797912598, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8535, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.5060505867004395, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8733, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 2.6549222469329834, + "learning_rate": 4.865962632006499e-06, + "loss": 1.8047, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.8636224269866943, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8543, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 2.644256830215454, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8853, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.949490547180176, + "learning_rate": 4.859870024370431e-06, + "loss": 1.8951, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 2.4576282501220703, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9069, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.6140542030334473, + "learning_rate": 4.855808285946385e-06, + "loss": 1.8573, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0159800052642822, + "eval_runtime": 44.2447, + "eval_samples_per_second": 22.602, + "eval_steps_per_second": 0.949, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 2.525766134262085, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.844, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.5479896068573, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8572, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.628570795059204, + "learning_rate": 4.849715678310317e-06, + "loss": 1.865, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 2.2768666744232178, + "learning_rate": 4.847684809098294e-06, + "loss": 1.8355, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.635516881942749, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8603, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.5307435989379883, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8482, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.3688952922821045, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8894, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.6000115871429443, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.8045, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7782504558563232, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8283, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.6914777755737305, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8765, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0086047649383545, + "eval_runtime": 48.245, + "eval_samples_per_second": 20.728, + "eval_steps_per_second": 0.871, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.598365306854248, + "learning_rate": 4.833468724614135e-06, + "loss": 1.8461, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.6718530654907227, + "learning_rate": 4.831437855402113e-06, + "loss": 1.7453, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.393528699874878, + "learning_rate": 4.82940698619009e-06, + "loss": 1.8275, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 3.0622947216033936, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8658, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.431041955947876, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8601, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.939424514770508, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8649, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.8116796016693115, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8303, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.511960029602051, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8383, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 3.8372581005096436, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8111, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.511948347091675, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8448, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 1.9998281002044678, + "eval_runtime": 44.7624, + "eval_samples_per_second": 22.34, + "eval_steps_per_second": 0.938, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 3.5780487060546875, + "learning_rate": 4.813160032493907e-06, + "loss": 1.8161, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.81740403175354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.8633, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.8117666244506836, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8385, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.5791451930999756, + "learning_rate": 4.80706742485784e-06, + "loss": 1.896, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.6816465854644775, + "learning_rate": 4.805036555645816e-06, + "loss": 1.8126, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.6419460773468018, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8753, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.666086196899414, + "learning_rate": 4.800974817221771e-06, + "loss": 1.8305, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8626585006713867, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8463, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.351825475692749, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.838, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.300717830657959, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8711, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 1.9904799461364746, + "eval_runtime": 43.9511, + "eval_samples_per_second": 22.753, + "eval_steps_per_second": 0.956, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 2.43884015083313, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8145, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.50593900680542, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8103, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.4643657207489014, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8162, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.4615302085876465, + "learning_rate": 4.786758732737612e-06, + "loss": 1.8637, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.6566433906555176, + "learning_rate": 4.784727863525589e-06, + "loss": 1.7805, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.5973432064056396, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8327, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.4996957778930664, + "learning_rate": 4.780666125101544e-06, + "loss": 1.7761, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 3.0266876220703125, + "learning_rate": 4.778635255889521e-06, + "loss": 1.8036, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.522369146347046, + "learning_rate": 4.776604386677498e-06, + "loss": 1.814, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.482727527618408, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8665, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 1.9888195991516113, + "eval_runtime": 44.0755, + "eval_samples_per_second": 22.688, + "eval_steps_per_second": 0.953, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 3.237544059753418, + "learning_rate": 4.772542648253453e-06, + "loss": 1.8642, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.393286943435669, + "learning_rate": 4.77051177904143e-06, + "loss": 1.7857, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.66072154045105, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8423, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.6725080013275146, + "learning_rate": 4.766450040617385e-06, + "loss": 1.815, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.7657666206359863, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8729, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 2.3153769969940186, + "learning_rate": 4.762388302193339e-06, + "loss": 1.8349, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5804553031921387, + "learning_rate": 4.760357432981316e-06, + "loss": 1.8155, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.7280807495117188, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8262, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.551940441131592, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8332, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.6137285232543945, + "learning_rate": 4.754264825345248e-06, + "loss": 1.7805, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 1.983920693397522, + "eval_runtime": 42.525, + "eval_samples_per_second": 23.516, + "eval_steps_per_second": 0.988, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.4460153579711914, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.7953, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.556032657623291, + "learning_rate": 4.750203086921203e-06, + "loss": 1.7393, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.335983991622925, + "learning_rate": 4.74817221770918e-06, + "loss": 1.8494, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.5057711601257324, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8663, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4447145462036133, + "learning_rate": 4.744110479285135e-06, + "loss": 1.7532, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.769984483718872, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7755, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.472193956375122, + "learning_rate": 4.740048740861089e-06, + "loss": 1.8056, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.4286932945251465, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8094, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.0248074531555176, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8497, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.607700824737549, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8741, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 1.9779797792434692, + "eval_runtime": 43.7734, + "eval_samples_per_second": 22.845, + "eval_steps_per_second": 0.959, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 3.035045623779297, + "learning_rate": 4.731925264012998e-06, + "loss": 1.8438, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.786921977996826, + "learning_rate": 4.729894394800975e-06, + "loss": 1.7666, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.4158425331115723, + "learning_rate": 4.727863525588953e-06, + "loss": 1.7464, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7772, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 2.6750917434692383, + "learning_rate": 4.725832656376929e-06, + "loss": 1.8447, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 3.08266282081604, + "learning_rate": 4.723801787164907e-06, + "loss": 1.7812, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.791749954223633, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8047, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6457693576812744, + "learning_rate": 4.719740048740861e-06, + "loss": 1.836, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.69480037689209, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8145, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.6356561183929443, + "learning_rate": 4.715678310316816e-06, + "loss": 1.7756, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 1.9723366498947144, + "eval_runtime": 45.5806, + "eval_samples_per_second": 21.939, + "eval_steps_per_second": 0.921, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.53474497795105, + "learning_rate": 4.713647441104794e-06, + "loss": 1.774, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.3631107807159424, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8517, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.594604730606079, + "learning_rate": 4.709585702680748e-06, + "loss": 1.8962, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.5280940532684326, + "learning_rate": 4.707554833468725e-06, + "loss": 1.7759, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.4596288204193115, + "learning_rate": 4.705523964256703e-06, + "loss": 1.7948, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.9123644828796387, + "learning_rate": 4.703493095044679e-06, + "loss": 1.7399, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.4354264736175537, + "learning_rate": 4.701462225832657e-06, + "loss": 1.8659, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.28625226020813, + "learning_rate": 4.699431356620634e-06, + "loss": 1.831, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.5445950031280518, + "learning_rate": 4.697400487408612e-06, + "loss": 1.8121, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.767825126647949, + "learning_rate": 4.695369618196588e-06, + "loss": 1.7874, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9653255939483643, + "eval_runtime": 45.2453, + "eval_samples_per_second": 22.102, + "eval_steps_per_second": 0.928, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4052765369415283, + "learning_rate": 4.693338748984566e-06, + "loss": 1.7539, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.498098373413086, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8343, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.821753740310669, + "learning_rate": 4.68927701056052e-06, + "loss": 1.8119, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.4158449172973633, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.7982, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.5229434967041016, + "learning_rate": 4.685215272136475e-06, + "loss": 1.806, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.729199171066284, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8768, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.5102570056915283, + "learning_rate": 4.681153533712429e-06, + "loss": 1.775, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4463515281677246, + "learning_rate": 4.679122664500407e-06, + "loss": 1.7493, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.505733013153076, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8462, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.5784912109375, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7381, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9640673398971558, + "eval_runtime": 44.2714, + "eval_samples_per_second": 22.588, + "eval_steps_per_second": 0.949, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.7864291667938232, + "learning_rate": 4.673030056864338e-06, + "loss": 1.7612, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.507809638977051, + "learning_rate": 4.670999187652316e-06, + "loss": 1.757, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.5499937534332275, + "learning_rate": 4.668968318440293e-06, + "loss": 1.8243, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.435472011566162, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7269, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.541031837463379, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.7959, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5052099227905273, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7191, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.6285109519958496, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8423, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.7825605869293213, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7178, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.677319288253784, + "learning_rate": 4.656783103168157e-06, + "loss": 1.7939, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.6830129623413086, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7842, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.960136890411377, + "eval_runtime": 45.9379, + "eval_samples_per_second": 21.769, + "eval_steps_per_second": 0.914, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.5368666648864746, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8352, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.8922412395477295, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.8252, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.8332598209381104, + "learning_rate": 4.648659626320066e-06, + "loss": 1.7805, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.420780658721924, + "learning_rate": 4.646628757108043e-06, + "loss": 1.7499, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.6684062480926514, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.7857, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.8547494411468506, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.8115, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.59279203414917, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8254, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.80961275100708, + "learning_rate": 4.638505280259952e-06, + "loss": 1.7901, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.9017655849456787, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8265, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.595454216003418, + "learning_rate": 4.634443541835907e-06, + "loss": 1.7833, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9576054811477661, + "eval_runtime": 50.7265, + "eval_samples_per_second": 19.714, + "eval_steps_per_second": 0.828, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 4.260303020477295, + "learning_rate": 4.632412672623883e-06, + "loss": 1.6741, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.366147756576538, + "learning_rate": 4.630381803411861e-06, + "loss": 1.7778, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.894336462020874, + "learning_rate": 4.628350934199838e-06, + "loss": 1.8268, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.7675349712371826, + "learning_rate": 4.626320064987816e-06, + "loss": 1.7748, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.461413621902466, + "learning_rate": 4.624289195775792e-06, + "loss": 1.7599, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.670187473297119, + "learning_rate": 4.62225832656377e-06, + "loss": 1.7736, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.8119454383850098, + "learning_rate": 4.620227457351747e-06, + "loss": 1.7837, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.8020036220550537, + "learning_rate": 4.618196588139724e-06, + "loss": 1.7508, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.5977094173431396, + "learning_rate": 4.616165718927701e-06, + "loss": 1.834, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.731919765472412, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.8137, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.9465162754058838, + "eval_runtime": 48.9829, + "eval_samples_per_second": 20.415, + "eval_steps_per_second": 0.857, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.604386568069458, + "learning_rate": 4.612103980503656e-06, + "loss": 1.7854, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.601107120513916, + "learning_rate": 4.610073111291633e-06, + "loss": 1.7749, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.6758370399475098, + "learning_rate": 4.60804224207961e-06, + "loss": 1.7826, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.631549119949341, + "learning_rate": 4.606011372867588e-06, + "loss": 1.7723, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.6704764366149902, + "learning_rate": 4.603980503655565e-06, + "loss": 1.7088, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 3.078573703765869, + "learning_rate": 4.601949634443542e-06, + "loss": 1.7254, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.7492356300354004, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7756, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 3.0763585567474365, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7806, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.5379223823547363, + "learning_rate": 4.595857026807474e-06, + "loss": 1.7516, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.8343496322631836, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8111, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.945749282836914, + "eval_runtime": 46.5453, + "eval_samples_per_second": 21.484, + "eval_steps_per_second": 0.902, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.809753179550171, + "learning_rate": 4.591795288383429e-06, + "loss": 1.7877, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.4196107387542725, + "learning_rate": 4.589764419171406e-06, + "loss": 1.7457, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.409287452697754, + "learning_rate": 4.587733549959383e-06, + "loss": 1.7338, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.414706230163574, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7185, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.591365337371826, + "learning_rate": 4.583671811535338e-06, + "loss": 1.8083, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.4648489952087402, + "learning_rate": 4.581640942323315e-06, + "loss": 1.8192, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.3804972171783447, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8077, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.283082962036133, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8294, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.61795711517334, + "learning_rate": 4.575548334687247e-06, + "loss": 1.7322, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.4035353660583496, + "learning_rate": 4.573517465475224e-06, + "loss": 1.7798, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9422866106033325, + "eval_runtime": 45.3068, + "eval_samples_per_second": 22.072, + "eval_steps_per_second": 0.927, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.7253756523132324, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7876, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.738295078277588, + "learning_rate": 4.569455727051178e-06, + "loss": 1.7933, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 3.1418697834014893, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8145, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.941603660583496, + "learning_rate": 4.565393988627132e-06, + "loss": 1.7671, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 3.063924551010132, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.7809, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 4.957335472106934, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7306, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.651716709136963, + "learning_rate": 4.559301380991065e-06, + "loss": 1.7428, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 4.777034282684326, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.8118, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.5549471378326416, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7294, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.474388599395752, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8338, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.9402854442596436, + "eval_runtime": 43.166, + "eval_samples_per_second": 23.166, + "eval_steps_per_second": 0.973, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.8968348503112793, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7432, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.6996726989746094, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.7837, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.4599392414093018, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7388, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.745487928390503, + "learning_rate": 4.545085296506905e-06, + "loss": 1.7439, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.6960670948028564, + "learning_rate": 4.543054427294882e-06, + "loss": 1.737, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.596876621246338, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7708, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.598267078399658, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7541, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.6069958209991455, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7756, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.490694761276245, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.7372, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.44179368019104, + "learning_rate": 4.532900081234769e-06, + "loss": 1.7652, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.934715986251831, + "eval_runtime": 44.688, + "eval_samples_per_second": 22.377, + "eval_steps_per_second": 0.94, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 3.018480062484741, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7596, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.5316643714904785, + "learning_rate": 4.528838342810723e-06, + "loss": 1.7715, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4893109798431396, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7549, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.524834156036377, + "learning_rate": 4.524776604386678e-06, + "loss": 1.8046, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.7527239322662354, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7583, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.5210752487182617, + "learning_rate": 4.520714865962632e-06, + "loss": 1.7896, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.5787770748138428, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7355, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.537195920944214, + "learning_rate": 4.516653127538587e-06, + "loss": 1.7272, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.6810014247894287, + "learning_rate": 4.514622258326564e-06, + "loss": 1.753, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.5868756771087646, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7964, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.934499740600586, + "eval_runtime": 45.6556, + "eval_samples_per_second": 21.903, + "eval_steps_per_second": 0.92, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.6830856800079346, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7072, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.6574594974517822, + "learning_rate": 4.508529650690496e-06, + "loss": 1.7004, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.4415769577026367, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7667, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.825115442276001, + "learning_rate": 4.504467912266451e-06, + "loss": 1.8356, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.949010133743286, + "learning_rate": 4.502437043054428e-06, + "loss": 1.7579, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.690986394882202, + "learning_rate": 4.500406173842405e-06, + "loss": 1.7657, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 3.1015520095825195, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.7432, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.608424186706543, + "learning_rate": 4.49634443541836e-06, + "loss": 1.8112, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.5238773822784424, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7724, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.7712161540985107, + "learning_rate": 4.492282696994314e-06, + "loss": 1.7111, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9326083660125732, + "eval_runtime": 48.5359, + "eval_samples_per_second": 20.603, + "eval_steps_per_second": 0.865, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.4763033390045166, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7268, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.6310081481933594, + "learning_rate": 4.488220958570269e-06, + "loss": 1.6885, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.5950350761413574, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7775, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.769026517868042, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7737, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.4898502826690674, + "learning_rate": 4.4821283509342e-06, + "loss": 1.751, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.668081521987915, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7178, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.441615104675293, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7511, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.727665901184082, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.8639, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.7964634895324707, + "learning_rate": 4.474004874086109e-06, + "loss": 1.6945, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.6142687797546387, + "learning_rate": 4.471974004874087e-06, + "loss": 1.8028, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9281476736068726, + "eval_runtime": 48.078, + "eval_samples_per_second": 20.8, + "eval_steps_per_second": 0.874, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.6655778884887695, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.7632, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.931162118911743, + "learning_rate": 4.467912266450041e-06, + "loss": 1.7244, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.469895124435425, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7773, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.4948830604553223, + "learning_rate": 4.463850528025995e-06, + "loss": 1.788, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5961945056915283, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.756, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.452946424484253, + "learning_rate": 4.45978878960195e-06, + "loss": 1.6968, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.9352009296417236, + "learning_rate": 4.457757920389927e-06, + "loss": 1.7414, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.7044763565063477, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7848, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.7200324535369873, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7699, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.6185197830200195, + "learning_rate": 4.451665312753859e-06, + "loss": 1.7816, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9209110736846924, + "eval_runtime": 47.3321, + "eval_samples_per_second": 21.127, + "eval_steps_per_second": 0.887, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.3639564514160156, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7164, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.7073519229888916, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.727, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.637080430984497, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7951, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.8013885021209717, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7333, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.512484312057495, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8048, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 3.3851466178894043, + "learning_rate": 4.439480097481723e-06, + "loss": 1.6667, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.8105392456054688, + "learning_rate": 4.4374492282697e-06, + "loss": 1.7622, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.555773973464966, + "learning_rate": 4.435418359057677e-06, + "loss": 1.7519, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.688596725463867, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7198, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.828646421432495, + "learning_rate": 4.431356620633632e-06, + "loss": 1.7478, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.919256567955017, + "eval_runtime": 50.5993, + "eval_samples_per_second": 19.763, + "eval_steps_per_second": 0.83, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.4523651599884033, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7344, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.4237422943115234, + "learning_rate": 4.427294882209586e-06, + "loss": 1.6664, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 3.2746193408966064, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.7446, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.502819061279297, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7691, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.7032017707824707, + "learning_rate": 4.421202274573518e-06, + "loss": 1.6448, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.5879440307617188, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.6908, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 3.2481982707977295, + "learning_rate": 4.417140536149473e-06, + "loss": 1.7725, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.4844820499420166, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7141, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.5596253871917725, + "learning_rate": 4.413078797725427e-06, + "loss": 1.7485, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 3.142839193344116, + "learning_rate": 4.411047928513404e-06, + "loss": 1.734, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9129430055618286, + "eval_runtime": 49.2608, + "eval_samples_per_second": 20.3, + "eval_steps_per_second": 0.853, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.5837552547454834, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7668, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.7043633460998535, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7497, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.9886765480041504, + "learning_rate": 4.404955320877336e-06, + "loss": 1.7082, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.3199501037597656, + "learning_rate": 4.402924451665313e-06, + "loss": 1.693, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.606682777404785, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7242, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.6967978477478027, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7516, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.351726531982422, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7338, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.590869188308716, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7462, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.3498220443725586, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7609, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.281002998352051, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7584, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.908841609954834, + "eval_runtime": 50.7771, + "eval_samples_per_second": 19.694, + "eval_steps_per_second": 0.827, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.468755006790161, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7654, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.3467307090759277, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7923, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.3723301887512207, + "learning_rate": 4.384646628757108e-06, + "loss": 1.74, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.4527792930603027, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.6988, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.3923704624176025, + "learning_rate": 4.380584890333063e-06, + "loss": 1.6406, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 2.740556240081787, + "learning_rate": 4.37855402112104e-06, + "loss": 1.8032, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.3013832569122314, + "learning_rate": 4.376523151909017e-06, + "loss": 1.7259, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.4632391929626465, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.7462, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.4878177642822266, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7665, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.610898494720459, + "learning_rate": 4.370430544272949e-06, + "loss": 1.7646, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9056707620620728, + "eval_runtime": 51.8738, + "eval_samples_per_second": 19.278, + "eval_steps_per_second": 0.81, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.387721300125122, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7454, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.9966042041778564, + "learning_rate": 4.366368805848904e-06, + "loss": 1.6789, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.664689064025879, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7486, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.4788103103637695, + "learning_rate": 4.362307067424858e-06, + "loss": 1.6981, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.4413328170776367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7279, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.6818156242370605, + "learning_rate": 4.358245329000813e-06, + "loss": 1.74, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.650348424911499, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7212, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.442805767059326, + "learning_rate": 4.354183590576767e-06, + "loss": 1.6951, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.6806585788726807, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7499, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.586792230606079, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7709, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.901674747467041, + "eval_runtime": 42.3186, + "eval_samples_per_second": 23.63, + "eval_steps_per_second": 0.992, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.547452688217163, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7458, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.5820348262786865, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.7232, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.4208321571350098, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7981, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.8791956901550293, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7372, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.6537868976593018, + "learning_rate": 4.339967506092607e-06, + "loss": 1.7216, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.416097402572632, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.701, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.5565483570098877, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7558, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.5875823497772217, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7439, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5459702014923096, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.7058, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.4195566177368164, + "learning_rate": 4.329813160032495e-06, + "loss": 1.6749, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9035066366195679, + "eval_runtime": 43.7962, + "eval_samples_per_second": 22.833, + "eval_steps_per_second": 0.959, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.6019818782806396, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7549, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.8627779483795166, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7257, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.582158088684082, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7138, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.4528613090515137, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7162, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 3.1956026554107666, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7556, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.3531901836395264, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.746, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.441988706588745, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7664, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.599506378173828, + "learning_rate": 4.313566206336312e-06, + "loss": 1.6955, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.4121932983398438, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7194, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.4803290367126465, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7386, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.8993914127349854, + "eval_runtime": 43.0525, + "eval_samples_per_second": 23.227, + "eval_steps_per_second": 0.976, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.701594114303589, + "learning_rate": 4.307473598700244e-06, + "loss": 1.6796, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.7144737243652344, + "learning_rate": 4.305442729488221e-06, + "loss": 1.763, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.395312786102295, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7522, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.3926732540130615, + "learning_rate": 4.301380991064176e-06, + "loss": 1.7705, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.6293702125549316, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7693, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.3746306896209717, + "learning_rate": 4.29731925264013e-06, + "loss": 1.72, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.5421926975250244, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.6868, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.6991186141967773, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7378, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.463897705078125, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7628, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5139036178588867, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7125, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.8909698724746704, + "eval_runtime": 42.5917, + "eval_samples_per_second": 23.479, + "eval_steps_per_second": 0.986, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.5220465660095215, + "learning_rate": 4.287164906580017e-06, + "loss": 1.6975, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.432988405227661, + "learning_rate": 4.285134037367994e-06, + "loss": 1.6904, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.5894253253936768, + "learning_rate": 4.283103168155971e-06, + "loss": 1.6919, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5334458351135254, + "learning_rate": 4.281072298943948e-06, + "loss": 1.6968, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.510019302368164, + "learning_rate": 4.279041429731926e-06, + "loss": 1.6791, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.531538248062134, + "learning_rate": 4.277010560519903e-06, + "loss": 1.723, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.5621440410614014, + "learning_rate": 4.27497969130788e-06, + "loss": 1.749, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.594329833984375, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7579, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.321031093597412, + "learning_rate": 4.270917952883835e-06, + "loss": 1.6983, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.6325764656066895, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7872, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.886696219444275, + "eval_runtime": 43.4143, + "eval_samples_per_second": 23.034, + "eval_steps_per_second": 0.967, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5342679023742676, + "learning_rate": 4.266856214459789e-06, + "loss": 1.6655, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.5810720920562744, + "learning_rate": 4.264825345247766e-06, + "loss": 1.7182, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.7768406867980957, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6534, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.5680463314056396, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7243, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7162973880767822, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.7146, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.697925090789795, + "learning_rate": 4.256701868399675e-06, + "loss": 1.7296, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.4458425045013428, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7466, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.6330654621124268, + "learning_rate": 4.252640129975629e-06, + "loss": 1.6949, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.350430727005005, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.6574, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.505995512008667, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7111, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.8887447118759155, + "eval_runtime": 43.9304, + "eval_samples_per_second": 22.763, + "eval_steps_per_second": 0.956, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 3.036954879760742, + "learning_rate": 4.246547522339562e-06, + "loss": 1.7182, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.563606023788452, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.6566, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.516315460205078, + "learning_rate": 4.242485783915517e-06, + "loss": 1.6543, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.6289877891540527, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6996, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.7563092708587646, + "learning_rate": 4.23842404549147e-06, + "loss": 1.666, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5743000507354736, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7338, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 3.1975982189178467, + "learning_rate": 4.234362307067425e-06, + "loss": 1.7444, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.6022894382476807, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7367, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.4466192722320557, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.726, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.579015016555786, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7146, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.8904188871383667, + "eval_runtime": 43.8091, + "eval_samples_per_second": 22.826, + "eval_steps_per_second": 0.959, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.719646453857422, + "learning_rate": 4.226238830219334e-06, + "loss": 1.7209, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.4423880577087402, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7288, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5883753299713135, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7089, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.447049140930176, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7095, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.453739881515503, + "learning_rate": 4.218115353371243e-06, + "loss": 1.6747, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4540836811065674, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7578, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.5468671321868896, + "learning_rate": 4.214053614947198e-06, + "loss": 1.6556, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.443814277648926, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7307, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.4998748302459717, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6768, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.667025566101074, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7493, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.8830010890960693, + "eval_runtime": 44.7027, + "eval_samples_per_second": 22.37, + "eval_steps_per_second": 0.94, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.416248321533203, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7381, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.378025770187378, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7335, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.633173704147339, + "learning_rate": 4.201868399675061e-06, + "loss": 1.6484, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.549384117126465, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7143, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.6492598056793213, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6907, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.444086790084839, + "learning_rate": 4.195775792038993e-06, + "loss": 1.636, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.628814935684204, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7342, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.83831524848938, + "learning_rate": 4.191714053614948e-06, + "loss": 1.7028, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.467339515686035, + "learning_rate": 4.189683184402924e-06, + "loss": 1.749, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.4529521465301514, + "learning_rate": 4.187652315190902e-06, + "loss": 1.682, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.8826342821121216, + "eval_runtime": 44.7757, + "eval_samples_per_second": 22.334, + "eval_steps_per_second": 0.938, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.7340762615203857, + "learning_rate": 4.185621445978879e-06, + "loss": 1.723, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.6028647422790527, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7566, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.3967695236206055, + "learning_rate": 4.181559707554833e-06, + "loss": 1.683, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.5475900173187256, + "learning_rate": 4.179528838342811e-06, + "loss": 1.6976, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.5411484241485596, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7489, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 3.458749294281006, + "learning_rate": 4.175467099918766e-06, + "loss": 1.6828, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.2977802753448486, + "learning_rate": 4.173436230706742e-06, + "loss": 1.6963, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.875307559967041, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.6857, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 4.097696304321289, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7225, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.6774847507476807, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7031, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.8783056735992432, + "eval_runtime": 44.0219, + "eval_samples_per_second": 22.716, + "eval_steps_per_second": 0.954, + "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_local3_10M_seed0/runs/checkpoint-700/zero_to_fp32.py b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-700/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/babylm_shuffle_local3_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-900/rng_state_0.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..5a018a68bd5303fcda57dcb6093948dc236b697a --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:34bfd29981738a04d7cb9b0b31f032749b5bd5fa1b120104d856a6f5b060af88 +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/rng_state_1.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..85d20c3dd29d7f71371d2ce18d57829a2ebd22bc --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9e7ca9eed1d152be37852a8850dde9294edfaf37097b42c75c54b278336ef61 +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/rng_state_2.pth b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..137d87c143da2b66811890c8cf1e8b013ef6eacf --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:62c45c6121ed4cfe9395ee5153d26d376e4f02f8dfa1bbceafb8a5416efcdd14 +size 16567 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/scheduler.pt b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..ebbc1e184a8a5d0eed24cb8ab7f032ba7496d97c --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ce99e1e4c08fd1d968baa3fd802f16cedd70dbf02bc87cb1076f41f9837387df +size 627 diff --git a/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/tokenizer.json b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..309789d5ccd06f418fbcfb7472a6a9affda68101 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/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-900/training_args.bin b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..6db59a81c34edebb00ef6a59f5669f3babfbb622 --- /dev/null +++ b/babylm_shuffle_local3_10M_seed0/runs/checkpoint-900/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:484dadb3af9420a36d698e37e41887c1bddbd89850c9b141ca33808576dc9968 +size 6011